jpn-jpan-10mb-after-ppt-Dp-100mb-ckpt500_seed455/checkpoint-21000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.423542989036353,
  "eval_steps": 3000,
  "global_step": 21000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 8.164552354812622,
      "epoch": 0.0005770340450086555,
      "grad_norm": 1.03125,
      "learning_rate": 2e-06,
      "loss": 7.7235,
      "mean_token_accuracy": 0.13572018668055535,
      "num_tokens": 13337.0,
      "step": 5
    },
    {
      "entropy": 8.246073150634766,
      "epoch": 0.001154068090017311,
      "grad_norm": 0.97265625,
      "learning_rate": 4.5e-06,
      "loss": 7.8919,
      "mean_token_accuracy": 0.1187807373702526,
      "num_tokens": 26339.0,
      "step": 10
    },
    {
      "entropy": 8.20343017578125,
      "epoch": 0.0017311021350259665,
      "grad_norm": 1.0703125,
      "learning_rate": 7e-06,
      "loss": 7.8009,
      "mean_token_accuracy": 0.12226058542728424,
      "num_tokens": 39118.0,
      "step": 15
    },
    {
      "entropy": 8.200080871582031,
      "epoch": 0.002308136180034622,
      "grad_norm": 1.015625,
      "learning_rate": 9.5e-06,
      "loss": 7.9304,
      "mean_token_accuracy": 0.11691192910075188,
      "num_tokens": 51367.0,
      "step": 20
    },
    {
      "entropy": 8.19017915725708,
      "epoch": 0.0028851702250432777,
      "grad_norm": 1.0234375,
      "learning_rate": 1.2e-05,
      "loss": 7.8692,
      "mean_token_accuracy": 0.12374739944934846,
      "num_tokens": 64639.0,
      "step": 25
    },
    {
      "entropy": 8.127242851257325,
      "epoch": 0.003462204270051933,
      "grad_norm": 0.9453125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 7.8742,
      "mean_token_accuracy": 0.12418468594551087,
      "num_tokens": 76244.0,
      "step": 30
    },
    {
      "entropy": 8.187537956237794,
      "epoch": 0.004039238315060588,
      "grad_norm": 0.86328125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.912,
      "mean_token_accuracy": 0.11309220269322395,
      "num_tokens": 88490.0,
      "step": 35
    },
    {
      "entropy": 8.093108320236206,
      "epoch": 0.004616272360069244,
      "grad_norm": 1.1015625,
      "learning_rate": 1.95e-05,
      "loss": 7.8296,
      "mean_token_accuracy": 0.12792499735951424,
      "num_tokens": 102022.0,
      "step": 40
    },
    {
      "entropy": 8.102341556549073,
      "epoch": 0.0051933064050779,
      "grad_norm": 0.83203125,
      "learning_rate": 2.2e-05,
      "loss": 7.8503,
      "mean_token_accuracy": 0.1249616153538227,
      "num_tokens": 114442.0,
      "step": 45
    },
    {
      "entropy": 8.1214994430542,
      "epoch": 0.005770340450086555,
      "grad_norm": 0.8828125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.9016,
      "mean_token_accuracy": 0.11642022728919983,
      "num_tokens": 128044.0,
      "step": 50
    },
    {
      "entropy": 8.168946361541748,
      "epoch": 0.006347374495095211,
      "grad_norm": 0.88671875,
      "learning_rate": 2.7e-05,
      "loss": 7.9196,
      "mean_token_accuracy": 0.11260023415088653,
      "num_tokens": 141818.0,
      "step": 55
    },
    {
      "entropy": 8.143735218048096,
      "epoch": 0.006924408540103866,
      "grad_norm": 0.88671875,
      "learning_rate": 2.95e-05,
      "loss": 7.8821,
      "mean_token_accuracy": 0.12241782322525978,
      "num_tokens": 155109.0,
      "step": 60
    },
    {
      "entropy": 8.149381828308105,
      "epoch": 0.0075014425851125215,
      "grad_norm": 0.875,
      "learning_rate": 3.2e-05,
      "loss": 7.8244,
      "mean_token_accuracy": 0.12313178777694703,
      "num_tokens": 167036.0,
      "step": 65
    },
    {
      "entropy": 8.136874580383301,
      "epoch": 0.008078476630121177,
      "grad_norm": 0.95703125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 7.9797,
      "mean_token_accuracy": 0.1158173181116581,
      "num_tokens": 181398.0,
      "step": 70
    },
    {
      "entropy": 8.139523601531982,
      "epoch": 0.008655510675129832,
      "grad_norm": 0.96875,
      "learning_rate": 3.7e-05,
      "loss": 7.8314,
      "mean_token_accuracy": 0.12376198023557664,
      "num_tokens": 193242.0,
      "step": 75
    },
    {
      "entropy": 8.148941707611083,
      "epoch": 0.009232544720138488,
      "grad_norm": 0.85546875,
      "learning_rate": 3.95e-05,
      "loss": 7.8335,
      "mean_token_accuracy": 0.11901628077030182,
      "num_tokens": 206744.0,
      "step": 80
    },
    {
      "entropy": 8.112703990936279,
      "epoch": 0.009809578765147143,
      "grad_norm": 0.78515625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 7.7994,
      "mean_token_accuracy": 0.1257437027990818,
      "num_tokens": 219120.0,
      "step": 85
    },
    {
      "entropy": 8.163854217529297,
      "epoch": 0.0103866128101558,
      "grad_norm": 0.9375,
      "learning_rate": 4.45e-05,
      "loss": 7.874,
      "mean_token_accuracy": 0.12007838636636733,
      "num_tokens": 231942.0,
      "step": 90
    },
    {
      "entropy": 8.188548183441162,
      "epoch": 0.010963646855164455,
      "grad_norm": 0.85546875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 7.8636,
      "mean_token_accuracy": 0.11973497197031975,
      "num_tokens": 244786.0,
      "step": 95
    },
    {
      "entropy": 8.129443168640137,
      "epoch": 0.01154068090017311,
      "grad_norm": 0.84765625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.7706,
      "mean_token_accuracy": 0.13497541323304177,
      "num_tokens": 257503.0,
      "step": 100
    },
    {
      "entropy": 8.024968385696411,
      "epoch": 0.012117714945181766,
      "grad_norm": 0.875,
      "learning_rate": 5.2e-05,
      "loss": 7.7253,
      "mean_token_accuracy": 0.12938223108649255,
      "num_tokens": 270330.0,
      "step": 105
    },
    {
      "entropy": 8.047209739685059,
      "epoch": 0.012694748990190421,
      "grad_norm": 0.953125,
      "learning_rate": 5.45e-05,
      "loss": 7.8195,
      "mean_token_accuracy": 0.13103519529104232,
      "num_tokens": 283866.0,
      "step": 110
    },
    {
      "entropy": 8.176537418365479,
      "epoch": 0.013271783035199077,
      "grad_norm": 0.84375,
      "learning_rate": 5.7e-05,
      "loss": 7.7892,
      "mean_token_accuracy": 0.11941086724400521,
      "num_tokens": 296715.0,
      "step": 115
    },
    {
      "entropy": 8.116231632232665,
      "epoch": 0.013848817080207732,
      "grad_norm": 0.85546875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.8106,
      "mean_token_accuracy": 0.12553205117583274,
      "num_tokens": 309268.0,
      "step": 120
    },
    {
      "entropy": 8.064244985580444,
      "epoch": 0.014425851125216388,
      "grad_norm": 1.15625,
      "learning_rate": 6.2e-05,
      "loss": 7.7923,
      "mean_token_accuracy": 0.13402299582958221,
      "num_tokens": 322471.0,
      "step": 125
    },
    {
      "entropy": 8.130724334716797,
      "epoch": 0.015002885170225043,
      "grad_norm": 0.9296875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.8188,
      "mean_token_accuracy": 0.12116066142916679,
      "num_tokens": 335728.0,
      "step": 130
    },
    {
      "entropy": 8.112252330780029,
      "epoch": 0.015579919215233698,
      "grad_norm": 0.95703125,
      "learning_rate": 6.7e-05,
      "loss": 7.7868,
      "mean_token_accuracy": 0.12612647488713263,
      "num_tokens": 347861.0,
      "step": 135
    },
    {
      "entropy": 8.101672506332397,
      "epoch": 0.016156953260242354,
      "grad_norm": 1.0703125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.8175,
      "mean_token_accuracy": 0.12923522740602494,
      "num_tokens": 360275.0,
      "step": 140
    },
    {
      "entropy": 8.119850635528564,
      "epoch": 0.01673398730525101,
      "grad_norm": 0.875,
      "learning_rate": 7.2e-05,
      "loss": 7.7957,
      "mean_token_accuracy": 0.12403815910220146,
      "num_tokens": 373948.0,
      "step": 145
    },
    {
      "entropy": 8.126962089538575,
      "epoch": 0.017311021350259664,
      "grad_norm": 0.9140625,
      "learning_rate": 7.45e-05,
      "loss": 7.7728,
      "mean_token_accuracy": 0.12742806524038314,
      "num_tokens": 386267.0,
      "step": 150
    },
    {
      "entropy": 8.088498830795288,
      "epoch": 0.01788805539526832,
      "grad_norm": 0.890625,
      "learning_rate": 7.7e-05,
      "loss": 7.7321,
      "mean_token_accuracy": 0.12699732854962348,
      "num_tokens": 398595.0,
      "step": 155
    },
    {
      "entropy": 8.019657421112061,
      "epoch": 0.018465089440276975,
      "grad_norm": 0.93359375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.7922,
      "mean_token_accuracy": 0.12140516936779022,
      "num_tokens": 410626.0,
      "step": 160
    },
    {
      "entropy": 8.144418334960937,
      "epoch": 0.019042123485285632,
      "grad_norm": 0.91015625,
      "learning_rate": 8.2e-05,
      "loss": 7.8247,
      "mean_token_accuracy": 0.12391015514731407,
      "num_tokens": 423176.0,
      "step": 165
    },
    {
      "entropy": 8.170842695236207,
      "epoch": 0.019619157530294286,
      "grad_norm": 0.96875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.7815,
      "mean_token_accuracy": 0.12869809195399284,
      "num_tokens": 435997.0,
      "step": 170
    },
    {
      "entropy": 8.032562351226806,
      "epoch": 0.020196191575302943,
      "grad_norm": 0.8671875,
      "learning_rate": 8.7e-05,
      "loss": 7.7859,
      "mean_token_accuracy": 0.1265020787715912,
      "num_tokens": 449295.0,
      "step": 175
    },
    {
      "entropy": 8.155485820770263,
      "epoch": 0.0207732256203116,
      "grad_norm": 0.88671875,
      "learning_rate": 8.95e-05,
      "loss": 7.805,
      "mean_token_accuracy": 0.12606418877840042,
      "num_tokens": 463983.0,
      "step": 180
    },
    {
      "entropy": 8.100854730606079,
      "epoch": 0.021350259665320254,
      "grad_norm": 0.9140625,
      "learning_rate": 9.2e-05,
      "loss": 7.814,
      "mean_token_accuracy": 0.12124907821416855,
      "num_tokens": 476849.0,
      "step": 185
    },
    {
      "entropy": 8.118375110626221,
      "epoch": 0.02192729371032891,
      "grad_norm": 0.98046875,
      "learning_rate": 9.45e-05,
      "loss": 7.8255,
      "mean_token_accuracy": 0.12398237809538841,
      "num_tokens": 489120.0,
      "step": 190
    },
    {
      "entropy": 8.123465538024902,
      "epoch": 0.022504327755337564,
      "grad_norm": 0.94140625,
      "learning_rate": 9.7e-05,
      "loss": 7.7552,
      "mean_token_accuracy": 0.1278926372528076,
      "num_tokens": 501085.0,
      "step": 195
    },
    {
      "entropy": 8.084616231918336,
      "epoch": 0.02308136180034622,
      "grad_norm": 0.9375,
      "learning_rate": 9.95e-05,
      "loss": 7.8169,
      "mean_token_accuracy": 0.1159471996128559,
      "num_tokens": 513207.0,
      "step": 200
    },
    {
      "entropy": 8.162526893615723,
      "epoch": 0.023658395845354875,
      "grad_norm": 1.015625,
      "learning_rate": 0.000102,
      "loss": 7.7468,
      "mean_token_accuracy": 0.11916335672140121,
      "num_tokens": 524809.0,
      "step": 205
    },
    {
      "entropy": 8.082560157775879,
      "epoch": 0.024235429890363532,
      "grad_norm": 1.140625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.8136,
      "mean_token_accuracy": 0.12392313480377197,
      "num_tokens": 538950.0,
      "step": 210
    },
    {
      "entropy": 8.15697078704834,
      "epoch": 0.024812463935372186,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000107,
      "loss": 7.793,
      "mean_token_accuracy": 0.13155161514878272,
      "num_tokens": 551199.0,
      "step": 215
    },
    {
      "entropy": 8.045954370498658,
      "epoch": 0.025389497980380843,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001095,
      "loss": 7.731,
      "mean_token_accuracy": 0.12966038063168525,
      "num_tokens": 563202.0,
      "step": 220
    },
    {
      "entropy": 8.128791332244873,
      "epoch": 0.025966532025389497,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000112,
      "loss": 7.7501,
      "mean_token_accuracy": 0.12665933594107628,
      "num_tokens": 576716.0,
      "step": 225
    },
    {
      "entropy": 8.056756114959716,
      "epoch": 0.026543566070398154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001145,
      "loss": 7.7642,
      "mean_token_accuracy": 0.12589010149240493,
      "num_tokens": 588413.0,
      "step": 230
    },
    {
      "entropy": 8.159916019439697,
      "epoch": 0.027120600115406807,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.8143,
      "mean_token_accuracy": 0.12119406759738922,
      "num_tokens": 602052.0,
      "step": 235
    },
    {
      "entropy": 8.061726570129395,
      "epoch": 0.027697634160415464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.7646,
      "mean_token_accuracy": 0.12605880349874496,
      "num_tokens": 614508.0,
      "step": 240
    },
    {
      "entropy": 8.145089435577393,
      "epoch": 0.02827466820542412,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000122,
      "loss": 7.8077,
      "mean_token_accuracy": 0.12516265287995337,
      "num_tokens": 628758.0,
      "step": 245
    },
    {
      "entropy": 8.007381439208984,
      "epoch": 0.028851702250432775,
      "grad_norm": 1.203125,
      "learning_rate": 0.0001245,
      "loss": 7.7227,
      "mean_token_accuracy": 0.1311620891094208,
      "num_tokens": 641046.0,
      "step": 250
    },
    {
      "entropy": 8.058531141281128,
      "epoch": 0.029428736295441432,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000127,
      "loss": 7.7143,
      "mean_token_accuracy": 0.12785313203930854,
      "num_tokens": 653910.0,
      "step": 255
    },
    {
      "entropy": 8.076497840881348,
      "epoch": 0.030005770340450086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001295,
      "loss": 7.7051,
      "mean_token_accuracy": 0.13272143229842187,
      "num_tokens": 665862.0,
      "step": 260
    },
    {
      "entropy": 7.979534912109375,
      "epoch": 0.030582804385458743,
      "grad_norm": 1.234375,
      "learning_rate": 0.000132,
      "loss": 7.7153,
      "mean_token_accuracy": 0.1262400984764099,
      "num_tokens": 678854.0,
      "step": 265
    },
    {
      "entropy": 8.1496657371521,
      "epoch": 0.031159838430467397,
      "grad_norm": 0.953125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.734,
      "mean_token_accuracy": 0.12493550479412079,
      "num_tokens": 690188.0,
      "step": 270
    },
    {
      "entropy": 7.9987037658691404,
      "epoch": 0.031736872475476054,
      "grad_norm": 1.140625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.7109,
      "mean_token_accuracy": 0.13045160546898843,
      "num_tokens": 702041.0,
      "step": 275
    },
    {
      "entropy": 8.06501841545105,
      "epoch": 0.03231390652048471,
      "grad_norm": 0.984375,
      "learning_rate": 0.0001395,
      "loss": 7.7275,
      "mean_token_accuracy": 0.1316023886203766,
      "num_tokens": 714109.0,
      "step": 280
    },
    {
      "entropy": 8.041949319839478,
      "epoch": 0.03289094056549336,
      "grad_norm": 1.140625,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.7499,
      "mean_token_accuracy": 0.12448728755116463,
      "num_tokens": 726120.0,
      "step": 285
    },
    {
      "entropy": 8.055285263061524,
      "epoch": 0.03346797461050202,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0001445,
      "loss": 7.6926,
      "mean_token_accuracy": 0.12664653435349466,
      "num_tokens": 737868.0,
      "step": 290
    },
    {
      "entropy": 8.153725147247314,
      "epoch": 0.034045008655510675,
      "grad_norm": 1.0,
      "learning_rate": 0.000147,
      "loss": 7.769,
      "mean_token_accuracy": 0.12764903903007507,
      "num_tokens": 751829.0,
      "step": 295
    },
    {
      "entropy": 8.056012821197509,
      "epoch": 0.03462204270051933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001495,
      "loss": 7.7237,
      "mean_token_accuracy": 0.12691396623849868,
      "num_tokens": 764457.0,
      "step": 300
    },
    {
      "entropy": 8.08794288635254,
      "epoch": 0.03519907674552799,
      "grad_norm": 1.0625,
      "learning_rate": 0.000152,
      "loss": 7.7548,
      "mean_token_accuracy": 0.12576484456658363,
      "num_tokens": 776887.0,
      "step": 305
    },
    {
      "entropy": 8.057601928710938,
      "epoch": 0.03577611079053664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.7485,
      "mean_token_accuracy": 0.12234131619334221,
      "num_tokens": 789001.0,
      "step": 310
    },
    {
      "entropy": 8.092532968521118,
      "epoch": 0.0363531448355453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000157,
      "loss": 7.7387,
      "mean_token_accuracy": 0.1253275267779827,
      "num_tokens": 802719.0,
      "step": 315
    },
    {
      "entropy": 8.023380422592163,
      "epoch": 0.03693017888055395,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001595,
      "loss": 7.6701,
      "mean_token_accuracy": 0.13013212531805038,
      "num_tokens": 815064.0,
      "step": 320
    },
    {
      "entropy": 8.050508165359497,
      "epoch": 0.03750721292556261,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000162,
      "loss": 7.7264,
      "mean_token_accuracy": 0.12830724567174911,
      "num_tokens": 827778.0,
      "step": 325
    },
    {
      "entropy": 8.036063766479492,
      "epoch": 0.038084246970571264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.6907,
      "mean_token_accuracy": 0.12928537279367447,
      "num_tokens": 840691.0,
      "step": 330
    },
    {
      "entropy": 8.05320930480957,
      "epoch": 0.03866128101557992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.7446,
      "mean_token_accuracy": 0.11841234192252159,
      "num_tokens": 852818.0,
      "step": 335
    },
    {
      "entropy": 7.989465665817261,
      "epoch": 0.03923831506058857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.7555,
      "mean_token_accuracy": 0.13145181983709336,
      "num_tokens": 866854.0,
      "step": 340
    },
    {
      "entropy": 8.109742450714112,
      "epoch": 0.03981534910559723,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.7113,
      "mean_token_accuracy": 0.12292054072022437,
      "num_tokens": 879234.0,
      "step": 345
    },
    {
      "entropy": 8.068915796279907,
      "epoch": 0.040392383150605886,
      "grad_norm": 0.96875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.6449,
      "mean_token_accuracy": 0.12445924580097198,
      "num_tokens": 890927.0,
      "step": 350
    },
    {
      "entropy": 8.012982606887817,
      "epoch": 0.04096941719561454,
      "grad_norm": 1.109375,
      "learning_rate": 0.000177,
      "loss": 7.6946,
      "mean_token_accuracy": 0.12666598856449127,
      "num_tokens": 902965.0,
      "step": 355
    },
    {
      "entropy": 8.02733473777771,
      "epoch": 0.0415464512406232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001795,
      "loss": 7.6597,
      "mean_token_accuracy": 0.1288968488574028,
      "num_tokens": 916384.0,
      "step": 360
    },
    {
      "entropy": 8.065688943862915,
      "epoch": 0.042123485285631854,
      "grad_norm": 0.921875,
      "learning_rate": 0.000182,
      "loss": 7.5615,
      "mean_token_accuracy": 0.13689733892679215,
      "num_tokens": 928618.0,
      "step": 365
    },
    {
      "entropy": 7.963602304458618,
      "epoch": 0.04270051933064051,
      "grad_norm": 1.375,
      "learning_rate": 0.0001845,
      "loss": 7.6959,
      "mean_token_accuracy": 0.12423681542277336,
      "num_tokens": 940087.0,
      "step": 370
    },
    {
      "entropy": 7.927715587615967,
      "epoch": 0.04327755337564916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000187,
      "loss": 7.6993,
      "mean_token_accuracy": 0.13103626742959024,
      "num_tokens": 953322.0,
      "step": 375
    },
    {
      "entropy": 8.1933123588562,
      "epoch": 0.04385458742065782,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001895,
      "loss": 7.7438,
      "mean_token_accuracy": 0.12380573004484177,
      "num_tokens": 965815.0,
      "step": 380
    },
    {
      "entropy": 7.940394258499145,
      "epoch": 0.044431621465666475,
      "grad_norm": 1.203125,
      "learning_rate": 0.000192,
      "loss": 7.6752,
      "mean_token_accuracy": 0.12669538110494613,
      "num_tokens": 978731.0,
      "step": 385
    },
    {
      "entropy": 8.027307653427124,
      "epoch": 0.04500865551067513,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001945,
      "loss": 7.5735,
      "mean_token_accuracy": 0.13436778634786606,
      "num_tokens": 989505.0,
      "step": 390
    },
    {
      "entropy": 7.926584243774414,
      "epoch": 0.04558568955568378,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.7088,
      "mean_token_accuracy": 0.12935978397727013,
      "num_tokens": 1001888.0,
      "step": 395
    },
    {
      "entropy": 7.996991300582886,
      "epoch": 0.04616272360069244,
      "grad_norm": 1.0,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.6416,
      "mean_token_accuracy": 0.13667909130454065,
      "num_tokens": 1015362.0,
      "step": 400
    },
    {
      "entropy": 8.05469207763672,
      "epoch": 0.0467397576457011,
      "grad_norm": 1.078125,
      "learning_rate": 0.000202,
      "loss": 7.7052,
      "mean_token_accuracy": 0.12611190229654312,
      "num_tokens": 1028700.0,
      "step": 405
    },
    {
      "entropy": 7.959253025054932,
      "epoch": 0.04731679169070975,
      "grad_norm": 0.96875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.7608,
      "mean_token_accuracy": 0.123697529733181,
      "num_tokens": 1040766.0,
      "step": 410
    },
    {
      "entropy": 8.0970055103302,
      "epoch": 0.04789382573571841,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000207,
      "loss": 7.5937,
      "mean_token_accuracy": 0.12767987027764321,
      "num_tokens": 1054321.0,
      "step": 415
    },
    {
      "entropy": 7.88906192779541,
      "epoch": 0.048470859780727064,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002095,
      "loss": 7.6707,
      "mean_token_accuracy": 0.13818718865513802,
      "num_tokens": 1065410.0,
      "step": 420
    },
    {
      "entropy": 8.037972116470337,
      "epoch": 0.04904789382573572,
      "grad_norm": 1.03125,
      "learning_rate": 0.000212,
      "loss": 7.695,
      "mean_token_accuracy": 0.11944770142436027,
      "num_tokens": 1077369.0,
      "step": 425
    },
    {
      "entropy": 8.041876745223998,
      "epoch": 0.04962492787074437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002145,
      "loss": 7.6535,
      "mean_token_accuracy": 0.1305831678211689,
      "num_tokens": 1089065.0,
      "step": 430
    },
    {
      "entropy": 7.918288230895996,
      "epoch": 0.05020196191575303,
      "grad_norm": 1.234375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.5798,
      "mean_token_accuracy": 0.13306143879890442,
      "num_tokens": 1101285.0,
      "step": 435
    },
    {
      "entropy": 7.921914911270141,
      "epoch": 0.050778995960761686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002195,
      "loss": 7.5956,
      "mean_token_accuracy": 0.1294034317135811,
      "num_tokens": 1114001.0,
      "step": 440
    },
    {
      "entropy": 8.043266582489014,
      "epoch": 0.05135603000577034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000222,
      "loss": 7.6698,
      "mean_token_accuracy": 0.12662654593586922,
      "num_tokens": 1126751.0,
      "step": 445
    },
    {
      "entropy": 7.9704231262207035,
      "epoch": 0.05193306405077899,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002245,
      "loss": 7.6779,
      "mean_token_accuracy": 0.125260828435421,
      "num_tokens": 1140737.0,
      "step": 450
    },
    {
      "entropy": 7.907954835891724,
      "epoch": 0.052510098095787654,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.5845,
      "mean_token_accuracy": 0.12618799656629562,
      "num_tokens": 1152569.0,
      "step": 455
    },
    {
      "entropy": 7.992079973220825,
      "epoch": 0.05308713214079631,
      "grad_norm": 0.96875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.6289,
      "mean_token_accuracy": 0.12820862010121345,
      "num_tokens": 1166163.0,
      "step": 460
    },
    {
      "entropy": 7.962404775619507,
      "epoch": 0.05366416618580496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.606,
      "mean_token_accuracy": 0.12718955501914025,
      "num_tokens": 1177647.0,
      "step": 465
    },
    {
      "entropy": 7.939283180236816,
      "epoch": 0.054241200230813615,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.6284,
      "mean_token_accuracy": 0.12961039319634438,
      "num_tokens": 1190884.0,
      "step": 470
    },
    {
      "entropy": 7.944163656234741,
      "epoch": 0.054818234275822275,
      "grad_norm": 1.046875,
      "learning_rate": 0.000237,
      "loss": 7.6643,
      "mean_token_accuracy": 0.12723161578178405,
      "num_tokens": 1203276.0,
      "step": 475
    },
    {
      "entropy": 7.883432531356812,
      "epoch": 0.05539526832083093,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002395,
      "loss": 7.543,
      "mean_token_accuracy": 0.14349528700113295,
      "num_tokens": 1216001.0,
      "step": 480
    },
    {
      "entropy": 8.038554286956787,
      "epoch": 0.05597230236583958,
      "grad_norm": 1.03125,
      "learning_rate": 0.000242,
      "loss": 7.6852,
      "mean_token_accuracy": 0.12811896130442618,
      "num_tokens": 1228724.0,
      "step": 485
    },
    {
      "entropy": 7.979045391082764,
      "epoch": 0.05654933641084824,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002445,
      "loss": 7.6783,
      "mean_token_accuracy": 0.13219430297613144,
      "num_tokens": 1242731.0,
      "step": 490
    },
    {
      "entropy": 7.960198593139649,
      "epoch": 0.0571263704558569,
      "grad_norm": 1.078125,
      "learning_rate": 0.000247,
      "loss": 7.6384,
      "mean_token_accuracy": 0.13144275397062302,
      "num_tokens": 1255275.0,
      "step": 495
    },
    {
      "entropy": 7.924617958068848,
      "epoch": 0.05770340450086555,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002495,
      "loss": 7.6371,
      "mean_token_accuracy": 0.13291670829057695,
      "num_tokens": 1267719.0,
      "step": 500
    },
    {
      "entropy": 7.997011137008667,
      "epoch": 0.058280438545874204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000252,
      "loss": 7.6101,
      "mean_token_accuracy": 0.12541425302624704,
      "num_tokens": 1279960.0,
      "step": 505
    },
    {
      "entropy": 7.934813690185547,
      "epoch": 0.058857472590882864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002545,
      "loss": 7.6068,
      "mean_token_accuracy": 0.1224808193743229,
      "num_tokens": 1292570.0,
      "step": 510
    },
    {
      "entropy": 7.883869075775147,
      "epoch": 0.05943450663589152,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000257,
      "loss": 7.6065,
      "mean_token_accuracy": 0.1239744633436203,
      "num_tokens": 1305548.0,
      "step": 515
    },
    {
      "entropy": 7.938631439208985,
      "epoch": 0.06001154068090017,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002595,
      "loss": 7.5338,
      "mean_token_accuracy": 0.13228121995925904,
      "num_tokens": 1318904.0,
      "step": 520
    },
    {
      "entropy": 7.933946323394776,
      "epoch": 0.060588574725908825,
      "grad_norm": 0.984375,
      "learning_rate": 0.000262,
      "loss": 7.626,
      "mean_token_accuracy": 0.12482062131166458,
      "num_tokens": 1332841.0,
      "step": 525
    },
    {
      "entropy": 7.883927583694458,
      "epoch": 0.061165608770917486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.6283,
      "mean_token_accuracy": 0.12460393533110618,
      "num_tokens": 1345708.0,
      "step": 530
    },
    {
      "entropy": 7.9526472091674805,
      "epoch": 0.06174264281592614,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.5758,
      "mean_token_accuracy": 0.1290095902979374,
      "num_tokens": 1358611.0,
      "step": 535
    },
    {
      "entropy": 7.871510028839111,
      "epoch": 0.06231967686093479,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.6509,
      "mean_token_accuracy": 0.1282058410346508,
      "num_tokens": 1372753.0,
      "step": 540
    },
    {
      "entropy": 7.92236442565918,
      "epoch": 0.06289671090594345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.5316,
      "mean_token_accuracy": 0.13794104903936386,
      "num_tokens": 1383811.0,
      "step": 545
    },
    {
      "entropy": 7.879396390914917,
      "epoch": 0.06347374495095211,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002745,
      "loss": 7.5916,
      "mean_token_accuracy": 0.12931896969676018,
      "num_tokens": 1395228.0,
      "step": 550
    },
    {
      "entropy": 7.91993293762207,
      "epoch": 0.06405077899596076,
      "grad_norm": 1.265625,
      "learning_rate": 0.000277,
      "loss": 7.6362,
      "mean_token_accuracy": 0.12369817793369293,
      "num_tokens": 1408869.0,
      "step": 555
    },
    {
      "entropy": 7.821833944320678,
      "epoch": 0.06462781304096941,
      "grad_norm": 1.0,
      "learning_rate": 0.0002795,
      "loss": 7.5481,
      "mean_token_accuracy": 0.13478931561112403,
      "num_tokens": 1420926.0,
      "step": 560
    },
    {
      "entropy": 7.885378456115722,
      "epoch": 0.06520484708597807,
      "grad_norm": 1.296875,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.5576,
      "mean_token_accuracy": 0.13291405290365219,
      "num_tokens": 1432435.0,
      "step": 565
    },
    {
      "entropy": 7.920798873901367,
      "epoch": 0.06578188113098672,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002845,
      "loss": 7.6686,
      "mean_token_accuracy": 0.12425242960453034,
      "num_tokens": 1444716.0,
      "step": 570
    },
    {
      "entropy": 7.834726428985595,
      "epoch": 0.06635891517599539,
      "grad_norm": 1.109375,
      "learning_rate": 0.000287,
      "loss": 7.5579,
      "mean_token_accuracy": 0.13059103563427926,
      "num_tokens": 1458129.0,
      "step": 575
    },
    {
      "entropy": 7.880359935760498,
      "epoch": 0.06693594922100404,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002895,
      "loss": 7.5193,
      "mean_token_accuracy": 0.13621216639876366,
      "num_tokens": 1472076.0,
      "step": 580
    },
    {
      "entropy": 7.737836599349976,
      "epoch": 0.0675129832660127,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000292,
      "loss": 7.5129,
      "mean_token_accuracy": 0.1413876920938492,
      "num_tokens": 1486026.0,
      "step": 585
    },
    {
      "entropy": 7.900694990158081,
      "epoch": 0.06809001731102135,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002945,
      "loss": 7.5203,
      "mean_token_accuracy": 0.14100683629512786,
      "num_tokens": 1498735.0,
      "step": 590
    },
    {
      "entropy": 7.771698427200318,
      "epoch": 0.06866705135603,
      "grad_norm": 1.09375,
      "learning_rate": 0.000297,
      "loss": 7.5156,
      "mean_token_accuracy": 0.13441314548254013,
      "num_tokens": 1510755.0,
      "step": 595
    },
    {
      "entropy": 7.878929662704468,
      "epoch": 0.06924408540103866,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002995,
      "loss": 7.5271,
      "mean_token_accuracy": 0.13422106131911277,
      "num_tokens": 1522386.0,
      "step": 600
    },
    {
      "entropy": 7.86671781539917,
      "epoch": 0.06982111944604731,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000302,
      "loss": 7.5695,
      "mean_token_accuracy": 0.13297740891575813,
      "num_tokens": 1534757.0,
      "step": 605
    },
    {
      "entropy": 7.748486709594727,
      "epoch": 0.07039815349105598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003045,
      "loss": 7.5497,
      "mean_token_accuracy": 0.1324236772954464,
      "num_tokens": 1547684.0,
      "step": 610
    },
    {
      "entropy": 7.883321237564087,
      "epoch": 0.07097518753606463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000307,
      "loss": 7.5574,
      "mean_token_accuracy": 0.12341172322630882,
      "num_tokens": 1560583.0,
      "step": 615
    },
    {
      "entropy": 7.867546463012696,
      "epoch": 0.07155222158107329,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003095,
      "loss": 7.5735,
      "mean_token_accuracy": 0.12901610583066941,
      "num_tokens": 1573174.0,
      "step": 620
    },
    {
      "entropy": 7.840350246429443,
      "epoch": 0.07212925562608194,
      "grad_norm": 1.078125,
      "learning_rate": 0.000312,
      "loss": 7.5389,
      "mean_token_accuracy": 0.1297079123556614,
      "num_tokens": 1585746.0,
      "step": 625
    },
    {
      "entropy": 7.879965114593506,
      "epoch": 0.0727062896710906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003145,
      "loss": 7.5187,
      "mean_token_accuracy": 0.12504662796854973,
      "num_tokens": 1598538.0,
      "step": 630
    },
    {
      "entropy": 7.854893636703491,
      "epoch": 0.07328332371609925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000317,
      "loss": 7.5054,
      "mean_token_accuracy": 0.13247747644782065,
      "num_tokens": 1612598.0,
      "step": 635
    },
    {
      "entropy": 7.714817237854004,
      "epoch": 0.0738603577611079,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003195,
      "loss": 7.4557,
      "mean_token_accuracy": 0.136217100918293,
      "num_tokens": 1624418.0,
      "step": 640
    },
    {
      "entropy": 7.805634260177612,
      "epoch": 0.07443739180611655,
      "grad_norm": 1.03125,
      "learning_rate": 0.000322,
      "loss": 7.5434,
      "mean_token_accuracy": 0.1354817859828472,
      "num_tokens": 1637512.0,
      "step": 645
    },
    {
      "entropy": 7.727578449249267,
      "epoch": 0.07501442585112522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.4692,
      "mean_token_accuracy": 0.13448835983872415,
      "num_tokens": 1650413.0,
      "step": 650
    },
    {
      "entropy": 7.775011730194092,
      "epoch": 0.07559145989613388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.4587,
      "mean_token_accuracy": 0.13572532832622528,
      "num_tokens": 1662225.0,
      "step": 655
    },
    {
      "entropy": 7.828440713882446,
      "epoch": 0.07616849394114253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.5138,
      "mean_token_accuracy": 0.13002411052584648,
      "num_tokens": 1674458.0,
      "step": 660
    },
    {
      "entropy": 7.777482557296753,
      "epoch": 0.07674552798615118,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.5272,
      "mean_token_accuracy": 0.12747668400406836,
      "num_tokens": 1686366.0,
      "step": 665
    },
    {
      "entropy": 7.848160314559936,
      "epoch": 0.07732256203115984,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.5094,
      "mean_token_accuracy": 0.13185340240597726,
      "num_tokens": 1699654.0,
      "step": 670
    },
    {
      "entropy": 7.789048290252685,
      "epoch": 0.07789959607616849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000337,
      "loss": 7.4846,
      "mean_token_accuracy": 0.13520109057426452,
      "num_tokens": 1711923.0,
      "step": 675
    },
    {
      "entropy": 7.776696729660034,
      "epoch": 0.07847663012117714,
      "grad_norm": 1.0,
      "learning_rate": 0.0003395,
      "loss": 7.4714,
      "mean_token_accuracy": 0.13042250275611877,
      "num_tokens": 1723413.0,
      "step": 680
    },
    {
      "entropy": 7.701392698287964,
      "epoch": 0.07905366416618581,
      "grad_norm": 1.03125,
      "learning_rate": 0.000342,
      "loss": 7.4525,
      "mean_token_accuracy": 0.13371195793151855,
      "num_tokens": 1735975.0,
      "step": 685
    },
    {
      "entropy": 7.6978672504425045,
      "epoch": 0.07963069821119446,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.4664,
      "mean_token_accuracy": 0.1311781272292137,
      "num_tokens": 1748563.0,
      "step": 690
    },
    {
      "entropy": 7.824566507339478,
      "epoch": 0.08020773225620312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000347,
      "loss": 7.4771,
      "mean_token_accuracy": 0.1280561789870262,
      "num_tokens": 1762914.0,
      "step": 695
    },
    {
      "entropy": 7.78171935081482,
      "epoch": 0.08078476630121177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003495,
      "loss": 7.4409,
      "mean_token_accuracy": 0.13339700549840927,
      "num_tokens": 1776084.0,
      "step": 700
    },
    {
      "entropy": 7.676154851913452,
      "epoch": 0.08136180034622043,
      "grad_norm": 1.1875,
      "learning_rate": 0.000352,
      "loss": 7.3918,
      "mean_token_accuracy": 0.1427437275648117,
      "num_tokens": 1789316.0,
      "step": 705
    },
    {
      "entropy": 7.7570672035217285,
      "epoch": 0.08193883439122908,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003545,
      "loss": 7.4787,
      "mean_token_accuracy": 0.13278622627258302,
      "num_tokens": 1801576.0,
      "step": 710
    },
    {
      "entropy": 7.686878871917725,
      "epoch": 0.08251586843623773,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000357,
      "loss": 7.4405,
      "mean_token_accuracy": 0.13206920996308327,
      "num_tokens": 1814039.0,
      "step": 715
    },
    {
      "entropy": 7.712433910369873,
      "epoch": 0.0830929024812464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003595,
      "loss": 7.3683,
      "mean_token_accuracy": 0.13764831200242042,
      "num_tokens": 1825193.0,
      "step": 720
    },
    {
      "entropy": 7.6329395294189455,
      "epoch": 0.08366993652625505,
      "grad_norm": 1.140625,
      "learning_rate": 0.000362,
      "loss": 7.4537,
      "mean_token_accuracy": 0.13079532310366632,
      "num_tokens": 1837983.0,
      "step": 725
    },
    {
      "entropy": 7.7968615055084225,
      "epoch": 0.08424697057126371,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003645,
      "loss": 7.3549,
      "mean_token_accuracy": 0.14052335917949677,
      "num_tokens": 1849830.0,
      "step": 730
    },
    {
      "entropy": 7.692524814605713,
      "epoch": 0.08482400461627236,
      "grad_norm": 1.65625,
      "learning_rate": 0.000367,
      "loss": 7.448,
      "mean_token_accuracy": 0.12845715060830115,
      "num_tokens": 1862577.0,
      "step": 735
    },
    {
      "entropy": 7.684691143035889,
      "epoch": 0.08540103866128101,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003695,
      "loss": 7.4389,
      "mean_token_accuracy": 0.1409561887383461,
      "num_tokens": 1874888.0,
      "step": 740
    },
    {
      "entropy": 7.630507850646973,
      "epoch": 0.08597807270628967,
      "grad_norm": 1.328125,
      "learning_rate": 0.000372,
      "loss": 7.38,
      "mean_token_accuracy": 0.13916135132312774,
      "num_tokens": 1888400.0,
      "step": 745
    },
    {
      "entropy": 7.7386709690094,
      "epoch": 0.08655510675129832,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003745,
      "loss": 7.4495,
      "mean_token_accuracy": 0.1248070776462555,
      "num_tokens": 1900961.0,
      "step": 750
    },
    {
      "entropy": 7.621120452880859,
      "epoch": 0.08713214079630698,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000377,
      "loss": 7.3786,
      "mean_token_accuracy": 0.1386454552412033,
      "num_tokens": 1913410.0,
      "step": 755
    },
    {
      "entropy": 7.710708427429199,
      "epoch": 0.08770917484131564,
      "grad_norm": 0.875,
      "learning_rate": 0.0003795,
      "loss": 7.4316,
      "mean_token_accuracy": 0.12876348271965982,
      "num_tokens": 1926602.0,
      "step": 760
    },
    {
      "entropy": 7.669585132598877,
      "epoch": 0.0882862088863243,
      "grad_norm": 1.078125,
      "learning_rate": 0.000382,
      "loss": 7.3984,
      "mean_token_accuracy": 0.12868618741631507,
      "num_tokens": 1938904.0,
      "step": 765
    },
    {
      "entropy": 7.6731616973876955,
      "epoch": 0.08886324293133295,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003845,
      "loss": 7.4165,
      "mean_token_accuracy": 0.12934457659721374,
      "num_tokens": 1952827.0,
      "step": 770
    },
    {
      "entropy": 7.672389221191406,
      "epoch": 0.0894402769763416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.4066,
      "mean_token_accuracy": 0.13497869223356246,
      "num_tokens": 1965182.0,
      "step": 775
    },
    {
      "entropy": 7.608702707290649,
      "epoch": 0.09001731102135026,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.3787,
      "mean_token_accuracy": 0.13133419081568717,
      "num_tokens": 1976953.0,
      "step": 780
    },
    {
      "entropy": 7.742263889312744,
      "epoch": 0.09059434506635891,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.4563,
      "mean_token_accuracy": 0.12778021022677422,
      "num_tokens": 1990019.0,
      "step": 785
    },
    {
      "entropy": 7.656824254989624,
      "epoch": 0.09117137911136756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.4006,
      "mean_token_accuracy": 0.12971119731664657,
      "num_tokens": 2002976.0,
      "step": 790
    },
    {
      "entropy": 7.721537303924561,
      "epoch": 0.09174841315637623,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.4165,
      "mean_token_accuracy": 0.13233844116330146,
      "num_tokens": 2017368.0,
      "step": 795
    },
    {
      "entropy": 7.608252668380738,
      "epoch": 0.09232544720138489,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003995,
      "loss": 7.2751,
      "mean_token_accuracy": 0.14428590834140778,
      "num_tokens": 2029838.0,
      "step": 800
    },
    {
      "entropy": 7.679257488250732,
      "epoch": 0.09290248124639354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000402,
      "loss": 7.3901,
      "mean_token_accuracy": 0.12942394018173217,
      "num_tokens": 2042085.0,
      "step": 805
    },
    {
      "entropy": 7.609614467620849,
      "epoch": 0.0934795152914022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004045,
      "loss": 7.429,
      "mean_token_accuracy": 0.12986384257674216,
      "num_tokens": 2053268.0,
      "step": 810
    },
    {
      "entropy": 7.730389881134033,
      "epoch": 0.09405654933641085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.3995,
      "mean_token_accuracy": 0.1275894299149513,
      "num_tokens": 2066621.0,
      "step": 815
    },
    {
      "entropy": 7.573925876617432,
      "epoch": 0.0946335833814195,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004095,
      "loss": 7.2761,
      "mean_token_accuracy": 0.13354206904768945,
      "num_tokens": 2079590.0,
      "step": 820
    },
    {
      "entropy": 7.547151565551758,
      "epoch": 0.09521061742642815,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000412,
      "loss": 7.3921,
      "mean_token_accuracy": 0.1272098556160927,
      "num_tokens": 2091729.0,
      "step": 825
    },
    {
      "entropy": 7.617713403701782,
      "epoch": 0.09578765147143682,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004145,
      "loss": 7.3473,
      "mean_token_accuracy": 0.14121743738651277,
      "num_tokens": 2104254.0,
      "step": 830
    },
    {
      "entropy": 7.655541276931762,
      "epoch": 0.09636468551644548,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000417,
      "loss": 7.4288,
      "mean_token_accuracy": 0.12773946151137353,
      "num_tokens": 2116327.0,
      "step": 835
    },
    {
      "entropy": 7.666590738296509,
      "epoch": 0.09694171956145413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004195,
      "loss": 7.4027,
      "mean_token_accuracy": 0.1262945294380188,
      "num_tokens": 2129153.0,
      "step": 840
    },
    {
      "entropy": 7.557644271850586,
      "epoch": 0.09751875360646278,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000422,
      "loss": 7.3806,
      "mean_token_accuracy": 0.13962640687823297,
      "num_tokens": 2141705.0,
      "step": 845
    },
    {
      "entropy": 7.566726779937744,
      "epoch": 0.09809578765147144,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004245,
      "loss": 7.2944,
      "mean_token_accuracy": 0.14180327355861663,
      "num_tokens": 2153760.0,
      "step": 850
    },
    {
      "entropy": 7.562725973129273,
      "epoch": 0.09867282169648009,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000427,
      "loss": 7.3744,
      "mean_token_accuracy": 0.1379995971918106,
      "num_tokens": 2168206.0,
      "step": 855
    },
    {
      "entropy": 7.643193006515503,
      "epoch": 0.09924985574148874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004295,
      "loss": 7.3256,
      "mean_token_accuracy": 0.13964372426271437,
      "num_tokens": 2180779.0,
      "step": 860
    },
    {
      "entropy": 7.448547601699829,
      "epoch": 0.0998268897864974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000432,
      "loss": 7.2866,
      "mean_token_accuracy": 0.13754021003842354,
      "num_tokens": 2194075.0,
      "step": 865
    },
    {
      "entropy": 7.622214460372925,
      "epoch": 0.10040392383150606,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004345,
      "loss": 7.2761,
      "mean_token_accuracy": 0.13813476786017417,
      "num_tokens": 2206424.0,
      "step": 870
    },
    {
      "entropy": 7.570018768310547,
      "epoch": 0.10098095787651472,
      "grad_norm": 1.125,
      "learning_rate": 0.000437,
      "loss": 7.3298,
      "mean_token_accuracy": 0.13034077361226082,
      "num_tokens": 2218550.0,
      "step": 875
    },
    {
      "entropy": 7.488341760635376,
      "epoch": 0.10155799192152337,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004395,
      "loss": 7.3548,
      "mean_token_accuracy": 0.13045953586697578,
      "num_tokens": 2231857.0,
      "step": 880
    },
    {
      "entropy": 7.663240146636963,
      "epoch": 0.10213502596653203,
      "grad_norm": 0.984375,
      "learning_rate": 0.000442,
      "loss": 7.3455,
      "mean_token_accuracy": 0.1357438363134861,
      "num_tokens": 2244057.0,
      "step": 885
    },
    {
      "entropy": 7.607003974914551,
      "epoch": 0.10271206001154068,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004445,
      "loss": 7.3748,
      "mean_token_accuracy": 0.12887232899665832,
      "num_tokens": 2257361.0,
      "step": 890
    },
    {
      "entropy": 7.620429849624633,
      "epoch": 0.10328909405654933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000447,
      "loss": 7.3642,
      "mean_token_accuracy": 0.12847339510917663,
      "num_tokens": 2269599.0,
      "step": 895
    },
    {
      "entropy": 7.570056581497193,
      "epoch": 0.10386612810155799,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.3901,
      "mean_token_accuracy": 0.1326812021434307,
      "num_tokens": 2282791.0,
      "step": 900
    },
    {
      "entropy": 7.571472215652466,
      "epoch": 0.10444316214656665,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.3424,
      "mean_token_accuracy": 0.13455241098999976,
      "num_tokens": 2295336.0,
      "step": 905
    },
    {
      "entropy": 7.547875213623047,
      "epoch": 0.10502019619157531,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.3645,
      "mean_token_accuracy": 0.12806227579712867,
      "num_tokens": 2308813.0,
      "step": 910
    },
    {
      "entropy": 7.587298440933227,
      "epoch": 0.10559723023658396,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.3429,
      "mean_token_accuracy": 0.1311824142932892,
      "num_tokens": 2321374.0,
      "step": 915
    },
    {
      "entropy": 7.545827627182007,
      "epoch": 0.10617426428159261,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.338,
      "mean_token_accuracy": 0.1307897984981537,
      "num_tokens": 2333698.0,
      "step": 920
    },
    {
      "entropy": 7.575364637374878,
      "epoch": 0.10675129832660127,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462,
      "loss": 7.3168,
      "mean_token_accuracy": 0.13864337727427484,
      "num_tokens": 2346557.0,
      "step": 925
    },
    {
      "entropy": 7.4012375831604,
      "epoch": 0.10732833237160992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004645,
      "loss": 7.1836,
      "mean_token_accuracy": 0.14260239005088807,
      "num_tokens": 2360003.0,
      "step": 930
    },
    {
      "entropy": 7.531010293960572,
      "epoch": 0.10790536641661858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000467,
      "loss": 7.2266,
      "mean_token_accuracy": 0.13916042149066926,
      "num_tokens": 2372765.0,
      "step": 935
    },
    {
      "entropy": 7.4169941425323485,
      "epoch": 0.10848240046162723,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004695,
      "loss": 7.2525,
      "mean_token_accuracy": 0.14179157763719558,
      "num_tokens": 2385154.0,
      "step": 940
    },
    {
      "entropy": 7.489320898056031,
      "epoch": 0.1090594345066359,
      "grad_norm": 1.015625,
      "learning_rate": 0.000472,
      "loss": 7.2705,
      "mean_token_accuracy": 0.13921171203255653,
      "num_tokens": 2397264.0,
      "step": 945
    },
    {
      "entropy": 7.529940891265869,
      "epoch": 0.10963646855164455,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004745,
      "loss": 7.3586,
      "mean_token_accuracy": 0.1348928801715374,
      "num_tokens": 2408652.0,
      "step": 950
    },
    {
      "entropy": 7.531420660018921,
      "epoch": 0.1102135025966532,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000477,
      "loss": 7.331,
      "mean_token_accuracy": 0.1281139448285103,
      "num_tokens": 2421683.0,
      "step": 955
    },
    {
      "entropy": 7.433173418045044,
      "epoch": 0.11079053664166186,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004795,
      "loss": 7.2361,
      "mean_token_accuracy": 0.13509340435266495,
      "num_tokens": 2434622.0,
      "step": 960
    },
    {
      "entropy": 7.488950967788696,
      "epoch": 0.11136757068667051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000482,
      "loss": 7.2495,
      "mean_token_accuracy": 0.14383002668619155,
      "num_tokens": 2448157.0,
      "step": 965
    },
    {
      "entropy": 7.640981388092041,
      "epoch": 0.11194460473167916,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845,
      "loss": 7.2964,
      "mean_token_accuracy": 0.13189825639128686,
      "num_tokens": 2460284.0,
      "step": 970
    },
    {
      "entropy": 7.387933540344238,
      "epoch": 0.11252163877668782,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000487,
      "loss": 7.309,
      "mean_token_accuracy": 0.13392101153731345,
      "num_tokens": 2473031.0,
      "step": 975
    },
    {
      "entropy": 7.540426445007324,
      "epoch": 0.11309867282169649,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004895,
      "loss": 7.3065,
      "mean_token_accuracy": 0.13692763224244117,
      "num_tokens": 2486870.0,
      "step": 980
    },
    {
      "entropy": 7.408073234558105,
      "epoch": 0.11367570686670514,
      "grad_norm": 0.96875,
      "learning_rate": 0.000492,
      "loss": 7.2302,
      "mean_token_accuracy": 0.1348964937031269,
      "num_tokens": 2498921.0,
      "step": 985
    },
    {
      "entropy": 7.482239198684693,
      "epoch": 0.1142527409117138,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945,
      "loss": 7.2894,
      "mean_token_accuracy": 0.13312276005744933,
      "num_tokens": 2511122.0,
      "step": 990
    },
    {
      "entropy": 7.542117166519165,
      "epoch": 0.11482977495672245,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497,
      "loss": 7.3043,
      "mean_token_accuracy": 0.1307055488228798,
      "num_tokens": 2523726.0,
      "step": 995
    },
    {
      "entropy": 7.41837511062622,
      "epoch": 0.1154068090017311,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995,
      "loss": 7.1923,
      "mean_token_accuracy": 0.1403934270143509,
      "num_tokens": 2536192.0,
      "step": 1000
    },
    {
      "entropy": 7.390990495681763,
      "epoch": 0.11598384304673975,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999999975783157,
      "loss": 7.2816,
      "mean_token_accuracy": 0.1320057801902294,
      "num_tokens": 2550200.0,
      "step": 1005
    },
    {
      "entropy": 7.49788727760315,
      "epoch": 0.11656087709174841,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999877402236,
      "loss": 7.2259,
      "mean_token_accuracy": 0.13413601964712143,
      "num_tokens": 2564045.0,
      "step": 1010
    },
    {
      "entropy": 7.4572518825531,
      "epoch": 0.11713791113675708,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999999703343686,
      "loss": 7.238,
      "mean_token_accuracy": 0.13197707533836364,
      "num_tokens": 2576210.0,
      "step": 1015
    },
    {
      "entropy": 7.300177717208863,
      "epoch": 0.11771494518176573,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999999453607515,
      "loss": 7.2207,
      "mean_token_accuracy": 0.13198934867978096,
      "num_tokens": 2589325.0,
      "step": 1020
    },
    {
      "entropy": 7.501029205322266,
      "epoch": 0.11829197922677438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999128193729,
      "loss": 7.3114,
      "mean_token_accuracy": 0.13109516799449922,
      "num_tokens": 2601522.0,
      "step": 1025
    },
    {
      "entropy": 7.46219277381897,
      "epoch": 0.11886901327178304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999872710234,
      "loss": 7.2959,
      "mean_token_accuracy": 0.12850206419825555,
      "num_tokens": 2615364.0,
      "step": 1030
    },
    {
      "entropy": 7.4426497459411625,
      "epoch": 0.11944604731679169,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999998250333361,
      "loss": 7.2097,
      "mean_token_accuracy": 0.13902488797903062,
      "num_tokens": 2628528.0,
      "step": 1035
    },
    {
      "entropy": 7.394994068145752,
      "epoch": 0.12002308136180034,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499999769788681,
      "loss": 7.2178,
      "mean_token_accuracy": 0.13481198623776436,
      "num_tokens": 2641877.0,
      "step": 1040
    },
    {
      "entropy": 7.42519211769104,
      "epoch": 0.120600115406809,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999997069762703,
      "loss": 7.2311,
      "mean_token_accuracy": 0.13621581196784974,
      "num_tokens": 2655419.0,
      "step": 1045
    },
    {
      "entropy": 7.355701065063476,
      "epoch": 0.12117714945181765,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999996365961062,
      "loss": 7.1697,
      "mean_token_accuracy": 0.13674544021487237,
      "num_tokens": 2667551.0,
      "step": 1050
    },
    {
      "entropy": 7.418414783477783,
      "epoch": 0.12175418349682632,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999995586481912,
      "loss": 7.1974,
      "mean_token_accuracy": 0.14013378769159318,
      "num_tokens": 2679948.0,
      "step": 1055
    },
    {
      "entropy": 7.3376726627349855,
      "epoch": 0.12233121754183497,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999994731325276,
      "loss": 7.2412,
      "mean_token_accuracy": 0.13359286561608313,
      "num_tokens": 2691501.0,
      "step": 1060
    },
    {
      "entropy": 7.368611145019531,
      "epoch": 0.12290825158684363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999993800491187,
      "loss": 7.176,
      "mean_token_accuracy": 0.14094864577054977,
      "num_tokens": 2702689.0,
      "step": 1065
    },
    {
      "entropy": 7.382964181900024,
      "epoch": 0.12348528563185228,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999992793979672,
      "loss": 7.1841,
      "mean_token_accuracy": 0.13821902945637704,
      "num_tokens": 2715023.0,
      "step": 1070
    },
    {
      "entropy": 7.3155985355377195,
      "epoch": 0.12406231967686093,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999991711790769,
      "loss": 7.1123,
      "mean_token_accuracy": 0.13841997236013412,
      "num_tokens": 2729517.0,
      "step": 1075
    },
    {
      "entropy": 7.430340957641602,
      "epoch": 0.12463935372186959,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999990553924511,
      "loss": 7.2008,
      "mean_token_accuracy": 0.13296948969364167,
      "num_tokens": 2740984.0,
      "step": 1080
    },
    {
      "entropy": 7.35272068977356,
      "epoch": 0.12521638776687824,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999989320380939,
      "loss": 7.1488,
      "mean_token_accuracy": 0.1401528961956501,
      "num_tokens": 2752943.0,
      "step": 1085
    },
    {
      "entropy": 7.400033378601075,
      "epoch": 0.1257934218118869,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999988011160094,
      "loss": 7.2192,
      "mean_token_accuracy": 0.1346215933561325,
      "num_tokens": 2765613.0,
      "step": 1090
    },
    {
      "entropy": 7.316109132766724,
      "epoch": 0.12637045585689555,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499998662626202,
      "loss": 7.1648,
      "mean_token_accuracy": 0.1355959139764309,
      "num_tokens": 2776820.0,
      "step": 1095
    },
    {
      "entropy": 7.295491933822632,
      "epoch": 0.12694748990190421,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985165686764,
      "loss": 7.2006,
      "mean_token_accuracy": 0.13195707648992538,
      "num_tokens": 2788981.0,
      "step": 1100
    },
    {
      "entropy": 7.3825420379638675,
      "epoch": 0.12752452394691285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999983629434373,
      "loss": 7.0555,
      "mean_token_accuracy": 0.14337626621127128,
      "num_tokens": 2801975.0,
      "step": 1105
    },
    {
      "entropy": 7.324372148513794,
      "epoch": 0.12810155799192152,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999982017504901,
      "loss": 7.1557,
      "mean_token_accuracy": 0.13406250923871993,
      "num_tokens": 2814708.0,
      "step": 1110
    },
    {
      "entropy": 7.2684407234191895,
      "epoch": 0.1286785920369302,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999980329898401,
      "loss": 7.174,
      "mean_token_accuracy": 0.13960450887680054,
      "num_tokens": 2827903.0,
      "step": 1115
    },
    {
      "entropy": 7.3982825756073,
      "epoch": 0.12925562608193883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499997856661493,
      "loss": 7.1933,
      "mean_token_accuracy": 0.13698331117630005,
      "num_tokens": 2841077.0,
      "step": 1120
    },
    {
      "entropy": 7.370109128952026,
      "epoch": 0.1298326601269475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499997672765455,
      "loss": 7.1852,
      "mean_token_accuracy": 0.13384110406041144,
      "num_tokens": 2853235.0,
      "step": 1125
    },
    {
      "entropy": 7.2596477508544925,
      "epoch": 0.13040969417195614,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999974813017318,
      "loss": 7.1604,
      "mean_token_accuracy": 0.13414481207728385,
      "num_tokens": 2864937.0,
      "step": 1130
    },
    {
      "entropy": 7.35138201713562,
      "epoch": 0.1309867282169648,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999972822703301,
      "loss": 7.1043,
      "mean_token_accuracy": 0.14946642518043518,
      "num_tokens": 2877791.0,
      "step": 1135
    },
    {
      "entropy": 7.182547760009766,
      "epoch": 0.13156376226197344,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999970756712567,
      "loss": 7.1067,
      "mean_token_accuracy": 0.13508472070097924,
      "num_tokens": 2889897.0,
      "step": 1140
    },
    {
      "entropy": 7.388927268981933,
      "epoch": 0.1321407963069821,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999968615045183,
      "loss": 7.1212,
      "mean_token_accuracy": 0.13854536414146423,
      "num_tokens": 2902806.0,
      "step": 1145
    },
    {
      "entropy": 7.248739910125733,
      "epoch": 0.13271783035199078,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999966397701222,
      "loss": 7.0086,
      "mean_token_accuracy": 0.14631882458925247,
      "num_tokens": 2915703.0,
      "step": 1150
    },
    {
      "entropy": 7.229858493804931,
      "epoch": 0.13329486439699942,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999964104680759,
      "loss": 7.2181,
      "mean_token_accuracy": 0.12988507375121117,
      "num_tokens": 2928171.0,
      "step": 1155
    },
    {
      "entropy": 7.360022401809692,
      "epoch": 0.13387189844200809,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999961735983871,
      "loss": 7.0487,
      "mean_token_accuracy": 0.1509324714541435,
      "num_tokens": 2939909.0,
      "step": 1160
    },
    {
      "entropy": 7.3091637134552006,
      "epoch": 0.13444893248701673,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999959291610639,
      "loss": 7.1827,
      "mean_token_accuracy": 0.13319118991494178,
      "num_tokens": 2951510.0,
      "step": 1165
    },
    {
      "entropy": 7.270810317993164,
      "epoch": 0.1350259665320254,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999956771561143,
      "loss": 7.1046,
      "mean_token_accuracy": 0.1315624199807644,
      "num_tokens": 2963326.0,
      "step": 1170
    },
    {
      "entropy": 7.282600927352905,
      "epoch": 0.13560300057703403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999954175835469,
      "loss": 7.1236,
      "mean_token_accuracy": 0.14552151560783386,
      "num_tokens": 2976226.0,
      "step": 1175
    },
    {
      "entropy": 7.304236793518067,
      "epoch": 0.1361800346220427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999951504433703,
      "loss": 7.1306,
      "mean_token_accuracy": 0.1356511816382408,
      "num_tokens": 2988180.0,
      "step": 1180
    },
    {
      "entropy": 7.281073331832886,
      "epoch": 0.13675706866705137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999948757355935,
      "loss": 7.1185,
      "mean_token_accuracy": 0.13962753042578696,
      "num_tokens": 3000215.0,
      "step": 1185
    },
    {
      "entropy": 7.306526613235474,
      "epoch": 0.13733410271206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499994593460226,
      "loss": 7.1111,
      "mean_token_accuracy": 0.13763693794608117,
      "num_tokens": 3012516.0,
      "step": 1190
    },
    {
      "entropy": 7.259176111221313,
      "epoch": 0.13791113675706868,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999943036172771,
      "loss": 7.1082,
      "mean_token_accuracy": 0.13961452320218087,
      "num_tokens": 3024786.0,
      "step": 1195
    },
    {
      "entropy": 7.322293615341186,
      "epoch": 0.13848817080207732,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999940062067565,
      "loss": 7.04,
      "mean_token_accuracy": 0.1440444752573967,
      "num_tokens": 3037179.0,
      "step": 1200
    },
    {
      "entropy": 7.134593200683594,
      "epoch": 0.13906520484708598,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999937012286742,
      "loss": 7.1333,
      "mean_token_accuracy": 0.14015561193227768,
      "num_tokens": 3049443.0,
      "step": 1205
    },
    {
      "entropy": 7.305765962600708,
      "epoch": 0.13964223889209462,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999933886830405,
      "loss": 7.0521,
      "mean_token_accuracy": 0.133638371527195,
      "num_tokens": 3061857.0,
      "step": 1210
    },
    {
      "entropy": 7.215990972518921,
      "epoch": 0.1402192729371033,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499993068569866,
      "loss": 7.0849,
      "mean_token_accuracy": 0.1435427337884903,
      "num_tokens": 3075185.0,
      "step": 1215
    },
    {
      "entropy": 7.301103687286377,
      "epoch": 0.14079630698211196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999927408891614,
      "loss": 7.0919,
      "mean_token_accuracy": 0.14063736498355867,
      "num_tokens": 3087645.0,
      "step": 1220
    },
    {
      "entropy": 7.285601663589477,
      "epoch": 0.1413733410271206,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999924056409378,
      "loss": 7.0908,
      "mean_token_accuracy": 0.14075700491666793,
      "num_tokens": 3101185.0,
      "step": 1225
    },
    {
      "entropy": 7.270492601394653,
      "epoch": 0.14195037507212926,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999920628252063,
      "loss": 7.0766,
      "mean_token_accuracy": 0.13953982219099997,
      "num_tokens": 3113694.0,
      "step": 1230
    },
    {
      "entropy": 7.237199211120606,
      "epoch": 0.1425274091171379,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999917124419785,
      "loss": 7.1188,
      "mean_token_accuracy": 0.13823226392269133,
      "num_tokens": 3128225.0,
      "step": 1235
    },
    {
      "entropy": 7.233181381225586,
      "epoch": 0.14310444316214657,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999913544912664,
      "loss": 7.0533,
      "mean_token_accuracy": 0.1442795380949974,
      "num_tokens": 3140118.0,
      "step": 1240
    },
    {
      "entropy": 7.241433238983154,
      "epoch": 0.1436814772071552,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999909889730817,
      "loss": 7.1085,
      "mean_token_accuracy": 0.13870764151215553,
      "num_tokens": 3153479.0,
      "step": 1245
    },
    {
      "entropy": 7.260871505737304,
      "epoch": 0.14425851125216388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999906158874368,
      "loss": 7.1168,
      "mean_token_accuracy": 0.14061163738369942,
      "num_tokens": 3167062.0,
      "step": 1250
    },
    {
      "entropy": 7.2724611282348635,
      "epoch": 0.14483554529717255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999902352343444,
      "loss": 7.0729,
      "mean_token_accuracy": 0.1381714127957821,
      "num_tokens": 3179061.0,
      "step": 1255
    },
    {
      "entropy": 7.174509811401367,
      "epoch": 0.1454125793421812,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999898470138171,
      "loss": 7.0213,
      "mean_token_accuracy": 0.13980793803930283,
      "num_tokens": 3191333.0,
      "step": 1260
    },
    {
      "entropy": 7.224514389038086,
      "epoch": 0.14598961338718985,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999894512258681,
      "loss": 7.1164,
      "mean_token_accuracy": 0.13482679277658463,
      "num_tokens": 3204487.0,
      "step": 1265
    },
    {
      "entropy": 7.220690965652466,
      "epoch": 0.1465666474321985,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999890478705107,
      "loss": 7.0728,
      "mean_token_accuracy": 0.1409827023744583,
      "num_tokens": 3216733.0,
      "step": 1270
    },
    {
      "entropy": 7.218664693832397,
      "epoch": 0.14714368147720716,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999886369477585,
      "loss": 7.0233,
      "mean_token_accuracy": 0.14548835158348083,
      "num_tokens": 3229605.0,
      "step": 1275
    },
    {
      "entropy": 7.249156379699707,
      "epoch": 0.1477207155222158,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999882184576251,
      "loss": 7.0641,
      "mean_token_accuracy": 0.14037289768457412,
      "num_tokens": 3241023.0,
      "step": 1280
    },
    {
      "entropy": 7.183506107330322,
      "epoch": 0.14829774956722447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999877924001248,
      "loss": 7.0454,
      "mean_token_accuracy": 0.1420667700469494,
      "num_tokens": 3252435.0,
      "step": 1285
    },
    {
      "entropy": 7.202338171005249,
      "epoch": 0.1488747836122331,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999873587752718,
      "loss": 7.0315,
      "mean_token_accuracy": 0.14606617093086244,
      "num_tokens": 3265663.0,
      "step": 1290
    },
    {
      "entropy": 7.2627809047698975,
      "epoch": 0.14945181765724178,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999869175830808,
      "loss": 7.0998,
      "mean_token_accuracy": 0.13558180630207062,
      "num_tokens": 3278309.0,
      "step": 1295
    },
    {
      "entropy": 7.196434354782104,
      "epoch": 0.15002885170225044,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999864688235666,
      "loss": 7.0881,
      "mean_token_accuracy": 0.13663394004106522,
      "num_tokens": 3290899.0,
      "step": 1300
    },
    {
      "entropy": 7.256631898880005,
      "epoch": 0.15060588574725908,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999860124967442,
      "loss": 7.0591,
      "mean_token_accuracy": 0.1400838539004326,
      "num_tokens": 3304383.0,
      "step": 1305
    },
    {
      "entropy": 7.142842245101929,
      "epoch": 0.15118291979226775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999855486026291,
      "loss": 7.0381,
      "mean_token_accuracy": 0.14005750343203544,
      "num_tokens": 3316310.0,
      "step": 1310
    },
    {
      "entropy": 7.253703260421753,
      "epoch": 0.1517599538372764,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999850771412369,
      "loss": 7.1357,
      "mean_token_accuracy": 0.13773472383618354,
      "num_tokens": 3330018.0,
      "step": 1315
    },
    {
      "entropy": 7.299247360229492,
      "epoch": 0.15233698788228506,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999845981125832,
      "loss": 7.1805,
      "mean_token_accuracy": 0.13796070963144302,
      "num_tokens": 3343134.0,
      "step": 1320
    },
    {
      "entropy": 7.159939193725586,
      "epoch": 0.1529140219272937,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999841115166844,
      "loss": 6.9761,
      "mean_token_accuracy": 0.14545321986079215,
      "num_tokens": 3356262.0,
      "step": 1325
    },
    {
      "entropy": 7.279706287384033,
      "epoch": 0.15349105597230236,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999836173535568,
      "loss": 7.1249,
      "mean_token_accuracy": 0.13199433758854867,
      "num_tokens": 3369672.0,
      "step": 1330
    },
    {
      "entropy": 7.183068513870239,
      "epoch": 0.15406809001731103,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999831156232169,
      "loss": 7.0391,
      "mean_token_accuracy": 0.14253250360488892,
      "num_tokens": 3381677.0,
      "step": 1335
    },
    {
      "entropy": 7.151807451248169,
      "epoch": 0.15464512406231967,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999826063256818,
      "loss": 6.9944,
      "mean_token_accuracy": 0.14133523479104043,
      "num_tokens": 3394662.0,
      "step": 1340
    },
    {
      "entropy": 7.135809326171875,
      "epoch": 0.15522215810732834,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999820894609683,
      "loss": 7.0019,
      "mean_token_accuracy": 0.14013058245182036,
      "num_tokens": 3407286.0,
      "step": 1345
    },
    {
      "entropy": 7.337306499481201,
      "epoch": 0.15579919215233698,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999815650290941,
      "loss": 7.0935,
      "mean_token_accuracy": 0.13155431523919106,
      "num_tokens": 3419478.0,
      "step": 1350
    },
    {
      "entropy": 7.090397119522095,
      "epoch": 0.15637622619734565,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999810330300766,
      "loss": 7.0581,
      "mean_token_accuracy": 0.1393677696585655,
      "num_tokens": 3433188.0,
      "step": 1355
    },
    {
      "entropy": 7.227596616744995,
      "epoch": 0.1569532602423543,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499980493463934,
      "loss": 6.9917,
      "mean_token_accuracy": 0.13574201092123986,
      "num_tokens": 3445806.0,
      "step": 1360
    },
    {
      "entropy": 7.163070344924927,
      "epoch": 0.15753029428736295,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999799463306841,
      "loss": 7.0483,
      "mean_token_accuracy": 0.1425718478858471,
      "num_tokens": 3458931.0,
      "step": 1365
    },
    {
      "entropy": 7.1892472267150875,
      "epoch": 0.15810732833237162,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999793916303455,
      "loss": 7.0889,
      "mean_token_accuracy": 0.1362600401043892,
      "num_tokens": 3471338.0,
      "step": 1370
    },
    {
      "entropy": 7.238631296157837,
      "epoch": 0.15868436237738026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999788293629368,
      "loss": 7.1303,
      "mean_token_accuracy": 0.1389173984527588,
      "num_tokens": 3484865.0,
      "step": 1375
    },
    {
      "entropy": 7.231470727920533,
      "epoch": 0.15926139642238893,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499978259528477,
      "loss": 6.9649,
      "mean_token_accuracy": 0.14690266847610473,
      "num_tokens": 3496676.0,
      "step": 1380
    },
    {
      "entropy": 7.058669900894165,
      "epoch": 0.15983843046739757,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999776821269852,
      "loss": 6.9509,
      "mean_token_accuracy": 0.14693509563803672,
      "num_tokens": 3508748.0,
      "step": 1385
    },
    {
      "entropy": 7.182062768936158,
      "epoch": 0.16041546451240624,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999770971584807,
      "loss": 7.0749,
      "mean_token_accuracy": 0.1334662154316902,
      "num_tokens": 3521490.0,
      "step": 1390
    },
    {
      "entropy": 7.199066686630249,
      "epoch": 0.16099249855741488,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999765046229834,
      "loss": 7.0484,
      "mean_token_accuracy": 0.1401499927043915,
      "num_tokens": 3534092.0,
      "step": 1395
    },
    {
      "entropy": 7.191224384307861,
      "epoch": 0.16156953260242354,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499975904520513,
      "loss": 7.0213,
      "mean_token_accuracy": 0.1379491239786148,
      "num_tokens": 3546583.0,
      "step": 1400
    },
    {
      "entropy": 7.084802722930908,
      "epoch": 0.1621465666474322,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00049997529685109,
      "loss": 7.002,
      "mean_token_accuracy": 0.13953247293829918,
      "num_tokens": 3559409.0,
      "step": 1405
    },
    {
      "entropy": 7.161634111404419,
      "epoch": 0.16272360069244085,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999746816147344,
      "loss": 7.0043,
      "mean_token_accuracy": 0.14320720732212067,
      "num_tokens": 3572362.0,
      "step": 1410
    },
    {
      "entropy": 7.124744558334351,
      "epoch": 0.16330063473744952,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999740588114673,
      "loss": 6.9779,
      "mean_token_accuracy": 0.14698471277952194,
      "num_tokens": 3585505.0,
      "step": 1415
    },
    {
      "entropy": 7.198022842407227,
      "epoch": 0.16387766878245816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999734284413095,
      "loss": 7.0287,
      "mean_token_accuracy": 0.14301127195358276,
      "num_tokens": 3597322.0,
      "step": 1420
    },
    {
      "entropy": 7.142577838897705,
      "epoch": 0.16445470282746683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499972790504282,
      "loss": 7.0051,
      "mean_token_accuracy": 0.1411785215139389,
      "num_tokens": 3609941.0,
      "step": 1425
    },
    {
      "entropy": 7.169891500473023,
      "epoch": 0.16503173687247547,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999721450004067,
      "loss": 7.0594,
      "mean_token_accuracy": 0.1376806043088436,
      "num_tokens": 3621751.0,
      "step": 1430
    },
    {
      "entropy": 7.1877641677856445,
      "epoch": 0.16560877091748413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999714919297049,
      "loss": 6.9994,
      "mean_token_accuracy": 0.1362007051706314,
      "num_tokens": 3633070.0,
      "step": 1435
    },
    {
      "entropy": 7.156223154067993,
      "epoch": 0.1661858049624928,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999708312921987,
      "loss": 6.991,
      "mean_token_accuracy": 0.14234942495822905,
      "num_tokens": 3646730.0,
      "step": 1440
    },
    {
      "entropy": 7.137038898468018,
      "epoch": 0.16676283900750144,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999701630879103,
      "loss": 7.024,
      "mean_token_accuracy": 0.13512019515037538,
      "num_tokens": 3659976.0,
      "step": 1445
    },
    {
      "entropy": 7.2560724258422855,
      "epoch": 0.1673398730525101,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999694873168623,
      "loss": 6.9583,
      "mean_token_accuracy": 0.13890018165111542,
      "num_tokens": 3672607.0,
      "step": 1450
    },
    {
      "entropy": 6.892187547683716,
      "epoch": 0.16791690709751875,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999688039790773,
      "loss": 6.9249,
      "mean_token_accuracy": 0.14720971807837485,
      "num_tokens": 3686338.0,
      "step": 1455
    },
    {
      "entropy": 7.185595369338989,
      "epoch": 0.16849394114252741,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999681130745784,
      "loss": 6.919,
      "mean_token_accuracy": 0.14689514487981797,
      "num_tokens": 3698986.0,
      "step": 1460
    },
    {
      "entropy": 7.067193078994751,
      "epoch": 0.16907097518753605,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999674146033888,
      "loss": 6.9691,
      "mean_token_accuracy": 0.1407572127878666,
      "num_tokens": 3712453.0,
      "step": 1465
    },
    {
      "entropy": 7.098576354980469,
      "epoch": 0.16964800923254472,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999667085655318,
      "loss": 6.9661,
      "mean_token_accuracy": 0.14339267313480378,
      "num_tokens": 3724769.0,
      "step": 1470
    },
    {
      "entropy": 7.177932548522949,
      "epoch": 0.1702250432775534,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999659949610313,
      "loss": 7.0199,
      "mean_token_accuracy": 0.13096101582050323,
      "num_tokens": 3738052.0,
      "step": 1475
    },
    {
      "entropy": 7.143703556060791,
      "epoch": 0.17080207732256203,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999652737899114,
      "loss": 6.9883,
      "mean_token_accuracy": 0.1414633348584175,
      "num_tokens": 3751279.0,
      "step": 1480
    },
    {
      "entropy": 7.204730272293091,
      "epoch": 0.1713791113675707,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999645450521963,
      "loss": 7.0358,
      "mean_token_accuracy": 0.13535647615790367,
      "num_tokens": 3764688.0,
      "step": 1485
    },
    {
      "entropy": 7.067404890060425,
      "epoch": 0.17195614541257934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999638087479104,
      "loss": 7.0258,
      "mean_token_accuracy": 0.13848631605505943,
      "num_tokens": 3776892.0,
      "step": 1490
    },
    {
      "entropy": 7.166511583328247,
      "epoch": 0.172533179457588,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999630648770786,
      "loss": 6.9324,
      "mean_token_accuracy": 0.14192868992686272,
      "num_tokens": 3789287.0,
      "step": 1495
    },
    {
      "entropy": 7.094117498397827,
      "epoch": 0.17311021350259664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999623134397257,
      "loss": 6.9241,
      "mean_token_accuracy": 0.13871007338166236,
      "num_tokens": 3801467.0,
      "step": 1500
    },
    {
      "entropy": 7.110012578964233,
      "epoch": 0.1736872475476053,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999615544358774,
      "loss": 6.9988,
      "mean_token_accuracy": 0.14102682918310167,
      "num_tokens": 3814641.0,
      "step": 1505
    },
    {
      "entropy": 7.171355581283569,
      "epoch": 0.17426428159261395,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999607878655587,
      "loss": 7.058,
      "mean_token_accuracy": 0.1313117817044258,
      "num_tokens": 3827187.0,
      "step": 1510
    },
    {
      "entropy": 7.093271684646607,
      "epoch": 0.17484131563762262,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999600137287958,
      "loss": 6.8747,
      "mean_token_accuracy": 0.15039864480495452,
      "num_tokens": 3839832.0,
      "step": 1515
    },
    {
      "entropy": 7.007730913162232,
      "epoch": 0.17541834968263129,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999592320256146,
      "loss": 6.882,
      "mean_token_accuracy": 0.14504249542951583,
      "num_tokens": 3852026.0,
      "step": 1520
    },
    {
      "entropy": 7.090141820907593,
      "epoch": 0.17599538372763993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999584427560412,
      "loss": 6.9337,
      "mean_token_accuracy": 0.14259505569934844,
      "num_tokens": 3864481.0,
      "step": 1525
    },
    {
      "entropy": 7.11194920539856,
      "epoch": 0.1765724177726486,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999576459201024,
      "loss": 6.926,
      "mean_token_accuracy": 0.14007529392838478,
      "num_tokens": 3876896.0,
      "step": 1530
    },
    {
      "entropy": 7.059630441665649,
      "epoch": 0.17714945181765723,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499956841517825,
      "loss": 6.8702,
      "mean_token_accuracy": 0.14155263453722,
      "num_tokens": 3890230.0,
      "step": 1535
    },
    {
      "entropy": 7.151209115982056,
      "epoch": 0.1777264858626659,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499956029549236,
      "loss": 6.9884,
      "mean_token_accuracy": 0.1379861555993557,
      "num_tokens": 3903138.0,
      "step": 1540
    },
    {
      "entropy": 7.033334732055664,
      "epoch": 0.17830351990767454,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999552100143625,
      "loss": 6.9137,
      "mean_token_accuracy": 0.14293416365981101,
      "num_tokens": 3915709.0,
      "step": 1545
    },
    {
      "entropy": 7.103172302246094,
      "epoch": 0.1788805539526832,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999543829132324,
      "loss": 6.9886,
      "mean_token_accuracy": 0.14267231971025468,
      "num_tokens": 3929265.0,
      "step": 1550
    },
    {
      "entropy": 7.093789052963257,
      "epoch": 0.17945758799769188,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999535482458734,
      "loss": 6.9683,
      "mean_token_accuracy": 0.14312490001320838,
      "num_tokens": 3940416.0,
      "step": 1555
    },
    {
      "entropy": 7.07169542312622,
      "epoch": 0.18003462204270052,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999527060123135,
      "loss": 6.9805,
      "mean_token_accuracy": 0.1391856700181961,
      "num_tokens": 3952999.0,
      "step": 1560
    },
    {
      "entropy": 7.027859497070312,
      "epoch": 0.18061165608770918,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999518562125811,
      "loss": 7.0033,
      "mean_token_accuracy": 0.14257199838757514,
      "num_tokens": 3966852.0,
      "step": 1565
    },
    {
      "entropy": 7.128302907943725,
      "epoch": 0.18118869013271782,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999509988467047,
      "loss": 6.8997,
      "mean_token_accuracy": 0.14498577862977982,
      "num_tokens": 3978242.0,
      "step": 1570
    },
    {
      "entropy": 7.117135620117187,
      "epoch": 0.1817657241777265,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999501339147132,
      "loss": 6.9929,
      "mean_token_accuracy": 0.13448395878076552,
      "num_tokens": 3993122.0,
      "step": 1575
    },
    {
      "entropy": 7.120546340942383,
      "epoch": 0.18234275822273513,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999492614166357,
      "loss": 6.9519,
      "mean_token_accuracy": 0.13845855444669725,
      "num_tokens": 4005230.0,
      "step": 1580
    },
    {
      "entropy": 7.002993774414063,
      "epoch": 0.1829197922677438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999483813525014,
      "loss": 6.8912,
      "mean_token_accuracy": 0.1465035580098629,
      "num_tokens": 4017413.0,
      "step": 1585
    },
    {
      "entropy": 7.117990589141845,
      "epoch": 0.18349682631275246,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999474937223403,
      "loss": 6.913,
      "mean_token_accuracy": 0.13722215369343757,
      "num_tokens": 4029679.0,
      "step": 1590
    },
    {
      "entropy": 7.0321924686431885,
      "epoch": 0.1840738603577611,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999465985261818,
      "loss": 6.8641,
      "mean_token_accuracy": 0.14145687371492385,
      "num_tokens": 4041857.0,
      "step": 1595
    },
    {
      "entropy": 7.059554100036621,
      "epoch": 0.18465089440276977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999456957640562,
      "loss": 6.9627,
      "mean_token_accuracy": 0.1417013004422188,
      "num_tokens": 4054743.0,
      "step": 1600
    },
    {
      "entropy": 6.978120374679565,
      "epoch": 0.1852279284477784,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999447854359939,
      "loss": 6.8829,
      "mean_token_accuracy": 0.1442536674439907,
      "num_tokens": 4067554.0,
      "step": 1605
    },
    {
      "entropy": 7.109914064407349,
      "epoch": 0.18580496249278708,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999438675420255,
      "loss": 6.8864,
      "mean_token_accuracy": 0.14278573244810105,
      "num_tokens": 4079912.0,
      "step": 1610
    },
    {
      "entropy": 7.044018888473511,
      "epoch": 0.18638199653779572,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999429420821818,
      "loss": 6.9875,
      "mean_token_accuracy": 0.13912659734487534,
      "num_tokens": 4093217.0,
      "step": 1615
    },
    {
      "entropy": 7.054386758804322,
      "epoch": 0.1869590305828044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499942009056494,
      "loss": 6.9365,
      "mean_token_accuracy": 0.13861448243260382,
      "num_tokens": 4104328.0,
      "step": 1620
    },
    {
      "entropy": 7.088769435882568,
      "epoch": 0.18753606462781305,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999410684649935,
      "loss": 6.903,
      "mean_token_accuracy": 0.13839107751846313,
      "num_tokens": 4116711.0,
      "step": 1625
    },
    {
      "entropy": 7.108532571792603,
      "epoch": 0.1881130986728217,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999401203077119,
      "loss": 6.916,
      "mean_token_accuracy": 0.14068023785948752,
      "num_tokens": 4129066.0,
      "step": 1630
    },
    {
      "entropy": 6.976222705841065,
      "epoch": 0.18869013271783036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499939164584681,
      "loss": 6.8407,
      "mean_token_accuracy": 0.14638919979333878,
      "num_tokens": 4141312.0,
      "step": 1635
    },
    {
      "entropy": 6.951454305648804,
      "epoch": 0.189267166762839,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999382012959331,
      "loss": 6.891,
      "mean_token_accuracy": 0.14239912778139113,
      "num_tokens": 4154687.0,
      "step": 1640
    },
    {
      "entropy": 7.107205200195312,
      "epoch": 0.18984420080784767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999372304415005,
      "loss": 6.9926,
      "mean_token_accuracy": 0.1338139146566391,
      "num_tokens": 4168401.0,
      "step": 1645
    },
    {
      "entropy": 7.093311500549317,
      "epoch": 0.1904212348528563,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999362520214159,
      "loss": 6.9056,
      "mean_token_accuracy": 0.1434105969965458,
      "num_tokens": 4180707.0,
      "step": 1650
    },
    {
      "entropy": 7.006139135360717,
      "epoch": 0.19099826889786498,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999352660357122,
      "loss": 6.8472,
      "mean_token_accuracy": 0.14567850530147552,
      "num_tokens": 4192939.0,
      "step": 1655
    },
    {
      "entropy": 7.006642055511475,
      "epoch": 0.19157530294287364,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999342724844226,
      "loss": 6.8907,
      "mean_token_accuracy": 0.1400732360780239,
      "num_tokens": 4206277.0,
      "step": 1660
    },
    {
      "entropy": 7.036981534957886,
      "epoch": 0.19215233698788228,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999332713675804,
      "loss": 6.9061,
      "mean_token_accuracy": 0.14375606924295425,
      "num_tokens": 4218672.0,
      "step": 1665
    },
    {
      "entropy": 7.058644390106201,
      "epoch": 0.19272937103289095,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999322626852193,
      "loss": 6.8487,
      "mean_token_accuracy": 0.15019180476665497,
      "num_tokens": 4229002.0,
      "step": 1670
    },
    {
      "entropy": 6.931442832946777,
      "epoch": 0.1933064050778996,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999312464373734,
      "loss": 6.7568,
      "mean_token_accuracy": 0.15428706631064415,
      "num_tokens": 4241986.0,
      "step": 1675
    },
    {
      "entropy": 6.937554502487183,
      "epoch": 0.19388343912290826,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999302226240767,
      "loss": 6.8767,
      "mean_token_accuracy": 0.14542328268289567,
      "num_tokens": 4254670.0,
      "step": 1680
    },
    {
      "entropy": 7.057171106338501,
      "epoch": 0.1944604731679169,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999291912453637,
      "loss": 6.9243,
      "mean_token_accuracy": 0.14324737936258317,
      "num_tokens": 4267561.0,
      "step": 1685
    },
    {
      "entropy": 7.091326284408569,
      "epoch": 0.19503750721292556,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999281523012691,
      "loss": 6.8723,
      "mean_token_accuracy": 0.14641524106264114,
      "num_tokens": 4280145.0,
      "step": 1690
    },
    {
      "entropy": 6.949259901046753,
      "epoch": 0.1956145412579342,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999271057918278,
      "loss": 6.8951,
      "mean_token_accuracy": 0.1458272233605385,
      "num_tokens": 4294089.0,
      "step": 1695
    },
    {
      "entropy": 6.969711494445801,
      "epoch": 0.19619157530294287,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999260517170751,
      "loss": 6.8523,
      "mean_token_accuracy": 0.14484949558973312,
      "num_tokens": 4305926.0,
      "step": 1700
    },
    {
      "entropy": 7.015812873840332,
      "epoch": 0.19676860934795154,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999249900770463,
      "loss": 6.8777,
      "mean_token_accuracy": 0.14453437477350234,
      "num_tokens": 4319319.0,
      "step": 1705
    },
    {
      "entropy": 7.054222202301025,
      "epoch": 0.19734564339296018,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999239208717772,
      "loss": 6.88,
      "mean_token_accuracy": 0.1493258535861969,
      "num_tokens": 4331650.0,
      "step": 1710
    },
    {
      "entropy": 6.912113094329834,
      "epoch": 0.19792267743796885,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999228441013037,
      "loss": 6.8579,
      "mean_token_accuracy": 0.14917318373918534,
      "num_tokens": 4344289.0,
      "step": 1715
    },
    {
      "entropy": 7.053227233886719,
      "epoch": 0.1984997114829775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499921759765662,
      "loss": 6.8428,
      "mean_token_accuracy": 0.1453030899167061,
      "num_tokens": 4357316.0,
      "step": 1720
    },
    {
      "entropy": 6.902173471450806,
      "epoch": 0.19907674552798615,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999206678648888,
      "loss": 6.8593,
      "mean_token_accuracy": 0.14832349121570587,
      "num_tokens": 4369383.0,
      "step": 1725
    },
    {
      "entropy": 6.987811851501465,
      "epoch": 0.1996537795729948,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999195683990206,
      "loss": 6.8523,
      "mean_token_accuracy": 0.14121579453349115,
      "num_tokens": 4383509.0,
      "step": 1730
    },
    {
      "entropy": 7.036195564270019,
      "epoch": 0.20023081361800346,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999184613680945,
      "loss": 6.8422,
      "mean_token_accuracy": 0.14607394635677337,
      "num_tokens": 4395105.0,
      "step": 1735
    },
    {
      "entropy": 6.9437949657440186,
      "epoch": 0.20080784766301213,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999173467721476,
      "loss": 6.8231,
      "mean_token_accuracy": 0.1548911601305008,
      "num_tokens": 4407473.0,
      "step": 1740
    },
    {
      "entropy": 7.070507526397705,
      "epoch": 0.20138488170802077,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999162246112175,
      "loss": 6.9122,
      "mean_token_accuracy": 0.13833602294325828,
      "num_tokens": 4420186.0,
      "step": 1745
    },
    {
      "entropy": 6.968269634246826,
      "epoch": 0.20196191575302944,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999150948853419,
      "loss": 6.821,
      "mean_token_accuracy": 0.1459064930677414,
      "num_tokens": 4432944.0,
      "step": 1750
    },
    {
      "entropy": 7.0235504627227785,
      "epoch": 0.20253894979803808,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999139575945587,
      "loss": 6.8803,
      "mean_token_accuracy": 0.14378070682287217,
      "num_tokens": 4445784.0,
      "step": 1755
    },
    {
      "entropy": 6.927468872070312,
      "epoch": 0.20311598384304674,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999128127389065,
      "loss": 6.7843,
      "mean_token_accuracy": 0.15202904790639876,
      "num_tokens": 4458338.0,
      "step": 1760
    },
    {
      "entropy": 6.9563414573669435,
      "epoch": 0.20369301788805538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999116603184233,
      "loss": 6.7983,
      "mean_token_accuracy": 0.13958832919597625,
      "num_tokens": 4471058.0,
      "step": 1765
    },
    {
      "entropy": 6.899150657653808,
      "epoch": 0.20427005193306405,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999105003331482,
      "loss": 6.8266,
      "mean_token_accuracy": 0.14727174118161201,
      "num_tokens": 4485561.0,
      "step": 1770
    },
    {
      "entropy": 6.9349593162536625,
      "epoch": 0.20484708597807272,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999093327831202,
      "loss": 6.8535,
      "mean_token_accuracy": 0.14275807067751883,
      "num_tokens": 4497865.0,
      "step": 1775
    },
    {
      "entropy": 6.864494657516479,
      "epoch": 0.20542412002308136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999081576683784,
      "loss": 6.7372,
      "mean_token_accuracy": 0.15288822427392007,
      "num_tokens": 4510291.0,
      "step": 1780
    },
    {
      "entropy": 6.981197881698608,
      "epoch": 0.20600115406809003,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999069749889626,
      "loss": 6.8206,
      "mean_token_accuracy": 0.14839925467967988,
      "num_tokens": 4521855.0,
      "step": 1785
    },
    {
      "entropy": 6.962658452987671,
      "epoch": 0.20657818811309867,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999057847449123,
      "loss": 6.9264,
      "mean_token_accuracy": 0.1439384013414383,
      "num_tokens": 4534123.0,
      "step": 1790
    },
    {
      "entropy": 7.070970106124878,
      "epoch": 0.20715522215810733,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999045869362678,
      "loss": 6.8721,
      "mean_token_accuracy": 0.1396043874323368,
      "num_tokens": 4547298.0,
      "step": 1795
    },
    {
      "entropy": 6.949685668945312,
      "epoch": 0.20773225620311597,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499903381563069,
      "loss": 6.8077,
      "mean_token_accuracy": 0.14740960970520972,
      "num_tokens": 4560945.0,
      "step": 1800
    },
    {
      "entropy": 6.966697263717651,
      "epoch": 0.20830929024812464,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999021686253568,
      "loss": 6.8687,
      "mean_token_accuracy": 0.1465895041823387,
      "num_tokens": 4572696.0,
      "step": 1805
    },
    {
      "entropy": 6.888779878616333,
      "epoch": 0.2088863242931333,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999009481231719,
      "loss": 6.8272,
      "mean_token_accuracy": 0.14835924357175828,
      "num_tokens": 4585670.0,
      "step": 1810
    },
    {
      "entropy": 7.021191167831421,
      "epoch": 0.20946335833814195,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998997200565553,
      "loss": 6.8068,
      "mean_token_accuracy": 0.1528303876519203,
      "num_tokens": 4597694.0,
      "step": 1815
    },
    {
      "entropy": 6.919396781921387,
      "epoch": 0.21004039238315061,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998984844255483,
      "loss": 6.81,
      "mean_token_accuracy": 0.15265992432832717,
      "num_tokens": 4609924.0,
      "step": 1820
    },
    {
      "entropy": 6.99350414276123,
      "epoch": 0.21061742642815925,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998972412301925,
      "loss": 6.8249,
      "mean_token_accuracy": 0.1444974884390831,
      "num_tokens": 4621668.0,
      "step": 1825
    },
    {
      "entropy": 6.94732346534729,
      "epoch": 0.21119446047316792,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004998959904705297,
      "loss": 6.881,
      "mean_token_accuracy": 0.14270134419202804,
      "num_tokens": 4635970.0,
      "step": 1830
    },
    {
      "entropy": 7.050157785415649,
      "epoch": 0.21177149451817656,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998947321466021,
      "loss": 6.8651,
      "mean_token_accuracy": 0.14370742663741112,
      "num_tokens": 4649557.0,
      "step": 1835
    },
    {
      "entropy": 6.908232021331787,
      "epoch": 0.21234852856318523,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998934662584518,
      "loss": 6.8489,
      "mean_token_accuracy": 0.1472188040614128,
      "num_tokens": 4662572.0,
      "step": 1840
    },
    {
      "entropy": 7.015526866912841,
      "epoch": 0.2129255626081939,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998921928061214,
      "loss": 6.787,
      "mean_token_accuracy": 0.14296051636338233,
      "num_tokens": 4674783.0,
      "step": 1845
    },
    {
      "entropy": 6.846453762054443,
      "epoch": 0.21350259665320254,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998909117896539,
      "loss": 6.8143,
      "mean_token_accuracy": 0.15064377710223198,
      "num_tokens": 4687543.0,
      "step": 1850
    },
    {
      "entropy": 7.0557780265808105,
      "epoch": 0.2140796306982112,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998896232090922,
      "loss": 6.7774,
      "mean_token_accuracy": 0.14568573087453843,
      "num_tokens": 4700225.0,
      "step": 1855
    },
    {
      "entropy": 6.856030988693237,
      "epoch": 0.21465666474321984,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998883270644798,
      "loss": 6.8354,
      "mean_token_accuracy": 0.14540046080946922,
      "num_tokens": 4712950.0,
      "step": 1860
    },
    {
      "entropy": 7.031614017486572,
      "epoch": 0.2152336987882285,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998870233558602,
      "loss": 6.915,
      "mean_token_accuracy": 0.1439936079084873,
      "num_tokens": 4725563.0,
      "step": 1865
    },
    {
      "entropy": 6.960827159881592,
      "epoch": 0.21581073283323715,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998857120832774,
      "loss": 6.8384,
      "mean_token_accuracy": 0.14699823558330535,
      "num_tokens": 4739057.0,
      "step": 1870
    },
    {
      "entropy": 6.924222373962403,
      "epoch": 0.21638776687824582,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998843932467751,
      "loss": 6.8103,
      "mean_token_accuracy": 0.15074544101953508,
      "num_tokens": 4752782.0,
      "step": 1875
    },
    {
      "entropy": 6.948456811904907,
      "epoch": 0.21696480092325446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998830668463982,
      "loss": 6.7441,
      "mean_token_accuracy": 0.1475722536444664,
      "num_tokens": 4765740.0,
      "step": 1880
    },
    {
      "entropy": 6.868886423110962,
      "epoch": 0.21754183496826313,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998817328821909,
      "loss": 6.8159,
      "mean_token_accuracy": 0.1473565399646759,
      "num_tokens": 4778711.0,
      "step": 1885
    },
    {
      "entropy": 7.014859580993653,
      "epoch": 0.2181188690132718,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998803913541983,
      "loss": 6.8055,
      "mean_token_accuracy": 0.14383689314126968,
      "num_tokens": 4791402.0,
      "step": 1890
    },
    {
      "entropy": 6.904099798202514,
      "epoch": 0.21869590305828043,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998790422624654,
      "loss": 6.7913,
      "mean_token_accuracy": 0.14016161784529685,
      "num_tokens": 4804256.0,
      "step": 1895
    },
    {
      "entropy": 6.875801038742066,
      "epoch": 0.2192729371032891,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998776856070376,
      "loss": 6.738,
      "mean_token_accuracy": 0.14255951046943666,
      "num_tokens": 4817424.0,
      "step": 1900
    },
    {
      "entropy": 6.946020650863647,
      "epoch": 0.21984997114829774,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004998763213879606,
      "loss": 6.7723,
      "mean_token_accuracy": 0.14450910985469817,
      "num_tokens": 4829222.0,
      "step": 1905
    },
    {
      "entropy": 6.881495380401612,
      "epoch": 0.2204270051933064,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998749496052803,
      "loss": 6.7586,
      "mean_token_accuracy": 0.15232504308223724,
      "num_tokens": 4844011.0,
      "step": 1910
    },
    {
      "entropy": 6.943768501281738,
      "epoch": 0.22100403923831505,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998735702590426,
      "loss": 6.8173,
      "mean_token_accuracy": 0.14637627452611923,
      "num_tokens": 4857116.0,
      "step": 1915
    },
    {
      "entropy": 6.903482675552368,
      "epoch": 0.22158107328332372,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998721833492942,
      "loss": 6.7527,
      "mean_token_accuracy": 0.15328776836395264,
      "num_tokens": 4867680.0,
      "step": 1920
    },
    {
      "entropy": 6.915145015716552,
      "epoch": 0.22215810732833238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998707888760816,
      "loss": 6.7966,
      "mean_token_accuracy": 0.1467222422361374,
      "num_tokens": 4880553.0,
      "step": 1925
    },
    {
      "entropy": 6.887819623947143,
      "epoch": 0.22273514137334102,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998693868394516,
      "loss": 6.7723,
      "mean_token_accuracy": 0.14810495600104331,
      "num_tokens": 4893706.0,
      "step": 1930
    },
    {
      "entropy": 6.863858842849732,
      "epoch": 0.2233121754183497,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998679772394516,
      "loss": 6.8486,
      "mean_token_accuracy": 0.14729523956775664,
      "num_tokens": 4906016.0,
      "step": 1935
    },
    {
      "entropy": 6.896185255050659,
      "epoch": 0.22388920946335833,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998665600761289,
      "loss": 6.7278,
      "mean_token_accuracy": 0.15100182890892028,
      "num_tokens": 4918628.0,
      "step": 1940
    },
    {
      "entropy": 7.010975885391235,
      "epoch": 0.224466243508367,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499865135349531,
      "loss": 6.8144,
      "mean_token_accuracy": 0.14225947707891465,
      "num_tokens": 4930525.0,
      "step": 1945
    },
    {
      "entropy": 6.8868975162506105,
      "epoch": 0.22504327755337564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998637030597061,
      "loss": 6.8197,
      "mean_token_accuracy": 0.1457245357334614,
      "num_tokens": 4942510.0,
      "step": 1950
    },
    {
      "entropy": 6.970962142944336,
      "epoch": 0.2256203115983843,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998622632067022,
      "loss": 6.7952,
      "mean_token_accuracy": 0.14592548012733458,
      "num_tokens": 4954865.0,
      "step": 1955
    },
    {
      "entropy": 6.893493223190307,
      "epoch": 0.22619734564339297,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998608157905678,
      "loss": 6.8813,
      "mean_token_accuracy": 0.1398551307618618,
      "num_tokens": 4968078.0,
      "step": 1960
    },
    {
      "entropy": 6.975720071792603,
      "epoch": 0.2267743796884016,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998593608113515,
      "loss": 6.7472,
      "mean_token_accuracy": 0.14719463735818863,
      "num_tokens": 4980273.0,
      "step": 1965
    },
    {
      "entropy": 6.914252281188965,
      "epoch": 0.22735141373341028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998578982691024,
      "loss": 6.7927,
      "mean_token_accuracy": 0.1422052152454853,
      "num_tokens": 4992372.0,
      "step": 1970
    },
    {
      "entropy": 6.883531761169434,
      "epoch": 0.22792844777841892,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998564281638694,
      "loss": 6.7757,
      "mean_token_accuracy": 0.1407983958721161,
      "num_tokens": 5004766.0,
      "step": 1975
    },
    {
      "entropy": 6.94020791053772,
      "epoch": 0.2285054818234276,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998549504957023,
      "loss": 6.7176,
      "mean_token_accuracy": 0.14406620860099792,
      "num_tokens": 5017007.0,
      "step": 1980
    },
    {
      "entropy": 6.817730140686035,
      "epoch": 0.22908251586843623,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998534652646506,
      "loss": 6.7859,
      "mean_token_accuracy": 0.14582390040159227,
      "num_tokens": 5029506.0,
      "step": 1985
    },
    {
      "entropy": 6.877771329879761,
      "epoch": 0.2296595499134449,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998519724707642,
      "loss": 6.7609,
      "mean_token_accuracy": 0.14823370724916457,
      "num_tokens": 5042209.0,
      "step": 1990
    },
    {
      "entropy": 6.908406686782837,
      "epoch": 0.23023658395845356,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998504721140934,
      "loss": 6.8161,
      "mean_token_accuracy": 0.13886914625763894,
      "num_tokens": 5056026.0,
      "step": 1995
    },
    {
      "entropy": 6.977062129974366,
      "epoch": 0.2308136180034622,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998489641946887,
      "loss": 6.7594,
      "mean_token_accuracy": 0.14582734405994416,
      "num_tokens": 5068194.0,
      "step": 2000
    },
    {
      "entropy": 6.795575380325317,
      "epoch": 0.23139065204847087,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998474487126009,
      "loss": 6.6985,
      "mean_token_accuracy": 0.14483340829610825,
      "num_tokens": 5080468.0,
      "step": 2005
    },
    {
      "entropy": 6.980053567886353,
      "epoch": 0.2319676860934795,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998459256678806,
      "loss": 6.8987,
      "mean_token_accuracy": 0.13540890365839003,
      "num_tokens": 5093104.0,
      "step": 2010
    },
    {
      "entropy": 6.952063846588135,
      "epoch": 0.23254472013848818,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998443950605796,
      "loss": 6.7604,
      "mean_token_accuracy": 0.14487868547439575,
      "num_tokens": 5105710.0,
      "step": 2015
    },
    {
      "entropy": 6.916586208343506,
      "epoch": 0.23312175418349682,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998428568907488,
      "loss": 6.8412,
      "mean_token_accuracy": 0.141026022285223,
      "num_tokens": 5118946.0,
      "step": 2020
    },
    {
      "entropy": 6.842178583145142,
      "epoch": 0.23369878822850548,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998413111584403,
      "loss": 6.6824,
      "mean_token_accuracy": 0.15238308906555176,
      "num_tokens": 5132567.0,
      "step": 2025
    },
    {
      "entropy": 6.798287200927734,
      "epoch": 0.23427582227351415,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998397578637059,
      "loss": 6.7384,
      "mean_token_accuracy": 0.15025758892297744,
      "num_tokens": 5145669.0,
      "step": 2030
    },
    {
      "entropy": 6.840568113327026,
      "epoch": 0.2348528563185228,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499838197006598,
      "loss": 6.6806,
      "mean_token_accuracy": 0.15366968512535095,
      "num_tokens": 5157753.0,
      "step": 2035
    },
    {
      "entropy": 6.853817701339722,
      "epoch": 0.23542989036353146,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499836628587169,
      "loss": 6.7113,
      "mean_token_accuracy": 0.14877334460616112,
      "num_tokens": 5170167.0,
      "step": 2040
    },
    {
      "entropy": 6.85221381187439,
      "epoch": 0.2360069244085401,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998350526054716,
      "loss": 6.6912,
      "mean_token_accuracy": 0.14868487864732743,
      "num_tokens": 5182143.0,
      "step": 2045
    },
    {
      "entropy": 6.8682215213775635,
      "epoch": 0.23658395845354876,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998334690615591,
      "loss": 6.7324,
      "mean_token_accuracy": 0.1463753066956997,
      "num_tokens": 5195586.0,
      "step": 2050
    },
    {
      "entropy": 6.841140508651733,
      "epoch": 0.2371609924985574,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998318779554844,
      "loss": 6.7221,
      "mean_token_accuracy": 0.15004333406686782,
      "num_tokens": 5207774.0,
      "step": 2055
    },
    {
      "entropy": 6.816382312774659,
      "epoch": 0.23773802654356607,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998302792873012,
      "loss": 6.7348,
      "mean_token_accuracy": 0.145672906935215,
      "num_tokens": 5219330.0,
      "step": 2060
    },
    {
      "entropy": 6.898803949356079,
      "epoch": 0.23831506058857474,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998286730570631,
      "loss": 6.8011,
      "mean_token_accuracy": 0.1392673708498478,
      "num_tokens": 5233817.0,
      "step": 2065
    },
    {
      "entropy": 6.921881198883057,
      "epoch": 0.23889209463358338,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998270592648245,
      "loss": 6.8029,
      "mean_token_accuracy": 0.1426341101527214,
      "num_tokens": 5246483.0,
      "step": 2070
    },
    {
      "entropy": 6.877314949035645,
      "epoch": 0.23946912867859205,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998254379106394,
      "loss": 6.7872,
      "mean_token_accuracy": 0.14690051525831221,
      "num_tokens": 5258695.0,
      "step": 2075
    },
    {
      "entropy": 6.834933519363403,
      "epoch": 0.2400461627236007,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998238089945622,
      "loss": 6.7364,
      "mean_token_accuracy": 0.14769575297832488,
      "num_tokens": 5271355.0,
      "step": 2080
    },
    {
      "entropy": 6.983479166030884,
      "epoch": 0.24062319676860935,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998221725166479,
      "loss": 6.8541,
      "mean_token_accuracy": 0.1414130374789238,
      "num_tokens": 5284905.0,
      "step": 2085
    },
    {
      "entropy": 6.848435688018799,
      "epoch": 0.241200230813618,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998205284769516,
      "loss": 6.7898,
      "mean_token_accuracy": 0.14240173250436783,
      "num_tokens": 5298266.0,
      "step": 2090
    },
    {
      "entropy": 6.868869304656982,
      "epoch": 0.24177726485862666,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998188768755285,
      "loss": 6.8241,
      "mean_token_accuracy": 0.13813576549291612,
      "num_tokens": 5311148.0,
      "step": 2095
    },
    {
      "entropy": 6.894551563262939,
      "epoch": 0.2423542989036353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998172177124341,
      "loss": 6.7433,
      "mean_token_accuracy": 0.14466421604156493,
      "num_tokens": 5322638.0,
      "step": 2100
    },
    {
      "entropy": 6.798158168792725,
      "epoch": 0.24293133294864397,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998155509877242,
      "loss": 6.6199,
      "mean_token_accuracy": 0.15917377769947053,
      "num_tokens": 5336116.0,
      "step": 2105
    },
    {
      "entropy": 6.763039493560791,
      "epoch": 0.24350836699365264,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499813876701455,
      "loss": 6.7412,
      "mean_token_accuracy": 0.1486453168094158,
      "num_tokens": 5347561.0,
      "step": 2110
    },
    {
      "entropy": 6.927878141403198,
      "epoch": 0.24408540103866128,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998121948536828,
      "loss": 6.7747,
      "mean_token_accuracy": 0.14461961835622789,
      "num_tokens": 5360055.0,
      "step": 2115
    },
    {
      "entropy": 6.847961521148681,
      "epoch": 0.24466243508366994,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998105054444639,
      "loss": 6.7334,
      "mean_token_accuracy": 0.1498301789164543,
      "num_tokens": 5372722.0,
      "step": 2120
    },
    {
      "entropy": 6.897721290588379,
      "epoch": 0.24523946912867858,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998088084738555,
      "loss": 6.6907,
      "mean_token_accuracy": 0.1467824786901474,
      "num_tokens": 5385586.0,
      "step": 2125
    },
    {
      "entropy": 6.765969133377075,
      "epoch": 0.24581650317368725,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998071039419144,
      "loss": 6.7337,
      "mean_token_accuracy": 0.14697647988796234,
      "num_tokens": 5399266.0,
      "step": 2130
    },
    {
      "entropy": 6.8921960353851315,
      "epoch": 0.2463935372186959,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499805391848698,
      "loss": 6.7285,
      "mean_token_accuracy": 0.14361063987016678,
      "num_tokens": 5412093.0,
      "step": 2135
    },
    {
      "entropy": 6.846024703979492,
      "epoch": 0.24697057126370456,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499803672194264,
      "loss": 6.7348,
      "mean_token_accuracy": 0.14491766393184663,
      "num_tokens": 5425181.0,
      "step": 2140
    },
    {
      "entropy": 6.797625732421875,
      "epoch": 0.24754760530871323,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998019449786701,
      "loss": 6.6422,
      "mean_token_accuracy": 0.15672230273485183,
      "num_tokens": 5437392.0,
      "step": 2145
    },
    {
      "entropy": 6.836762523651123,
      "epoch": 0.24812463935372187,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998002102019744,
      "loss": 6.716,
      "mean_token_accuracy": 0.15314831882715224,
      "num_tokens": 5449455.0,
      "step": 2150
    },
    {
      "entropy": 6.800679159164429,
      "epoch": 0.24870167339873053,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997984678642354,
      "loss": 6.7343,
      "mean_token_accuracy": 0.14900478795170785,
      "num_tokens": 5463095.0,
      "step": 2155
    },
    {
      "entropy": 6.874364376068115,
      "epoch": 0.24927870744373917,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997967179655115,
      "loss": 6.687,
      "mean_token_accuracy": 0.1538333460688591,
      "num_tokens": 5474440.0,
      "step": 2160
    },
    {
      "entropy": 6.821845054626465,
      "epoch": 0.24985574148874784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997949605058617,
      "loss": 6.688,
      "mean_token_accuracy": 0.15788259357213974,
      "num_tokens": 5487365.0,
      "step": 2165
    },
    {
      "entropy": 6.809021139144898,
      "epoch": 0.2504327755337565,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004997931954853451,
      "loss": 6.7504,
      "mean_token_accuracy": 0.14503707811236383,
      "num_tokens": 5502356.0,
      "step": 2170
    },
    {
      "entropy": 6.827416038513183,
      "epoch": 0.2510098095787651,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499791422904021,
      "loss": 6.7697,
      "mean_token_accuracy": 0.14718132317066193,
      "num_tokens": 5514504.0,
      "step": 2175
    },
    {
      "entropy": 6.879432010650635,
      "epoch": 0.2515868436237738,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997896427619491,
      "loss": 6.787,
      "mean_token_accuracy": 0.14120125472545625,
      "num_tokens": 5526415.0,
      "step": 2180
    },
    {
      "entropy": 6.782674407958984,
      "epoch": 0.25216387766878245,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997878550591892,
      "loss": 6.6442,
      "mean_token_accuracy": 0.1523044466972351,
      "num_tokens": 5538484.0,
      "step": 2185
    },
    {
      "entropy": 6.856536865234375,
      "epoch": 0.2527409117137911,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997860597958013,
      "loss": 6.7129,
      "mean_token_accuracy": 0.15341712385416031,
      "num_tokens": 5551315.0,
      "step": 2190
    },
    {
      "entropy": 6.747131633758545,
      "epoch": 0.2533179457587998,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997842569718461,
      "loss": 6.6025,
      "mean_token_accuracy": 0.1544630780816078,
      "num_tokens": 5564179.0,
      "step": 2195
    },
    {
      "entropy": 6.77038607597351,
      "epoch": 0.25389497980380843,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499782446587384,
      "loss": 6.735,
      "mean_token_accuracy": 0.14310188591480255,
      "num_tokens": 5577060.0,
      "step": 2200
    },
    {
      "entropy": 6.7083740234375,
      "epoch": 0.25447201384881707,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499780628642476,
      "loss": 6.548,
      "mean_token_accuracy": 0.16547798812389375,
      "num_tokens": 5589170.0,
      "step": 2205
    },
    {
      "entropy": 6.812865543365478,
      "epoch": 0.2550490478938257,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997788031371834,
      "loss": 6.7259,
      "mean_token_accuracy": 0.14614113718271254,
      "num_tokens": 5601747.0,
      "step": 2210
    },
    {
      "entropy": 6.895756387710572,
      "epoch": 0.2556260819388344,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997769700715672,
      "loss": 6.7757,
      "mean_token_accuracy": 0.1394893616437912,
      "num_tokens": 5614308.0,
      "step": 2215
    },
    {
      "entropy": 6.80255651473999,
      "epoch": 0.25620311598384304,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997751294456892,
      "loss": 6.7322,
      "mean_token_accuracy": 0.13944590389728545,
      "num_tokens": 5626351.0,
      "step": 2220
    },
    {
      "entropy": 6.84980354309082,
      "epoch": 0.2567801500288517,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997732812596114,
      "loss": 6.7099,
      "mean_token_accuracy": 0.15208475142717362,
      "num_tokens": 5638339.0,
      "step": 2225
    },
    {
      "entropy": 6.850376224517822,
      "epoch": 0.2573571840738604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997714255133961,
      "loss": 6.7641,
      "mean_token_accuracy": 0.14269857704639435,
      "num_tokens": 5651210.0,
      "step": 2230
    },
    {
      "entropy": 6.788392782211304,
      "epoch": 0.257934218118869,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997695622071054,
      "loss": 6.7186,
      "mean_token_accuracy": 0.14826106950640677,
      "num_tokens": 5664150.0,
      "step": 2235
    },
    {
      "entropy": 6.822963619232178,
      "epoch": 0.25851125216387766,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997676913408021,
      "loss": 6.6615,
      "mean_token_accuracy": 0.15033092349767685,
      "num_tokens": 5676009.0,
      "step": 2240
    },
    {
      "entropy": 6.927838134765625,
      "epoch": 0.2590882862088863,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997658129145493,
      "loss": 6.8377,
      "mean_token_accuracy": 0.1423884242773056,
      "num_tokens": 5688557.0,
      "step": 2245
    },
    {
      "entropy": 6.772985315322876,
      "epoch": 0.259665320253895,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997639269284099,
      "loss": 6.7958,
      "mean_token_accuracy": 0.14698884561657904,
      "num_tokens": 5702032.0,
      "step": 2250
    },
    {
      "entropy": 6.828845500946045,
      "epoch": 0.26024235429890363,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997620333824476,
      "loss": 6.6183,
      "mean_token_accuracy": 0.15371278077363967,
      "num_tokens": 5714623.0,
      "step": 2255
    },
    {
      "entropy": 6.7304175853729244,
      "epoch": 0.2608193883439123,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997601322767258,
      "loss": 6.6328,
      "mean_token_accuracy": 0.15081430226564407,
      "num_tokens": 5727415.0,
      "step": 2260
    },
    {
      "entropy": 6.825870084762573,
      "epoch": 0.26139642238892097,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997582236113087,
      "loss": 6.6852,
      "mean_token_accuracy": 0.14831779673695564,
      "num_tokens": 5740129.0,
      "step": 2265
    },
    {
      "entropy": 6.793635606765747,
      "epoch": 0.2619734564339296,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997563073862603,
      "loss": 6.6556,
      "mean_token_accuracy": 0.1551389992237091,
      "num_tokens": 5752469.0,
      "step": 2270
    },
    {
      "entropy": 6.750844240188599,
      "epoch": 0.26255049047893825,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997543836016453,
      "loss": 6.7453,
      "mean_token_accuracy": 0.1527159981429577,
      "num_tokens": 5765408.0,
      "step": 2275
    },
    {
      "entropy": 6.8902946472167965,
      "epoch": 0.2631275245239469,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997524522575281,
      "loss": 6.6628,
      "mean_token_accuracy": 0.14618581384420395,
      "num_tokens": 5777450.0,
      "step": 2280
    },
    {
      "entropy": 6.762203741073608,
      "epoch": 0.2637045585689556,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997505133539738,
      "loss": 6.737,
      "mean_token_accuracy": 0.1433141678571701,
      "num_tokens": 5790435.0,
      "step": 2285
    },
    {
      "entropy": 6.87997465133667,
      "epoch": 0.2642815926139642,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997485668910476,
      "loss": 6.7089,
      "mean_token_accuracy": 0.14799562245607376,
      "num_tokens": 5803517.0,
      "step": 2290
    },
    {
      "entropy": 6.723853969573975,
      "epoch": 0.26485862665897286,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997466128688151,
      "loss": 6.6455,
      "mean_token_accuracy": 0.15912182182073592,
      "num_tokens": 5815802.0,
      "step": 2295
    },
    {
      "entropy": 6.725922870635986,
      "epoch": 0.26543566070398156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499744651287342,
      "loss": 6.6842,
      "mean_token_accuracy": 0.1475355327129364,
      "num_tokens": 5827807.0,
      "step": 2300
    },
    {
      "entropy": 6.819564151763916,
      "epoch": 0.2660126947489902,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499742682146694,
      "loss": 6.7051,
      "mean_token_accuracy": 0.1480211853981018,
      "num_tokens": 5840759.0,
      "step": 2305
    },
    {
      "entropy": 6.844948768615723,
      "epoch": 0.26658972879399884,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997407054469377,
      "loss": 6.7258,
      "mean_token_accuracy": 0.1477212592959404,
      "num_tokens": 5853117.0,
      "step": 2310
    },
    {
      "entropy": 6.71142544746399,
      "epoch": 0.2671667628390075,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997387211881392,
      "loss": 6.6453,
      "mean_token_accuracy": 0.154846253991127,
      "num_tokens": 5866717.0,
      "step": 2315
    },
    {
      "entropy": 6.773086881637573,
      "epoch": 0.26774379688401617,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997367293703656,
      "loss": 6.699,
      "mean_token_accuracy": 0.1512269750237465,
      "num_tokens": 5879017.0,
      "step": 2320
    },
    {
      "entropy": 6.793334484100342,
      "epoch": 0.2683208309290248,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997347299936837,
      "loss": 6.6803,
      "mean_token_accuracy": 0.15060366541147233,
      "num_tokens": 5891481.0,
      "step": 2325
    },
    {
      "entropy": 6.785113191604614,
      "epoch": 0.26889786497403345,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997327230581608,
      "loss": 6.6951,
      "mean_token_accuracy": 0.14829754531383516,
      "num_tokens": 5903389.0,
      "step": 2330
    },
    {
      "entropy": 6.79466233253479,
      "epoch": 0.26947489901904215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997307085638644,
      "loss": 6.6014,
      "mean_token_accuracy": 0.15171003192663193,
      "num_tokens": 5916084.0,
      "step": 2335
    },
    {
      "entropy": 6.751241445541382,
      "epoch": 0.2700519330640508,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004997286865108621,
      "loss": 6.6476,
      "mean_token_accuracy": 0.15270082801580429,
      "num_tokens": 5927797.0,
      "step": 2340
    },
    {
      "entropy": 6.715710830688477,
      "epoch": 0.2706289671090594,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997266568992222,
      "loss": 6.5641,
      "mean_token_accuracy": 0.15439386516809464,
      "num_tokens": 5941886.0,
      "step": 2345
    },
    {
      "entropy": 6.745696973800659,
      "epoch": 0.27120600115406807,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004997246197290128,
      "loss": 6.6874,
      "mean_token_accuracy": 0.15251569598913192,
      "num_tokens": 5953655.0,
      "step": 2350
    },
    {
      "entropy": 6.920389127731323,
      "epoch": 0.27178303519907676,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997225750003024,
      "loss": 6.6943,
      "mean_token_accuracy": 0.14675507098436355,
      "num_tokens": 5967818.0,
      "step": 2355
    },
    {
      "entropy": 6.726615953445434,
      "epoch": 0.2723600692440854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997205227131599,
      "loss": 6.6363,
      "mean_token_accuracy": 0.15357123017311097,
      "num_tokens": 5980742.0,
      "step": 2360
    },
    {
      "entropy": 6.803928184509277,
      "epoch": 0.27293710328909404,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997184628676542,
      "loss": 6.6679,
      "mean_token_accuracy": 0.1451663374900818,
      "num_tokens": 5993263.0,
      "step": 2365
    },
    {
      "entropy": 6.834203672409058,
      "epoch": 0.27351413733410274,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997163954638546,
      "loss": 6.749,
      "mean_token_accuracy": 0.13940424844622612,
      "num_tokens": 6007820.0,
      "step": 2370
    },
    {
      "entropy": 6.820643520355224,
      "epoch": 0.2740911713791114,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997143205018306,
      "loss": 6.6946,
      "mean_token_accuracy": 0.14526703357696533,
      "num_tokens": 6020665.0,
      "step": 2375
    },
    {
      "entropy": 6.768551397323608,
      "epoch": 0.27466820542412,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997122379816523,
      "loss": 6.6328,
      "mean_token_accuracy": 0.15664122849702836,
      "num_tokens": 6032847.0,
      "step": 2380
    },
    {
      "entropy": 6.738122892379761,
      "epoch": 0.27524523946912866,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997101479033893,
      "loss": 6.6885,
      "mean_token_accuracy": 0.1447917029261589,
      "num_tokens": 6045291.0,
      "step": 2385
    },
    {
      "entropy": 6.767072296142578,
      "epoch": 0.27582227351413735,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997080502671122,
      "loss": 6.6632,
      "mean_token_accuracy": 0.152347169816494,
      "num_tokens": 6057872.0,
      "step": 2390
    },
    {
      "entropy": 6.724477624893188,
      "epoch": 0.276399307559146,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997059450728913,
      "loss": 6.6387,
      "mean_token_accuracy": 0.14954917132854462,
      "num_tokens": 6070252.0,
      "step": 2395
    },
    {
      "entropy": 6.8534856796264645,
      "epoch": 0.27697634160415463,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997038323207977,
      "loss": 6.7345,
      "mean_token_accuracy": 0.14333323687314986,
      "num_tokens": 6083728.0,
      "step": 2400
    },
    {
      "entropy": 6.765815353393554,
      "epoch": 0.2775533756491633,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997017120109022,
      "loss": 6.6816,
      "mean_token_accuracy": 0.14334091544151306,
      "num_tokens": 6097140.0,
      "step": 2405
    },
    {
      "entropy": 6.6975670337677,
      "epoch": 0.27813040969417197,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996995841432762,
      "loss": 6.6616,
      "mean_token_accuracy": 0.14706796407699585,
      "num_tokens": 6109518.0,
      "step": 2410
    },
    {
      "entropy": 6.822784471511841,
      "epoch": 0.2787074437391806,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996974487179915,
      "loss": 6.6353,
      "mean_token_accuracy": 0.1506706103682518,
      "num_tokens": 6122160.0,
      "step": 2415
    },
    {
      "entropy": 6.683734941482544,
      "epoch": 0.27928447778418924,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996953057351195,
      "loss": 6.6792,
      "mean_token_accuracy": 0.14972876906394958,
      "num_tokens": 6134638.0,
      "step": 2420
    },
    {
      "entropy": 6.794296360015869,
      "epoch": 0.27986151182919794,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996931551947327,
      "loss": 6.716,
      "mean_token_accuracy": 0.14403074458241463,
      "num_tokens": 6147142.0,
      "step": 2425
    },
    {
      "entropy": 6.91330852508545,
      "epoch": 0.2804385458742066,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996909970969031,
      "loss": 6.7142,
      "mean_token_accuracy": 0.1503919020295143,
      "num_tokens": 6159137.0,
      "step": 2430
    },
    {
      "entropy": 6.667835092544555,
      "epoch": 0.2810155799192152,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996888314417034,
      "loss": 6.7049,
      "mean_token_accuracy": 0.14543969705700874,
      "num_tokens": 6171663.0,
      "step": 2435
    },
    {
      "entropy": 6.919120836257934,
      "epoch": 0.2815926139642239,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996866582292067,
      "loss": 6.6955,
      "mean_token_accuracy": 0.15087713599205016,
      "num_tokens": 6184282.0,
      "step": 2440
    },
    {
      "entropy": 6.769654130935669,
      "epoch": 0.28216964800923255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996844774594856,
      "loss": 6.6964,
      "mean_token_accuracy": 0.145013540238142,
      "num_tokens": 6197243.0,
      "step": 2445
    },
    {
      "entropy": 6.731862545013428,
      "epoch": 0.2827466820542412,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004996822891326138,
      "loss": 6.6172,
      "mean_token_accuracy": 0.14766587913036347,
      "num_tokens": 6209061.0,
      "step": 2450
    },
    {
      "entropy": 6.731410026550293,
      "epoch": 0.28332371609924983,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004996800932486648,
      "loss": 6.5871,
      "mean_token_accuracy": 0.15210840851068497,
      "num_tokens": 6223497.0,
      "step": 2455
    },
    {
      "entropy": 6.796885824203491,
      "epoch": 0.28390075014425853,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996778898077126,
      "loss": 6.6323,
      "mean_token_accuracy": 0.15437232851982116,
      "num_tokens": 6236395.0,
      "step": 2460
    },
    {
      "entropy": 6.646933364868164,
      "epoch": 0.28447778418926717,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996756788098309,
      "loss": 6.6089,
      "mean_token_accuracy": 0.15198341608047486,
      "num_tokens": 6248700.0,
      "step": 2465
    },
    {
      "entropy": 6.803117084503174,
      "epoch": 0.2850548182342758,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004996734602550945,
      "loss": 6.6497,
      "mean_token_accuracy": 0.14864367991685867,
      "num_tokens": 6262728.0,
      "step": 2470
    },
    {
      "entropy": 6.692935037612915,
      "epoch": 0.2856318522792845,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996712341435779,
      "loss": 6.6138,
      "mean_token_accuracy": 0.1492878705263138,
      "num_tokens": 6276007.0,
      "step": 2475
    },
    {
      "entropy": 6.8393807888031,
      "epoch": 0.28620888632429314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996690004753559,
      "loss": 6.7461,
      "mean_token_accuracy": 0.14636218845844268,
      "num_tokens": 6289033.0,
      "step": 2480
    },
    {
      "entropy": 6.737071847915649,
      "epoch": 0.2867859203693018,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996667592505037,
      "loss": 6.6844,
      "mean_token_accuracy": 0.14224475771188735,
      "num_tokens": 6300994.0,
      "step": 2485
    },
    {
      "entropy": 6.738783359527588,
      "epoch": 0.2873629544143104,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996645104690967,
      "loss": 6.6377,
      "mean_token_accuracy": 0.15198902413249016,
      "num_tokens": 6313447.0,
      "step": 2490
    },
    {
      "entropy": 6.803956127166748,
      "epoch": 0.2879399884593191,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996622541312103,
      "loss": 6.623,
      "mean_token_accuracy": 0.151736593991518,
      "num_tokens": 6326622.0,
      "step": 2495
    },
    {
      "entropy": 6.725099420547485,
      "epoch": 0.28851702250432776,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996599902369207,
      "loss": 6.6404,
      "mean_token_accuracy": 0.1515824615955353,
      "num_tokens": 6339286.0,
      "step": 2500
    },
    {
      "entropy": 6.731478691101074,
      "epoch": 0.2890940565493364,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996577187863039,
      "loss": 6.6548,
      "mean_token_accuracy": 0.14619150757789612,
      "num_tokens": 6351869.0,
      "step": 2505
    },
    {
      "entropy": 6.72581992149353,
      "epoch": 0.2896710905943451,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996554397794364,
      "loss": 6.5781,
      "mean_token_accuracy": 0.15530783385038377,
      "num_tokens": 6363974.0,
      "step": 2510
    },
    {
      "entropy": 6.732598114013672,
      "epoch": 0.29024812463935373,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996531532163947,
      "loss": 6.5792,
      "mean_token_accuracy": 0.15085918009281157,
      "num_tokens": 6376569.0,
      "step": 2515
    },
    {
      "entropy": 6.7090099334716795,
      "epoch": 0.2908251586843624,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996508590972558,
      "loss": 6.5992,
      "mean_token_accuracy": 0.15587113946676254,
      "num_tokens": 6388830.0,
      "step": 2520
    },
    {
      "entropy": 6.735125112533569,
      "epoch": 0.291402192729371,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996485574220969,
      "loss": 6.609,
      "mean_token_accuracy": 0.1497942864894867,
      "num_tokens": 6401542.0,
      "step": 2525
    },
    {
      "entropy": 6.62659764289856,
      "epoch": 0.2919792267743797,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996462481909953,
      "loss": 6.5982,
      "mean_token_accuracy": 0.15247822999954225,
      "num_tokens": 6414123.0,
      "step": 2530
    },
    {
      "entropy": 6.759478330612183,
      "epoch": 0.29255626081938835,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996439314040287,
      "loss": 6.6562,
      "mean_token_accuracy": 0.1472500592470169,
      "num_tokens": 6427269.0,
      "step": 2535
    },
    {
      "entropy": 6.744736099243164,
      "epoch": 0.293133294864397,
      "grad_norm": 0.875,
      "learning_rate": 0.000499641607061275,
      "loss": 6.7066,
      "mean_token_accuracy": 0.1504643440246582,
      "num_tokens": 6442296.0,
      "step": 2540
    },
    {
      "entropy": 6.817773962020874,
      "epoch": 0.2937103289094057,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996392751628126,
      "loss": 6.7218,
      "mean_token_accuracy": 0.1421102836728096,
      "num_tokens": 6456954.0,
      "step": 2545
    },
    {
      "entropy": 6.750686311721802,
      "epoch": 0.2942873629544143,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004996369357087196,
      "loss": 6.6965,
      "mean_token_accuracy": 0.14400525987148285,
      "num_tokens": 6470139.0,
      "step": 2550
    },
    {
      "entropy": 6.839855909347534,
      "epoch": 0.29486439699942296,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499634588699075,
      "loss": 6.6334,
      "mean_token_accuracy": 0.15152083933353425,
      "num_tokens": 6483106.0,
      "step": 2555
    },
    {
      "entropy": 6.647458600997925,
      "epoch": 0.2954414310444316,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004996322341339575,
      "loss": 6.6561,
      "mean_token_accuracy": 0.14724331647157668,
      "num_tokens": 6495921.0,
      "step": 2560
    },
    {
      "entropy": 6.759826612472534,
      "epoch": 0.2960184650894403,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996298720134465,
      "loss": 6.5553,
      "mean_token_accuracy": 0.1497732073068619,
      "num_tokens": 6508408.0,
      "step": 2565
    },
    {
      "entropy": 6.677283096313476,
      "epoch": 0.29659549913444894,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996275023376213,
      "loss": 6.6444,
      "mean_token_accuracy": 0.1454513743519783,
      "num_tokens": 6522081.0,
      "step": 2570
    },
    {
      "entropy": 6.73228816986084,
      "epoch": 0.2971725331794576,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996251251065615,
      "loss": 6.6204,
      "mean_token_accuracy": 0.14908404648303986,
      "num_tokens": 6535065.0,
      "step": 2575
    },
    {
      "entropy": 6.740736770629883,
      "epoch": 0.2977495672244662,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996227403203473,
      "loss": 6.5985,
      "mean_token_accuracy": 0.1506045341491699,
      "num_tokens": 6547972.0,
      "step": 2580
    },
    {
      "entropy": 6.668179225921631,
      "epoch": 0.2983266012694749,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004996203479790589,
      "loss": 6.5494,
      "mean_token_accuracy": 0.156391242146492,
      "num_tokens": 6559935.0,
      "step": 2585
    },
    {
      "entropy": 6.756309700012207,
      "epoch": 0.29890363531448355,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996179480827764,
      "loss": 6.5639,
      "mean_token_accuracy": 0.1533658154308796,
      "num_tokens": 6572373.0,
      "step": 2590
    },
    {
      "entropy": 6.736523103713989,
      "epoch": 0.2994806693594922,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499615540631581,
      "loss": 6.6492,
      "mean_token_accuracy": 0.14879586845636367,
      "num_tokens": 6585846.0,
      "step": 2595
    },
    {
      "entropy": 6.774682474136353,
      "epoch": 0.3000577034045009,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996131256255534,
      "loss": 6.7078,
      "mean_token_accuracy": 0.14609526395797728,
      "num_tokens": 6600387.0,
      "step": 2600
    },
    {
      "entropy": 6.721484851837158,
      "epoch": 0.3006347374495095,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996107030647749,
      "loss": 6.6359,
      "mean_token_accuracy": 0.1481912150979042,
      "num_tokens": 6613142.0,
      "step": 2605
    },
    {
      "entropy": 6.684381151199341,
      "epoch": 0.30121177149451817,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996082729493269,
      "loss": 6.5359,
      "mean_token_accuracy": 0.15520427972078324,
      "num_tokens": 6626273.0,
      "step": 2610
    },
    {
      "entropy": 6.717306995391846,
      "epoch": 0.3017888055395268,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004996058352792913,
      "loss": 6.6311,
      "mean_token_accuracy": 0.14817688912153243,
      "num_tokens": 6639831.0,
      "step": 2615
    },
    {
      "entropy": 6.759371042251587,
      "epoch": 0.3023658395845355,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049960339005475,
      "loss": 6.6912,
      "mean_token_accuracy": 0.1456967532634735,
      "num_tokens": 6652776.0,
      "step": 2620
    },
    {
      "entropy": 6.834853887557983,
      "epoch": 0.30294287362954414,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996009372757852,
      "loss": 6.6504,
      "mean_token_accuracy": 0.15175580829381943,
      "num_tokens": 6665465.0,
      "step": 2625
    },
    {
      "entropy": 6.598134279251099,
      "epoch": 0.3035199076745528,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004995984769424795,
      "loss": 6.521,
      "mean_token_accuracy": 0.16483442038297652,
      "num_tokens": 6677472.0,
      "step": 2630
    },
    {
      "entropy": 6.655421113967895,
      "epoch": 0.3040969417195615,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995960090549155,
      "loss": 6.5558,
      "mean_token_accuracy": 0.15292261838912963,
      "num_tokens": 6689145.0,
      "step": 2635
    },
    {
      "entropy": 6.692539262771606,
      "epoch": 0.3046739757645701,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995935336131764,
      "loss": 6.5359,
      "mean_token_accuracy": 0.15541132986545564,
      "num_tokens": 6701520.0,
      "step": 2640
    },
    {
      "entropy": 6.7484640121459964,
      "epoch": 0.30525100980957875,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995910506173452,
      "loss": 6.6392,
      "mean_token_accuracy": 0.15335483923554422,
      "num_tokens": 6715615.0,
      "step": 2645
    },
    {
      "entropy": 6.786902809143067,
      "epoch": 0.3058280438545874,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995885600675059,
      "loss": 6.6493,
      "mean_token_accuracy": 0.14820180535316468,
      "num_tokens": 6728772.0,
      "step": 2650
    },
    {
      "entropy": 6.723455142974854,
      "epoch": 0.3064050778995961,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995860619637415,
      "loss": 6.5904,
      "mean_token_accuracy": 0.1488635666668415,
      "num_tokens": 6741454.0,
      "step": 2655
    },
    {
      "entropy": 6.680184316635132,
      "epoch": 0.30698211194460473,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995835563061367,
      "loss": 6.6329,
      "mean_token_accuracy": 0.15059148371219636,
      "num_tokens": 6753983.0,
      "step": 2660
    },
    {
      "entropy": 6.8095183849334715,
      "epoch": 0.30755914598961337,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995810430947756,
      "loss": 6.6587,
      "mean_token_accuracy": 0.14956516027450562,
      "num_tokens": 6767377.0,
      "step": 2665
    },
    {
      "entropy": 6.731013202667237,
      "epoch": 0.30813618003462206,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995785223297426,
      "loss": 6.6165,
      "mean_token_accuracy": 0.15526754558086395,
      "num_tokens": 6779220.0,
      "step": 2670
    },
    {
      "entropy": 6.739847898483276,
      "epoch": 0.3087132140796307,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995759940111225,
      "loss": 6.5862,
      "mean_token_accuracy": 0.15217508971691132,
      "num_tokens": 6792202.0,
      "step": 2675
    },
    {
      "entropy": 6.649475145339966,
      "epoch": 0.30929024812463934,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995734581390005,
      "loss": 6.5915,
      "mean_token_accuracy": 0.15380042493343354,
      "num_tokens": 6804988.0,
      "step": 2680
    },
    {
      "entropy": 6.723600292205811,
      "epoch": 0.309867282169648,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995709147134617,
      "loss": 6.625,
      "mean_token_accuracy": 0.15084846913814545,
      "num_tokens": 6817311.0,
      "step": 2685
    },
    {
      "entropy": 6.757253742218017,
      "epoch": 0.3104443162146567,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004995683637345919,
      "loss": 6.5985,
      "mean_token_accuracy": 0.15270360708236694,
      "num_tokens": 6829875.0,
      "step": 2690
    },
    {
      "entropy": 6.733402490615845,
      "epoch": 0.3110213502596653,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004995658052024765,
      "loss": 6.5826,
      "mean_token_accuracy": 0.1535247102379799,
      "num_tokens": 6843080.0,
      "step": 2695
    },
    {
      "entropy": 6.633068037033081,
      "epoch": 0.31159838430467396,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995632391172019,
      "loss": 6.5728,
      "mean_token_accuracy": 0.15194655805826188,
      "num_tokens": 6857202.0,
      "step": 2700
    },
    {
      "entropy": 6.745692777633667,
      "epoch": 0.31217541834968265,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004995606654788543,
      "loss": 6.6612,
      "mean_token_accuracy": 0.1453693374991417,
      "num_tokens": 6870801.0,
      "step": 2705
    },
    {
      "entropy": 6.764104557037354,
      "epoch": 0.3127524523946913,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004995580842875202,
      "loss": 6.6112,
      "mean_token_accuracy": 0.14864688366651535,
      "num_tokens": 6883084.0,
      "step": 2710
    },
    {
      "entropy": 6.743693923950195,
      "epoch": 0.31332948643969993,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995554955432866,
      "loss": 6.6421,
      "mean_token_accuracy": 0.14455916807055474,
      "num_tokens": 6895309.0,
      "step": 2715
    },
    {
      "entropy": 6.723890781402588,
      "epoch": 0.3139065204847086,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995528992462404,
      "loss": 6.6268,
      "mean_token_accuracy": 0.15532680079340935,
      "num_tokens": 6907646.0,
      "step": 2720
    },
    {
      "entropy": 6.747570991516113,
      "epoch": 0.31448355452971727,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499550295396469,
      "loss": 6.5977,
      "mean_token_accuracy": 0.15922294557094574,
      "num_tokens": 6919769.0,
      "step": 2725
    },
    {
      "entropy": 6.665530633926392,
      "epoch": 0.3150605885747259,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995476839940598,
      "loss": 6.5651,
      "mean_token_accuracy": 0.15450860559940338,
      "num_tokens": 6933348.0,
      "step": 2730
    },
    {
      "entropy": 6.678213691711425,
      "epoch": 0.31563762261973455,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499545065039101,
      "loss": 6.6257,
      "mean_token_accuracy": 0.14856788590550424,
      "num_tokens": 6945105.0,
      "step": 2735
    },
    {
      "entropy": 6.677836799621582,
      "epoch": 0.31621465666474324,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995424385316803,
      "loss": 6.6046,
      "mean_token_accuracy": 0.147177591919899,
      "num_tokens": 6958009.0,
      "step": 2740
    },
    {
      "entropy": 6.699324274063111,
      "epoch": 0.3167916907097519,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995398044718863,
      "loss": 6.5217,
      "mean_token_accuracy": 0.15502108931541442,
      "num_tokens": 6970156.0,
      "step": 2745
    },
    {
      "entropy": 6.679452276229858,
      "epoch": 0.3173687247547605,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004995371628598074,
      "loss": 6.6122,
      "mean_token_accuracy": 0.15149652659893037,
      "num_tokens": 6983431.0,
      "step": 2750
    },
    {
      "entropy": 6.764488601684571,
      "epoch": 0.31794575879976916,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995345136955328,
      "loss": 6.6527,
      "mean_token_accuracy": 0.14685067385435105,
      "num_tokens": 6997470.0,
      "step": 2755
    },
    {
      "entropy": 6.6844353675842285,
      "epoch": 0.31852279284477786,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995318569791512,
      "loss": 6.6035,
      "mean_token_accuracy": 0.1576828345656395,
      "num_tokens": 7010337.0,
      "step": 2760
    },
    {
      "entropy": 6.792466688156128,
      "epoch": 0.3190998268897865,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004995291927107522,
      "loss": 6.635,
      "mean_token_accuracy": 0.1491129755973816,
      "num_tokens": 7023950.0,
      "step": 2765
    },
    {
      "entropy": 6.659047603607178,
      "epoch": 0.31967686093479514,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995265208904252,
      "loss": 6.5956,
      "mean_token_accuracy": 0.1566797286272049,
      "num_tokens": 7036409.0,
      "step": 2770
    },
    {
      "entropy": 6.750910663604737,
      "epoch": 0.32025389497980383,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995238415182603,
      "loss": 6.6074,
      "mean_token_accuracy": 0.14797423183918,
      "num_tokens": 7049409.0,
      "step": 2775
    },
    {
      "entropy": 6.705856990814209,
      "epoch": 0.3208309290248125,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995211545943476,
      "loss": 6.6512,
      "mean_token_accuracy": 0.15140025615692138,
      "num_tokens": 7062120.0,
      "step": 2780
    },
    {
      "entropy": 6.757962846755982,
      "epoch": 0.3214079630698211,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995184601187772,
      "loss": 6.6431,
      "mean_token_accuracy": 0.14404307901859284,
      "num_tokens": 7073150.0,
      "step": 2785
    },
    {
      "entropy": 6.742756462097168,
      "epoch": 0.32198499711482975,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049951575809164,
      "loss": 6.5825,
      "mean_token_accuracy": 0.14299354702234268,
      "num_tokens": 7085927.0,
      "step": 2790
    },
    {
      "entropy": 6.657029342651367,
      "epoch": 0.32256203115983845,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995130485130269,
      "loss": 6.6388,
      "mean_token_accuracy": 0.1517026126384735,
      "num_tokens": 7097607.0,
      "step": 2795
    },
    {
      "entropy": 6.784142065048218,
      "epoch": 0.3231390652048471,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995103313830289,
      "loss": 6.5707,
      "mean_token_accuracy": 0.1529722362756729,
      "num_tokens": 7109771.0,
      "step": 2800
    },
    {
      "entropy": 6.7327779769897464,
      "epoch": 0.3237160992498557,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995076067017373,
      "loss": 6.6322,
      "mean_token_accuracy": 0.1470279797911644,
      "num_tokens": 7122757.0,
      "step": 2805
    },
    {
      "entropy": 6.654575204849243,
      "epoch": 0.3242931332948644,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004995048744692439,
      "loss": 6.6093,
      "mean_token_accuracy": 0.14694169014692307,
      "num_tokens": 7135516.0,
      "step": 2810
    },
    {
      "entropy": 6.756155252456665,
      "epoch": 0.32487016733987306,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995021346856405,
      "loss": 6.6842,
      "mean_token_accuracy": 0.14694853872060776,
      "num_tokens": 7149125.0,
      "step": 2815
    },
    {
      "entropy": 6.73067193031311,
      "epoch": 0.3254472013848817,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994993873510196,
      "loss": 6.6307,
      "mean_token_accuracy": 0.14298643171787262,
      "num_tokens": 7160723.0,
      "step": 2820
    },
    {
      "entropy": 6.687489080429077,
      "epoch": 0.32602423542989034,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994966324654731,
      "loss": 6.6089,
      "mean_token_accuracy": 0.14513298645615577,
      "num_tokens": 7174551.0,
      "step": 2825
    },
    {
      "entropy": 6.7687946319580075,
      "epoch": 0.32660126947489904,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499493870029094,
      "loss": 6.5889,
      "mean_token_accuracy": 0.15016230791807175,
      "num_tokens": 7186764.0,
      "step": 2830
    },
    {
      "entropy": 6.696437358856201,
      "epoch": 0.3271783035199077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994911000419749,
      "loss": 6.5842,
      "mean_token_accuracy": 0.15165507793426514,
      "num_tokens": 7199645.0,
      "step": 2835
    },
    {
      "entropy": 6.612226581573486,
      "epoch": 0.3277553375649163,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994883225042093,
      "loss": 6.5105,
      "mean_token_accuracy": 0.16406965851783753,
      "num_tokens": 7212579.0,
      "step": 2840
    },
    {
      "entropy": 6.673196649551391,
      "epoch": 0.328332371609925,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994855374158905,
      "loss": 6.5433,
      "mean_token_accuracy": 0.1573927402496338,
      "num_tokens": 7225139.0,
      "step": 2845
    },
    {
      "entropy": 6.728290367126465,
      "epoch": 0.32890940565493365,
      "grad_norm": 0.875,
      "learning_rate": 0.000499482744777112,
      "loss": 6.6863,
      "mean_token_accuracy": 0.14344870671629906,
      "num_tokens": 7239220.0,
      "step": 2850
    },
    {
      "entropy": 6.801548290252685,
      "epoch": 0.3294864396999423,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499479944587968,
      "loss": 6.7103,
      "mean_token_accuracy": 0.15049158781766891,
      "num_tokens": 7252613.0,
      "step": 2855
    },
    {
      "entropy": 6.632060527801514,
      "epoch": 0.33006347374495093,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994771368485526,
      "loss": 6.536,
      "mean_token_accuracy": 0.15601779520511627,
      "num_tokens": 7263658.0,
      "step": 2860
    },
    {
      "entropy": 6.740309047698974,
      "epoch": 0.3306405077899596,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994743215589602,
      "loss": 6.5617,
      "mean_token_accuracy": 0.1619183152914047,
      "num_tokens": 7275782.0,
      "step": 2865
    },
    {
      "entropy": 6.594912481307984,
      "epoch": 0.33121754183496827,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994714987192854,
      "loss": 6.618,
      "mean_token_accuracy": 0.1574898988008499,
      "num_tokens": 7288536.0,
      "step": 2870
    },
    {
      "entropy": 6.694319057464599,
      "epoch": 0.3317945758799769,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994686683296234,
      "loss": 6.576,
      "mean_token_accuracy": 0.15647158324718474,
      "num_tokens": 7300480.0,
      "step": 2875
    },
    {
      "entropy": 6.702283906936645,
      "epoch": 0.3323716099249856,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994658303900691,
      "loss": 6.6199,
      "mean_token_accuracy": 0.15028671324253082,
      "num_tokens": 7313168.0,
      "step": 2880
    },
    {
      "entropy": 6.6668816089630125,
      "epoch": 0.33294864396999424,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994629849007182,
      "loss": 6.5767,
      "mean_token_accuracy": 0.15539561957120895,
      "num_tokens": 7327347.0,
      "step": 2885
    },
    {
      "entropy": 6.722062492370606,
      "epoch": 0.3335256780150029,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994601318616663,
      "loss": 6.61,
      "mean_token_accuracy": 0.14843571111559867,
      "num_tokens": 7339955.0,
      "step": 2890
    },
    {
      "entropy": 6.70476131439209,
      "epoch": 0.3341027120600115,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994572712730092,
      "loss": 6.5879,
      "mean_token_accuracy": 0.148174050450325,
      "num_tokens": 7353586.0,
      "step": 2895
    },
    {
      "entropy": 6.692771768569946,
      "epoch": 0.3346797461050202,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994544031348434,
      "loss": 6.5577,
      "mean_token_accuracy": 0.15284216701984404,
      "num_tokens": 7365802.0,
      "step": 2900
    },
    {
      "entropy": 6.6868119716644285,
      "epoch": 0.33525678015002885,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994515274472653,
      "loss": 6.5256,
      "mean_token_accuracy": 0.14853521138429643,
      "num_tokens": 7377864.0,
      "step": 2905
    },
    {
      "entropy": 6.647774076461792,
      "epoch": 0.3358338141950375,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994486442103715,
      "loss": 6.4741,
      "mean_token_accuracy": 0.15462038069963455,
      "num_tokens": 7389921.0,
      "step": 2910
    },
    {
      "entropy": 6.653657722473144,
      "epoch": 0.3364108482400462,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499445753424259,
      "loss": 6.54,
      "mean_token_accuracy": 0.15041984468698502,
      "num_tokens": 7402765.0,
      "step": 2915
    },
    {
      "entropy": 6.737467956542969,
      "epoch": 0.33698788228505483,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004994428550890251,
      "loss": 6.5701,
      "mean_token_accuracy": 0.14831982553005219,
      "num_tokens": 7416330.0,
      "step": 2920
    },
    {
      "entropy": 6.6251367092132565,
      "epoch": 0.33756491633006347,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994399492047672,
      "loss": 6.515,
      "mean_token_accuracy": 0.16390403658151625,
      "num_tokens": 7428701.0,
      "step": 2925
    },
    {
      "entropy": 6.723229122161865,
      "epoch": 0.3381419503750721,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994370357715832,
      "loss": 6.5249,
      "mean_token_accuracy": 0.15543388575315475,
      "num_tokens": 7440687.0,
      "step": 2930
    },
    {
      "entropy": 6.654600191116333,
      "epoch": 0.3387189844200808,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499434114789571,
      "loss": 6.6147,
      "mean_token_accuracy": 0.14533780366182328,
      "num_tokens": 7454448.0,
      "step": 2935
    },
    {
      "entropy": 6.746973848342895,
      "epoch": 0.33929601846508944,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994311862588286,
      "loss": 6.5318,
      "mean_token_accuracy": 0.15422469675540923,
      "num_tokens": 7466722.0,
      "step": 2940
    },
    {
      "entropy": 6.612504863739014,
      "epoch": 0.3398730525100981,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994282501794551,
      "loss": 6.5434,
      "mean_token_accuracy": 0.15694365054368972,
      "num_tokens": 7478574.0,
      "step": 2945
    },
    {
      "entropy": 6.630198383331299,
      "epoch": 0.3404500865551068,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994253065515486,
      "loss": 6.4416,
      "mean_token_accuracy": 0.1585552930831909,
      "num_tokens": 7490424.0,
      "step": 2950
    },
    {
      "entropy": 6.687573194503784,
      "epoch": 0.3410271206001154,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994223553752084,
      "loss": 6.5932,
      "mean_token_accuracy": 0.15581245720386505,
      "num_tokens": 7502832.0,
      "step": 2955
    },
    {
      "entropy": 6.616685962677002,
      "epoch": 0.34160415464512406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994193966505339,
      "loss": 6.5195,
      "mean_token_accuracy": 0.1566762462258339,
      "num_tokens": 7514151.0,
      "step": 2960
    },
    {
      "entropy": 6.685769557952881,
      "epoch": 0.3421811886901327,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004994164303776244,
      "loss": 6.5417,
      "mean_token_accuracy": 0.15701379925012587,
      "num_tokens": 7526877.0,
      "step": 2965
    },
    {
      "entropy": 6.673000526428223,
      "epoch": 0.3427582227351414,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994134565565797,
      "loss": 6.5302,
      "mean_token_accuracy": 0.156466044485569,
      "num_tokens": 7538858.0,
      "step": 2970
    },
    {
      "entropy": 6.721205186843872,
      "epoch": 0.34333525678015003,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994104751875,
      "loss": 6.56,
      "mean_token_accuracy": 0.1546006865799427,
      "num_tokens": 7550804.0,
      "step": 2975
    },
    {
      "entropy": 6.626689672470093,
      "epoch": 0.3439122908251587,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994074862704854,
      "loss": 6.4896,
      "mean_token_accuracy": 0.16036185026168823,
      "num_tokens": 7562269.0,
      "step": 2980
    },
    {
      "entropy": 6.6253760814666744,
      "epoch": 0.3444893248701673,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994044898056363,
      "loss": 6.5821,
      "mean_token_accuracy": 0.14939415752887725,
      "num_tokens": 7573695.0,
      "step": 2985
    },
    {
      "entropy": 6.688880825042725,
      "epoch": 0.345066358915176,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994014857930538,
      "loss": 6.5974,
      "mean_token_accuracy": 0.14651145711541175,
      "num_tokens": 7585943.0,
      "step": 2990
    },
    {
      "entropy": 6.667485666275025,
      "epoch": 0.34564339296018465,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993984742328387,
      "loss": 6.5789,
      "mean_token_accuracy": 0.1516515865921974,
      "num_tokens": 7598326.0,
      "step": 2995
    },
    {
      "entropy": 6.715835285186768,
      "epoch": 0.3462204270051933,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993954551250925,
      "loss": 6.5636,
      "mean_token_accuracy": 0.14854548275470733,
      "num_tokens": 7609950.0,
      "step": 3000
    },
    {
      "epoch": 0.3462204270051933,
      "eval_entropy": 6.571954663856687,
      "eval_loss": 6.600578308105469,
      "eval_mean_token_accuracy": 0.15436312960360685,
      "eval_num_tokens": 7609950.0,
      "eval_runtime": 17.7878,
      "eval_samples_per_second": 1581.762,
      "eval_steps_per_second": 197.72,
      "step": 3000
    },
    {
      "entropy": 6.67913236618042,
      "epoch": 0.346797461050202,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004993924284699165,
      "loss": 6.5187,
      "mean_token_accuracy": 0.15324495881795883,
      "num_tokens": 7621987.0,
      "step": 3005
    },
    {
      "entropy": 6.662750720977783,
      "epoch": 0.3473744950952106,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993893942674127,
      "loss": 6.5837,
      "mean_token_accuracy": 0.15105323493480682,
      "num_tokens": 7634801.0,
      "step": 3010
    },
    {
      "entropy": 6.713637208938598,
      "epoch": 0.34795152914021926,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499386352517683,
      "loss": 6.5376,
      "mean_token_accuracy": 0.15901586338877677,
      "num_tokens": 7647942.0,
      "step": 3015
    },
    {
      "entropy": 6.683552551269531,
      "epoch": 0.3485285631852279,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00049938330322083,
      "loss": 6.5491,
      "mean_token_accuracy": 0.1520805388689041,
      "num_tokens": 7662361.0,
      "step": 3020
    },
    {
      "entropy": 6.628801774978638,
      "epoch": 0.3491055972302366,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993802463769558,
      "loss": 6.5425,
      "mean_token_accuracy": 0.15337807536125184,
      "num_tokens": 7674161.0,
      "step": 3025
    },
    {
      "entropy": 6.704723119735718,
      "epoch": 0.34968263127524524,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993771819861636,
      "loss": 6.5468,
      "mean_token_accuracy": 0.15563344210386276,
      "num_tokens": 7686957.0,
      "step": 3030
    },
    {
      "entropy": 6.638054227828979,
      "epoch": 0.3502596653202539,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993741100485564,
      "loss": 6.5082,
      "mean_token_accuracy": 0.15760557949543,
      "num_tokens": 7699436.0,
      "step": 3035
    },
    {
      "entropy": 6.569077968597412,
      "epoch": 0.35083669936526257,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993710305642374,
      "loss": 6.5267,
      "mean_token_accuracy": 0.15751972198486328,
      "num_tokens": 7711794.0,
      "step": 3040
    },
    {
      "entropy": 6.666848373413086,
      "epoch": 0.3514137334102712,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993679435333102,
      "loss": 6.478,
      "mean_token_accuracy": 0.1532701849937439,
      "num_tokens": 7723496.0,
      "step": 3045
    },
    {
      "entropy": 6.619588375091553,
      "epoch": 0.35199076745527985,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993648489558788,
      "loss": 6.497,
      "mean_token_accuracy": 0.1570889875292778,
      "num_tokens": 7735980.0,
      "step": 3050
    },
    {
      "entropy": 6.641847801208496,
      "epoch": 0.3525678015002885,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993617468320471,
      "loss": 6.5698,
      "mean_token_accuracy": 0.15105772390961647,
      "num_tokens": 7750145.0,
      "step": 3055
    },
    {
      "entropy": 6.659021091461182,
      "epoch": 0.3531448355452972,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993586371619193,
      "loss": 6.5764,
      "mean_token_accuracy": 0.1522957906126976,
      "num_tokens": 7762284.0,
      "step": 3060
    },
    {
      "entropy": 6.5531312942504885,
      "epoch": 0.3537218695903058,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993555199456004,
      "loss": 6.5305,
      "mean_token_accuracy": 0.1619846776127815,
      "num_tokens": 7776225.0,
      "step": 3065
    },
    {
      "entropy": 6.609269666671753,
      "epoch": 0.35429890363531447,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993523951831949,
      "loss": 6.4629,
      "mean_token_accuracy": 0.16548260748386384,
      "num_tokens": 7789408.0,
      "step": 3070
    },
    {
      "entropy": 6.703277111053467,
      "epoch": 0.35487593768032316,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993492628748081,
      "loss": 6.5309,
      "mean_token_accuracy": 0.14926318675279618,
      "num_tokens": 7801824.0,
      "step": 3075
    },
    {
      "entropy": 6.5838579654693605,
      "epoch": 0.3554529717253318,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004993461230205453,
      "loss": 6.5234,
      "mean_token_accuracy": 0.1576508030295372,
      "num_tokens": 7814167.0,
      "step": 3080
    },
    {
      "entropy": 6.691663599014282,
      "epoch": 0.35603000577034044,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004993429756205121,
      "loss": 6.5595,
      "mean_token_accuracy": 0.1552141159772873,
      "num_tokens": 7826736.0,
      "step": 3085
    },
    {
      "entropy": 6.6293723583221436,
      "epoch": 0.3566070398153491,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993398206748142,
      "loss": 6.5219,
      "mean_token_accuracy": 0.15335862934589387,
      "num_tokens": 7838626.0,
      "step": 3090
    },
    {
      "entropy": 6.691901588439942,
      "epoch": 0.3571840738603578,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993366581835581,
      "loss": 6.5186,
      "mean_token_accuracy": 0.15390921831130983,
      "num_tokens": 7850608.0,
      "step": 3095
    },
    {
      "entropy": 6.63633918762207,
      "epoch": 0.3577611079053664,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004993334881468498,
      "loss": 6.5063,
      "mean_token_accuracy": 0.1534891128540039,
      "num_tokens": 7862800.0,
      "step": 3100
    },
    {
      "entropy": 6.715207052230835,
      "epoch": 0.35833814195037506,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993303105647961,
      "loss": 6.5752,
      "mean_token_accuracy": 0.15088694244623185,
      "num_tokens": 7876198.0,
      "step": 3105
    },
    {
      "entropy": 6.708647918701172,
      "epoch": 0.35891517599538375,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993271254375038,
      "loss": 6.5954,
      "mean_token_accuracy": 0.14844163358211518,
      "num_tokens": 7888129.0,
      "step": 3110
    },
    {
      "entropy": 6.652002191543579,
      "epoch": 0.3594922100403924,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00049932393276508,
      "loss": 6.5536,
      "mean_token_accuracy": 0.15066457390785218,
      "num_tokens": 7900453.0,
      "step": 3115
    },
    {
      "entropy": 6.685463333129883,
      "epoch": 0.36006924408540103,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993207325476323,
      "loss": 6.5567,
      "mean_token_accuracy": 0.14931840747594832,
      "num_tokens": 7913538.0,
      "step": 3120
    },
    {
      "entropy": 6.679757308959961,
      "epoch": 0.36064627813040967,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993175247852681,
      "loss": 6.5217,
      "mean_token_accuracy": 0.15380802154541015,
      "num_tokens": 7927744.0,
      "step": 3125
    },
    {
      "entropy": 6.594105291366577,
      "epoch": 0.36122331217541837,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993143094780954,
      "loss": 6.4676,
      "mean_token_accuracy": 0.1560771107673645,
      "num_tokens": 7939975.0,
      "step": 3130
    },
    {
      "entropy": 6.670260858535767,
      "epoch": 0.361800346220427,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993110866262224,
      "loss": 6.522,
      "mean_token_accuracy": 0.15731455832719804,
      "num_tokens": 7951692.0,
      "step": 3135
    },
    {
      "entropy": 6.630655765533447,
      "epoch": 0.36237738026543564,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993078562297573,
      "loss": 6.5325,
      "mean_token_accuracy": 0.15282207280397414,
      "num_tokens": 7964887.0,
      "step": 3140
    },
    {
      "entropy": 6.624306344985962,
      "epoch": 0.36295441431044434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993046182888089,
      "loss": 6.5486,
      "mean_token_accuracy": 0.15523382872343064,
      "num_tokens": 7977608.0,
      "step": 3145
    },
    {
      "entropy": 6.6301249980926515,
      "epoch": 0.363531448355453,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993013728034861,
      "loss": 6.4665,
      "mean_token_accuracy": 0.15506149977445602,
      "num_tokens": 7990094.0,
      "step": 3150
    },
    {
      "entropy": 6.691576719284058,
      "epoch": 0.3641084824004616,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499298119773898,
      "loss": 6.5336,
      "mean_token_accuracy": 0.15450148284435272,
      "num_tokens": 8003477.0,
      "step": 3155
    },
    {
      "entropy": 6.611363458633423,
      "epoch": 0.36468551644547026,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992948592001541,
      "loss": 6.4861,
      "mean_token_accuracy": 0.1541801556944847,
      "num_tokens": 8016241.0,
      "step": 3160
    },
    {
      "entropy": 6.59058780670166,
      "epoch": 0.36526255049047895,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499291591082364,
      "loss": 6.4575,
      "mean_token_accuracy": 0.16361010670661927,
      "num_tokens": 8028352.0,
      "step": 3165
    },
    {
      "entropy": 6.609756946563721,
      "epoch": 0.3658395845354876,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992883154206377,
      "loss": 6.5508,
      "mean_token_accuracy": 0.1522105261683464,
      "num_tokens": 8040215.0,
      "step": 3170
    },
    {
      "entropy": 6.655618906021118,
      "epoch": 0.36641661858049623,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992850322150853,
      "loss": 6.4528,
      "mean_token_accuracy": 0.16034831553697587,
      "num_tokens": 8051633.0,
      "step": 3175
    },
    {
      "entropy": 6.64651517868042,
      "epoch": 0.36699365262550493,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992817414658172,
      "loss": 6.5185,
      "mean_token_accuracy": 0.1551498845219612,
      "num_tokens": 8063390.0,
      "step": 3180
    },
    {
      "entropy": 6.624983882904052,
      "epoch": 0.36757068667051357,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992784431729442,
      "loss": 6.5044,
      "mean_token_accuracy": 0.15131230503320695,
      "num_tokens": 8077561.0,
      "step": 3185
    },
    {
      "entropy": 6.6048698902130125,
      "epoch": 0.3681477207155222,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992751373365771,
      "loss": 6.5723,
      "mean_token_accuracy": 0.15384764671325685,
      "num_tokens": 8091345.0,
      "step": 3190
    },
    {
      "entropy": 6.599915266036987,
      "epoch": 0.36872475476053085,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992718239568273,
      "loss": 6.4195,
      "mean_token_accuracy": 0.16362291127443312,
      "num_tokens": 8104179.0,
      "step": 3195
    },
    {
      "entropy": 6.637220239639282,
      "epoch": 0.36930178880553954,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499268503033806,
      "loss": 6.5777,
      "mean_token_accuracy": 0.14956896454095842,
      "num_tokens": 8115831.0,
      "step": 3200
    },
    {
      "entropy": 6.605573606491089,
      "epoch": 0.3698788228505482,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004992651745676249,
      "loss": 6.5124,
      "mean_token_accuracy": 0.15962110906839372,
      "num_tokens": 8129048.0,
      "step": 3205
    },
    {
      "entropy": 6.649493169784546,
      "epoch": 0.3704558568955568,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992618385583962,
      "loss": 6.5172,
      "mean_token_accuracy": 0.14941440969705583,
      "num_tokens": 8141679.0,
      "step": 3210
    },
    {
      "entropy": 6.617839097976685,
      "epoch": 0.3710328909405655,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992584950062319,
      "loss": 6.4865,
      "mean_token_accuracy": 0.15477514415979385,
      "num_tokens": 8155004.0,
      "step": 3215
    },
    {
      "entropy": 6.538065576553345,
      "epoch": 0.37160992498557416,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992551439112446,
      "loss": 6.4259,
      "mean_token_accuracy": 0.1586252972483635,
      "num_tokens": 8166718.0,
      "step": 3220
    },
    {
      "entropy": 6.632493591308593,
      "epoch": 0.3721869590305828,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992517852735469,
      "loss": 6.6244,
      "mean_token_accuracy": 0.14812181890010834,
      "num_tokens": 8180011.0,
      "step": 3225
    },
    {
      "entropy": 6.646015453338623,
      "epoch": 0.37276399307559144,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992484190932517,
      "loss": 6.4352,
      "mean_token_accuracy": 0.1578085280954838,
      "num_tokens": 8194800.0,
      "step": 3230
    },
    {
      "entropy": 6.664063167572022,
      "epoch": 0.37334102712060013,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992450453704723,
      "loss": 6.5604,
      "mean_token_accuracy": 0.15252675265073776,
      "num_tokens": 8207153.0,
      "step": 3235
    },
    {
      "entropy": 6.711281061172485,
      "epoch": 0.3739180611656088,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992416641053223,
      "loss": 6.5446,
      "mean_token_accuracy": 0.15358139276504518,
      "num_tokens": 8221135.0,
      "step": 3240
    },
    {
      "entropy": 6.494730043411255,
      "epoch": 0.3744950952106174,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992382752979153,
      "loss": 6.4123,
      "mean_token_accuracy": 0.16235440969467163,
      "num_tokens": 8234271.0,
      "step": 3245
    },
    {
      "entropy": 6.650647115707398,
      "epoch": 0.3750721292556261,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992348789483651,
      "loss": 6.5545,
      "mean_token_accuracy": 0.14874081835150718,
      "num_tokens": 8246710.0,
      "step": 3250
    },
    {
      "entropy": 6.687919473648071,
      "epoch": 0.37564916330063475,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004992314750567864,
      "loss": 6.6077,
      "mean_token_accuracy": 0.15173837468028067,
      "num_tokens": 8259817.0,
      "step": 3255
    },
    {
      "entropy": 6.620677471160889,
      "epoch": 0.3762261973456434,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992280636232933,
      "loss": 6.5395,
      "mean_token_accuracy": 0.16084663420915604,
      "num_tokens": 8274359.0,
      "step": 3260
    },
    {
      "entropy": 6.721046495437622,
      "epoch": 0.376803231390652,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992246446480007,
      "loss": 6.5053,
      "mean_token_accuracy": 0.15376538634300232,
      "num_tokens": 8286355.0,
      "step": 3265
    },
    {
      "entropy": 6.596151685714721,
      "epoch": 0.3773802654356607,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992212181310235,
      "loss": 6.4847,
      "mean_token_accuracy": 0.15147178918123244,
      "num_tokens": 8298338.0,
      "step": 3270
    },
    {
      "entropy": 6.520630645751953,
      "epoch": 0.37795729948066936,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499217784072477,
      "loss": 6.4627,
      "mean_token_accuracy": 0.15949076861143113,
      "num_tokens": 8312367.0,
      "step": 3275
    },
    {
      "entropy": 6.663025856018066,
      "epoch": 0.378534333525678,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004992143424724768,
      "loss": 6.5539,
      "mean_token_accuracy": 0.14739133268594742,
      "num_tokens": 8325577.0,
      "step": 3280
    },
    {
      "entropy": 6.645350456237793,
      "epoch": 0.3791113675706867,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992108933311385,
      "loss": 6.5688,
      "mean_token_accuracy": 0.15200639963150026,
      "num_tokens": 8338965.0,
      "step": 3285
    },
    {
      "entropy": 6.735022687911988,
      "epoch": 0.37968840161569534,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992074366485782,
      "loss": 6.5193,
      "mean_token_accuracy": 0.14615978449583053,
      "num_tokens": 8351669.0,
      "step": 3290
    },
    {
      "entropy": 6.575160884857178,
      "epoch": 0.380265435660704,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499203972424912,
      "loss": 6.5674,
      "mean_token_accuracy": 0.15271605104207991,
      "num_tokens": 8364194.0,
      "step": 3295
    },
    {
      "entropy": 6.621188068389893,
      "epoch": 0.3808424697057126,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004992005006602567,
      "loss": 6.5759,
      "mean_token_accuracy": 0.15298435539007188,
      "num_tokens": 8377001.0,
      "step": 3300
    },
    {
      "entropy": 6.630554580688477,
      "epoch": 0.3814195037507213,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991970213547289,
      "loss": 6.5695,
      "mean_token_accuracy": 0.1499115854501724,
      "num_tokens": 8389247.0,
      "step": 3305
    },
    {
      "entropy": 6.665640449523925,
      "epoch": 0.38199653779572995,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991935345084457,
      "loss": 6.4557,
      "mean_token_accuracy": 0.15480488911271095,
      "num_tokens": 8402915.0,
      "step": 3310
    },
    {
      "entropy": 6.620656824111938,
      "epoch": 0.3825735718407386,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991900401215243,
      "loss": 6.566,
      "mean_token_accuracy": 0.15210421830415727,
      "num_tokens": 8414590.0,
      "step": 3315
    },
    {
      "entropy": 6.6845362186431885,
      "epoch": 0.3831506058857473,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991865381940822,
      "loss": 6.5415,
      "mean_token_accuracy": 0.15634194612503052,
      "num_tokens": 8427337.0,
      "step": 3320
    },
    {
      "entropy": 6.532070064544678,
      "epoch": 0.3837276399307559,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991830287262374,
      "loss": 6.4353,
      "mean_token_accuracy": 0.16090213805437087,
      "num_tokens": 8440095.0,
      "step": 3325
    },
    {
      "entropy": 6.582758522033691,
      "epoch": 0.38430467397576457,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991795117181077,
      "loss": 6.4573,
      "mean_token_accuracy": 0.1525129869580269,
      "num_tokens": 8452448.0,
      "step": 3330
    },
    {
      "entropy": 6.6684760570526125,
      "epoch": 0.3848817080207732,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991759871698113,
      "loss": 6.4847,
      "mean_token_accuracy": 0.1601237565279007,
      "num_tokens": 8465289.0,
      "step": 3335
    },
    {
      "entropy": 6.593679332733155,
      "epoch": 0.3854587420657819,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991724550814671,
      "loss": 6.5024,
      "mean_token_accuracy": 0.15915523767471312,
      "num_tokens": 8477365.0,
      "step": 3340
    },
    {
      "entropy": 6.6565875053405765,
      "epoch": 0.38603577611079054,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991689154531937,
      "loss": 6.5706,
      "mean_token_accuracy": 0.15299642086029053,
      "num_tokens": 8489104.0,
      "step": 3345
    },
    {
      "entropy": 6.565132427215576,
      "epoch": 0.3866128101557992,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004991653682851102,
      "loss": 6.4436,
      "mean_token_accuracy": 0.1581375189125538,
      "num_tokens": 8501509.0,
      "step": 3350
    },
    {
      "entropy": 6.575328779220581,
      "epoch": 0.3871898442008078,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000499161813577336,
      "loss": 6.4948,
      "mean_token_accuracy": 0.15168995559215545,
      "num_tokens": 8514224.0,
      "step": 3355
    },
    {
      "entropy": 6.675197601318359,
      "epoch": 0.3877668782458165,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991582513299903,
      "loss": 6.5789,
      "mean_token_accuracy": 0.14614480137825012,
      "num_tokens": 8526689.0,
      "step": 3360
    },
    {
      "entropy": 6.582357358932495,
      "epoch": 0.38834391229082516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991546815431932,
      "loss": 6.6078,
      "mean_token_accuracy": 0.1479400023818016,
      "num_tokens": 8539952.0,
      "step": 3365
    },
    {
      "entropy": 6.705312252044678,
      "epoch": 0.3889209463358338,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499151104217065,
      "loss": 6.5948,
      "mean_token_accuracy": 0.1492103323340416,
      "num_tokens": 8553428.0,
      "step": 3370
    },
    {
      "entropy": 6.649981832504272,
      "epoch": 0.3894979803808425,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991475193517254,
      "loss": 6.4562,
      "mean_token_accuracy": 0.15850950852036477,
      "num_tokens": 8566066.0,
      "step": 3375
    },
    {
      "entropy": 6.554257202148437,
      "epoch": 0.39007501442585113,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991439269472956,
      "loss": 6.41,
      "mean_token_accuracy": 0.15376016795635222,
      "num_tokens": 8579211.0,
      "step": 3380
    },
    {
      "entropy": 6.668615388870239,
      "epoch": 0.39065204847085977,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991403270038961,
      "loss": 6.521,
      "mean_token_accuracy": 0.1519201323390007,
      "num_tokens": 8591647.0,
      "step": 3385
    },
    {
      "entropy": 6.518864393234253,
      "epoch": 0.3912290825158684,
      "grad_norm": 0.875,
      "learning_rate": 0.000499136719521648,
      "loss": 6.5075,
      "mean_token_accuracy": 0.15120508521795273,
      "num_tokens": 8604279.0,
      "step": 3390
    },
    {
      "entropy": 6.625417041778564,
      "epoch": 0.3918061165608771,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004991331045006726,
      "loss": 6.4162,
      "mean_token_accuracy": 0.1614367976784706,
      "num_tokens": 8617722.0,
      "step": 3395
    },
    {
      "entropy": 6.527361679077148,
      "epoch": 0.39238315060588574,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991294819410916,
      "loss": 6.4441,
      "mean_token_accuracy": 0.16820953413844109,
      "num_tokens": 8630250.0,
      "step": 3400
    },
    {
      "entropy": 6.482899379730225,
      "epoch": 0.3929601846508944,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991258518430268,
      "loss": 6.3346,
      "mean_token_accuracy": 0.16297557950019836,
      "num_tokens": 8641900.0,
      "step": 3405
    },
    {
      "entropy": 6.555102777481079,
      "epoch": 0.3935372186959031,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991222142066003,
      "loss": 6.4545,
      "mean_token_accuracy": 0.15731057673692703,
      "num_tokens": 8654588.0,
      "step": 3410
    },
    {
      "entropy": 6.648878288269043,
      "epoch": 0.3941142527409117,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991185690319345,
      "loss": 6.4915,
      "mean_token_accuracy": 0.14755185917019845,
      "num_tokens": 8668111.0,
      "step": 3415
    },
    {
      "entropy": 6.570490169525146,
      "epoch": 0.39469128678592036,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000499114916319152,
      "loss": 6.5662,
      "mean_token_accuracy": 0.1540154278278351,
      "num_tokens": 8681461.0,
      "step": 3420
    },
    {
      "entropy": 6.658483409881592,
      "epoch": 0.395268320830929,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991112560683755,
      "loss": 6.4903,
      "mean_token_accuracy": 0.1537899613380432,
      "num_tokens": 8692893.0,
      "step": 3425
    },
    {
      "entropy": 6.593234586715698,
      "epoch": 0.3958453548759377,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991075882797283,
      "loss": 6.3817,
      "mean_token_accuracy": 0.16625459790229796,
      "num_tokens": 8704704.0,
      "step": 3430
    },
    {
      "entropy": 6.585725784301758,
      "epoch": 0.39642238892094633,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991039129533337,
      "loss": 6.5528,
      "mean_token_accuracy": 0.15255533307790756,
      "num_tokens": 8717595.0,
      "step": 3435
    },
    {
      "entropy": 6.544450187683106,
      "epoch": 0.396999422965955,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991002300893152,
      "loss": 6.4927,
      "mean_token_accuracy": 0.15547696650028228,
      "num_tokens": 8729712.0,
      "step": 3440
    },
    {
      "entropy": 6.623501539230347,
      "epoch": 0.39757645701096367,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990965396877969,
      "loss": 6.4136,
      "mean_token_accuracy": 0.15489595979452134,
      "num_tokens": 8742668.0,
      "step": 3445
    },
    {
      "entropy": 6.637320518493652,
      "epoch": 0.3981534910559723,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990928417489027,
      "loss": 6.5008,
      "mean_token_accuracy": 0.16184985637664795,
      "num_tokens": 8754755.0,
      "step": 3450
    },
    {
      "entropy": 6.598210096359253,
      "epoch": 0.39873052510098095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990891362727572,
      "loss": 6.4325,
      "mean_token_accuracy": 0.16320631355047227,
      "num_tokens": 8766063.0,
      "step": 3455
    },
    {
      "entropy": 6.624905776977539,
      "epoch": 0.3993075591459896,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990854232594848,
      "loss": 6.4722,
      "mean_token_accuracy": 0.163354229927063,
      "num_tokens": 8779226.0,
      "step": 3460
    },
    {
      "entropy": 6.56977071762085,
      "epoch": 0.3998845931909983,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990817027092106,
      "loss": 6.5052,
      "mean_token_accuracy": 0.1581609547138214,
      "num_tokens": 8791404.0,
      "step": 3465
    },
    {
      "entropy": 6.596388483047486,
      "epoch": 0.4004616272360069,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004990779746220595,
      "loss": 6.5466,
      "mean_token_accuracy": 0.1522767573595047,
      "num_tokens": 8805306.0,
      "step": 3470
    },
    {
      "entropy": 6.5737920761108395,
      "epoch": 0.40103866128101556,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990742389981572,
      "loss": 6.4273,
      "mean_token_accuracy": 0.16027601957321166,
      "num_tokens": 8817850.0,
      "step": 3475
    },
    {
      "entropy": 6.558056163787842,
      "epoch": 0.40161569532602426,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499070495837629,
      "loss": 6.4522,
      "mean_token_accuracy": 0.16423497945070267,
      "num_tokens": 8830765.0,
      "step": 3480
    },
    {
      "entropy": 6.598183536529541,
      "epoch": 0.4021927293710329,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004990667451406012,
      "loss": 6.4909,
      "mean_token_accuracy": 0.15179530829191207,
      "num_tokens": 8843329.0,
      "step": 3485
    },
    {
      "entropy": 6.6653501987457275,
      "epoch": 0.40276976341604154,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004990629869071995,
      "loss": 6.4857,
      "mean_token_accuracy": 0.15301345139741898,
      "num_tokens": 8856712.0,
      "step": 3490
    },
    {
      "entropy": 6.470282936096192,
      "epoch": 0.4033467974610502,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990592211375506,
      "loss": 6.4469,
      "mean_token_accuracy": 0.1624319449067116,
      "num_tokens": 8868364.0,
      "step": 3495
    },
    {
      "entropy": 6.570513820648193,
      "epoch": 0.4039238315060589,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004990554478317811,
      "loss": 6.4316,
      "mean_token_accuracy": 0.15485603660345076,
      "num_tokens": 8881335.0,
      "step": 3500
    },
    {
      "entropy": 6.5659960269927975,
      "epoch": 0.4045008655510675,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990516669900179,
      "loss": 6.5002,
      "mean_token_accuracy": 0.15902469158172608,
      "num_tokens": 8894049.0,
      "step": 3505
    },
    {
      "entropy": 6.558427095413208,
      "epoch": 0.40507789959607615,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990478786123882,
      "loss": 6.4024,
      "mean_token_accuracy": 0.1596820518374443,
      "num_tokens": 8906056.0,
      "step": 3510
    },
    {
      "entropy": 6.625905513763428,
      "epoch": 0.40565493364108485,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990440826990193,
      "loss": 6.4801,
      "mean_token_accuracy": 0.15464828312397003,
      "num_tokens": 8919606.0,
      "step": 3515
    },
    {
      "entropy": 6.57070255279541,
      "epoch": 0.4062319676860935,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990402792500392,
      "loss": 6.4484,
      "mean_token_accuracy": 0.16032859534025193,
      "num_tokens": 8934733.0,
      "step": 3520
    },
    {
      "entropy": 6.487190055847168,
      "epoch": 0.4068090017311021,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990364682655754,
      "loss": 6.4802,
      "mean_token_accuracy": 0.15721859708428382,
      "num_tokens": 8948331.0,
      "step": 3525
    },
    {
      "entropy": 6.619977712631226,
      "epoch": 0.40738603577611077,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990326497457564,
      "loss": 6.4614,
      "mean_token_accuracy": 0.1596258044242859,
      "num_tokens": 8961171.0,
      "step": 3530
    },
    {
      "entropy": 6.556022262573242,
      "epoch": 0.40796306982111946,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990288236907104,
      "loss": 6.5243,
      "mean_token_accuracy": 0.15410750806331636,
      "num_tokens": 8973164.0,
      "step": 3535
    },
    {
      "entropy": 6.643644666671753,
      "epoch": 0.4085401038661281,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990249901005661,
      "loss": 6.5542,
      "mean_token_accuracy": 0.14475183486938475,
      "num_tokens": 8985981.0,
      "step": 3540
    },
    {
      "entropy": 6.619608354568482,
      "epoch": 0.40911713791113674,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990211489754527,
      "loss": 6.4342,
      "mean_token_accuracy": 0.15859266966581345,
      "num_tokens": 8999473.0,
      "step": 3545
    },
    {
      "entropy": 6.4508928775787355,
      "epoch": 0.40969417195614544,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004990173003154993,
      "loss": 6.3771,
      "mean_token_accuracy": 0.16602665036916733,
      "num_tokens": 9010876.0,
      "step": 3550
    },
    {
      "entropy": 6.635811138153076,
      "epoch": 0.4102712060011541,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004990134441208351,
      "loss": 6.5016,
      "mean_token_accuracy": 0.1551976703107357,
      "num_tokens": 9024704.0,
      "step": 3555
    },
    {
      "entropy": 6.547682857513427,
      "epoch": 0.4108482400461627,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004990095803915901,
      "loss": 6.4859,
      "mean_token_accuracy": 0.15543424487113952,
      "num_tokens": 9037006.0,
      "step": 3560
    },
    {
      "entropy": 6.650226306915283,
      "epoch": 0.41142527409117136,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990057091278942,
      "loss": 6.5471,
      "mean_token_accuracy": 0.14769547432661057,
      "num_tokens": 9050205.0,
      "step": 3565
    },
    {
      "entropy": 6.662499713897705,
      "epoch": 0.41200230813618005,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004990018303298773,
      "loss": 6.4379,
      "mean_token_accuracy": 0.15080624148249627,
      "num_tokens": 9062900.0,
      "step": 3570
    },
    {
      "entropy": 6.431814289093017,
      "epoch": 0.4125793421811887,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989979439976703,
      "loss": 6.395,
      "mean_token_accuracy": 0.15929965376853944,
      "num_tokens": 9074767.0,
      "step": 3575
    },
    {
      "entropy": 6.664768648147583,
      "epoch": 0.41315637622619733,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989940501314037,
      "loss": 6.5162,
      "mean_token_accuracy": 0.15215157121419906,
      "num_tokens": 9087314.0,
      "step": 3580
    },
    {
      "entropy": 6.590992450714111,
      "epoch": 0.413733410271206,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004989901487312084,
      "loss": 6.4838,
      "mean_token_accuracy": 0.15035763382911682,
      "num_tokens": 9099543.0,
      "step": 3585
    },
    {
      "entropy": 6.6410894870758055,
      "epoch": 0.41431044431621467,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989862397972157,
      "loss": 6.425,
      "mean_token_accuracy": 0.15633125454187394,
      "num_tokens": 9112262.0,
      "step": 3590
    },
    {
      "entropy": 6.4641694068908695,
      "epoch": 0.4148874783612233,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989823233295572,
      "loss": 6.3407,
      "mean_token_accuracy": 0.16114797592163085,
      "num_tokens": 9124444.0,
      "step": 3595
    },
    {
      "entropy": 6.454566192626953,
      "epoch": 0.41546451240623195,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004989783993283644,
      "loss": 6.4023,
      "mean_token_accuracy": 0.16097055971622468,
      "num_tokens": 9137848.0,
      "step": 3600
    },
    {
      "entropy": 6.589309740066528,
      "epoch": 0.41604154645124064,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989744677937694,
      "loss": 6.4211,
      "mean_token_accuracy": 0.15603204369544982,
      "num_tokens": 9148985.0,
      "step": 3605
    },
    {
      "entropy": 6.562459945678711,
      "epoch": 0.4166185804962493,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989705287259046,
      "loss": 6.4676,
      "mean_token_accuracy": 0.15446533113718033,
      "num_tokens": 9161313.0,
      "step": 3610
    },
    {
      "entropy": 6.5443140983581545,
      "epoch": 0.4171956145412579,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004989665821249021,
      "loss": 6.4517,
      "mean_token_accuracy": 0.16134824752807617,
      "num_tokens": 9173616.0,
      "step": 3615
    },
    {
      "entropy": 6.630168104171753,
      "epoch": 0.4177726485862666,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498962627990895,
      "loss": 6.5187,
      "mean_token_accuracy": 0.14512302130460739,
      "num_tokens": 9185951.0,
      "step": 3620
    },
    {
      "entropy": 6.5845588684082035,
      "epoch": 0.41834968263127525,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989586663240161,
      "loss": 6.5113,
      "mean_token_accuracy": 0.15076626762747763,
      "num_tokens": 9199844.0,
      "step": 3625
    },
    {
      "entropy": 6.495019483566284,
      "epoch": 0.4189267166762839,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989546971243988,
      "loss": 6.421,
      "mean_token_accuracy": 0.16512578576803208,
      "num_tokens": 9213034.0,
      "step": 3630
    },
    {
      "entropy": 6.572731828689575,
      "epoch": 0.41950375072129253,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004989507203921763,
      "loss": 6.4122,
      "mean_token_accuracy": 0.1600019007921219,
      "num_tokens": 9224885.0,
      "step": 3635
    },
    {
      "entropy": 6.603681850433349,
      "epoch": 0.42008078476630123,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989467361274828,
      "loss": 6.4817,
      "mean_token_accuracy": 0.14893933087587358,
      "num_tokens": 9238960.0,
      "step": 3640
    },
    {
      "entropy": 6.601598501205444,
      "epoch": 0.42065781881130987,
      "grad_norm": 0.75,
      "learning_rate": 0.0004989427443304519,
      "loss": 6.4878,
      "mean_token_accuracy": 0.1529766984283924,
      "num_tokens": 9253075.0,
      "step": 3645
    },
    {
      "entropy": 6.541758871078491,
      "epoch": 0.4212348528563185,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498938745001218,
      "loss": 6.4409,
      "mean_token_accuracy": 0.1553158700466156,
      "num_tokens": 9266865.0,
      "step": 3650
    },
    {
      "entropy": 6.626510763168335,
      "epoch": 0.4218118869013272,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989347381399158,
      "loss": 6.4682,
      "mean_token_accuracy": 0.15190115571022034,
      "num_tokens": 9278046.0,
      "step": 3655
    },
    {
      "entropy": 6.4989166259765625,
      "epoch": 0.42238892094633584,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989307237466799,
      "loss": 6.3544,
      "mean_token_accuracy": 0.16506405919790268,
      "num_tokens": 9291998.0,
      "step": 3660
    },
    {
      "entropy": 6.5267595767974855,
      "epoch": 0.4229659549913445,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989267018216453,
      "loss": 6.4365,
      "mean_token_accuracy": 0.16413115411996843,
      "num_tokens": 9305023.0,
      "step": 3665
    },
    {
      "entropy": 6.513909244537354,
      "epoch": 0.4235429890363531,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989226723649473,
      "loss": 6.3962,
      "mean_token_accuracy": 0.15708214789628983,
      "num_tokens": 9316635.0,
      "step": 3670
    },
    {
      "entropy": 6.60302848815918,
      "epoch": 0.4241200230813618,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004989186353767214,
      "loss": 6.4542,
      "mean_token_accuracy": 0.15462954044342042,
      "num_tokens": 9329671.0,
      "step": 3675
    },
    {
      "entropy": 6.509460067749023,
      "epoch": 0.42469705712637046,
      "grad_norm": 0.75,
      "learning_rate": 0.0004989145908571035,
      "loss": 6.4708,
      "mean_token_accuracy": 0.14989196062088012,
      "num_tokens": 9342634.0,
      "step": 3680
    },
    {
      "entropy": 6.68083233833313,
      "epoch": 0.4252740911713791,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989105388062295,
      "loss": 6.4503,
      "mean_token_accuracy": 0.15909998267889022,
      "num_tokens": 9355096.0,
      "step": 3685
    },
    {
      "entropy": 6.558935117721558,
      "epoch": 0.4258511252163878,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989064792242358,
      "loss": 6.4575,
      "mean_token_accuracy": 0.15320901870727538,
      "num_tokens": 9367722.0,
      "step": 3690
    },
    {
      "entropy": 6.580481433868409,
      "epoch": 0.42642815926139643,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989024121112589,
      "loss": 6.5058,
      "mean_token_accuracy": 0.14978662133216858,
      "num_tokens": 9380247.0,
      "step": 3695
    },
    {
      "entropy": 6.554038381576538,
      "epoch": 0.4270051933064051,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988983374674356,
      "loss": 6.3572,
      "mean_token_accuracy": 0.16555936485528946,
      "num_tokens": 9392216.0,
      "step": 3700
    },
    {
      "entropy": 6.588551568984985,
      "epoch": 0.4275822273514137,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988942552929029,
      "loss": 6.4779,
      "mean_token_accuracy": 0.1517507925629616,
      "num_tokens": 9405215.0,
      "step": 3705
    },
    {
      "entropy": 6.522645664215088,
      "epoch": 0.4281592613964224,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988901655877981,
      "loss": 6.4372,
      "mean_token_accuracy": 0.15472906231880187,
      "num_tokens": 9417582.0,
      "step": 3710
    },
    {
      "entropy": 6.53475399017334,
      "epoch": 0.42873629544143105,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988860683522589,
      "loss": 6.4331,
      "mean_token_accuracy": 0.1537720039486885,
      "num_tokens": 9428680.0,
      "step": 3715
    },
    {
      "entropy": 6.597131443023682,
      "epoch": 0.4293133294864397,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498881963586423,
      "loss": 6.3344,
      "mean_token_accuracy": 0.1615703895688057,
      "num_tokens": 9441669.0,
      "step": 3720
    },
    {
      "entropy": 6.50098934173584,
      "epoch": 0.4298903635314484,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988778512904282,
      "loss": 6.4868,
      "mean_token_accuracy": 0.15616406500339508,
      "num_tokens": 9455214.0,
      "step": 3725
    },
    {
      "entropy": 6.5591755390167235,
      "epoch": 0.430467397576457,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988737314644134,
      "loss": 6.4619,
      "mean_token_accuracy": 0.15522471219301223,
      "num_tokens": 9468180.0,
      "step": 3730
    },
    {
      "entropy": 6.5727219581604,
      "epoch": 0.43104443162146566,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004988696041085168,
      "loss": 6.4449,
      "mean_token_accuracy": 0.16246393471956252,
      "num_tokens": 9481243.0,
      "step": 3735
    },
    {
      "entropy": 6.487103366851807,
      "epoch": 0.4316214656664743,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988654692228772,
      "loss": 6.3551,
      "mean_token_accuracy": 0.1678798109292984,
      "num_tokens": 9493511.0,
      "step": 3740
    },
    {
      "entropy": 6.497088432312012,
      "epoch": 0.432198499711483,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004988613268076335,
      "loss": 6.4093,
      "mean_token_accuracy": 0.1654896542429924,
      "num_tokens": 9505450.0,
      "step": 3745
    },
    {
      "entropy": 6.585122728347779,
      "epoch": 0.43277553375649164,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988571768629257,
      "loss": 6.4596,
      "mean_token_accuracy": 0.1581433355808258,
      "num_tokens": 9518684.0,
      "step": 3750
    },
    {
      "entropy": 6.498142337799072,
      "epoch": 0.4333525678015003,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004988530193888927,
      "loss": 6.4145,
      "mean_token_accuracy": 0.15784785747528077,
      "num_tokens": 9532064.0,
      "step": 3755
    },
    {
      "entropy": 6.487857294082642,
      "epoch": 0.4339296018465089,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988488543856747,
      "loss": 6.3626,
      "mean_token_accuracy": 0.1681128889322281,
      "num_tokens": 9543810.0,
      "step": 3760
    },
    {
      "entropy": 6.606877851486206,
      "epoch": 0.4345066358915176,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004988446818534115,
      "loss": 6.5149,
      "mean_token_accuracy": 0.14516288191080093,
      "num_tokens": 9556261.0,
      "step": 3765
    },
    {
      "entropy": 6.537605476379395,
      "epoch": 0.43508366993652625,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004988405017922438,
      "loss": 6.346,
      "mean_token_accuracy": 0.16130636185407637,
      "num_tokens": 9569627.0,
      "step": 3770
    },
    {
      "entropy": 6.5598901271820065,
      "epoch": 0.4356607039815349,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498836314202312,
      "loss": 6.4565,
      "mean_token_accuracy": 0.1545551188290119,
      "num_tokens": 9582263.0,
      "step": 3775
    },
    {
      "entropy": 6.507501220703125,
      "epoch": 0.4362377380265436,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988321190837568,
      "loss": 6.3332,
      "mean_token_accuracy": 0.16725715696811677,
      "num_tokens": 9594282.0,
      "step": 3780
    },
    {
      "entropy": 6.494114971160888,
      "epoch": 0.4368147720715522,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988279164367196,
      "loss": 6.3586,
      "mean_token_accuracy": 0.16182285398244858,
      "num_tokens": 9607159.0,
      "step": 3785
    },
    {
      "entropy": 6.535694646835327,
      "epoch": 0.43739180611656087,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004988237062613415,
      "loss": 6.3988,
      "mean_token_accuracy": 0.16683767586946488,
      "num_tokens": 9619221.0,
      "step": 3790
    },
    {
      "entropy": 6.552090501785278,
      "epoch": 0.4379688401615695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988194885577644,
      "loss": 6.4308,
      "mean_token_accuracy": 0.15698195844888688,
      "num_tokens": 9631956.0,
      "step": 3795
    },
    {
      "entropy": 6.558296966552734,
      "epoch": 0.4385458742065782,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988152633261299,
      "loss": 6.4351,
      "mean_token_accuracy": 0.1557241663336754,
      "num_tokens": 9643557.0,
      "step": 3800
    },
    {
      "entropy": 6.54206166267395,
      "epoch": 0.43912290825158684,
      "grad_norm": 0.875,
      "learning_rate": 0.00049881103056658,
      "loss": 6.397,
      "mean_token_accuracy": 0.15948179066181184,
      "num_tokens": 9655422.0,
      "step": 3805
    },
    {
      "entropy": 6.583641576766968,
      "epoch": 0.4396999422965955,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004988067902792575,
      "loss": 6.4727,
      "mean_token_accuracy": 0.15186914056539536,
      "num_tokens": 9669522.0,
      "step": 3810
    },
    {
      "entropy": 6.567377662658691,
      "epoch": 0.4402769763416042,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988025424643047,
      "loss": 6.4853,
      "mean_token_accuracy": 0.15547452867031097,
      "num_tokens": 9681301.0,
      "step": 3815
    },
    {
      "entropy": 6.49556827545166,
      "epoch": 0.4408540103866128,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004987982871218645,
      "loss": 6.4058,
      "mean_token_accuracy": 0.15341770350933076,
      "num_tokens": 9693894.0,
      "step": 3820
    },
    {
      "entropy": 6.581149435043335,
      "epoch": 0.44143104443162146,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004987940242520802,
      "loss": 6.503,
      "mean_token_accuracy": 0.14905156195163727,
      "num_tokens": 9706499.0,
      "step": 3825
    },
    {
      "entropy": 6.495326089859009,
      "epoch": 0.4420080784766301,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000498789753855095,
      "loss": 6.3531,
      "mean_token_accuracy": 0.16193671226501466,
      "num_tokens": 9718632.0,
      "step": 3830
    },
    {
      "entropy": 6.544300889968872,
      "epoch": 0.4425851125216388,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987854759310526,
      "loss": 6.4331,
      "mean_token_accuracy": 0.15969545245170594,
      "num_tokens": 9730451.0,
      "step": 3835
    },
    {
      "entropy": 6.594881200790406,
      "epoch": 0.44316214656664743,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987811904800968,
      "loss": 6.4535,
      "mean_token_accuracy": 0.15569856613874436,
      "num_tokens": 9742693.0,
      "step": 3840
    },
    {
      "entropy": 6.533166885375977,
      "epoch": 0.44373918061165607,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987768975023719,
      "loss": 6.4415,
      "mean_token_accuracy": 0.15766530483961105,
      "num_tokens": 9755642.0,
      "step": 3845
    },
    {
      "entropy": 6.5512227535247805,
      "epoch": 0.44431621465666477,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987725969980222,
      "loss": 6.4597,
      "mean_token_accuracy": 0.1554524153470993,
      "num_tokens": 9767674.0,
      "step": 3850
    },
    {
      "entropy": 6.490002059936524,
      "epoch": 0.4448932487016734,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987682889671923,
      "loss": 6.3795,
      "mean_token_accuracy": 0.16682353913784026,
      "num_tokens": 9781018.0,
      "step": 3855
    },
    {
      "entropy": 6.5620664119720455,
      "epoch": 0.44547028274668204,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004987639734100272,
      "loss": 6.4492,
      "mean_token_accuracy": 0.1582857385277748,
      "num_tokens": 9792521.0,
      "step": 3860
    },
    {
      "entropy": 6.507644081115723,
      "epoch": 0.4460473167916907,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004987596503266721,
      "loss": 6.4171,
      "mean_token_accuracy": 0.1616949424147606,
      "num_tokens": 9805356.0,
      "step": 3865
    },
    {
      "entropy": 6.488741016387939,
      "epoch": 0.4466243508366994,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987553197172722,
      "loss": 6.3975,
      "mean_token_accuracy": 0.16268157660961152,
      "num_tokens": 9818779.0,
      "step": 3870
    },
    {
      "entropy": 6.604960584640503,
      "epoch": 0.447201384881708,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004987509815819732,
      "loss": 6.4664,
      "mean_token_accuracy": 0.1494177907705307,
      "num_tokens": 9832328.0,
      "step": 3875
    },
    {
      "entropy": 6.522948360443115,
      "epoch": 0.44777841892671666,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987466359209213,
      "loss": 6.4074,
      "mean_token_accuracy": 0.1558647111058235,
      "num_tokens": 9846281.0,
      "step": 3880
    },
    {
      "entropy": 6.517709732055664,
      "epoch": 0.44835545297172535,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987422827342622,
      "loss": 6.4635,
      "mean_token_accuracy": 0.1561240956187248,
      "num_tokens": 9858737.0,
      "step": 3885
    },
    {
      "entropy": 6.549286508560181,
      "epoch": 0.448932487016734,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987379220221426,
      "loss": 6.4766,
      "mean_token_accuracy": 0.15599623322486877,
      "num_tokens": 9872852.0,
      "step": 3890
    },
    {
      "entropy": 6.555766773223877,
      "epoch": 0.44950952106174263,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987335537847092,
      "loss": 6.3896,
      "mean_token_accuracy": 0.1689491853117943,
      "num_tokens": 9886339.0,
      "step": 3895
    },
    {
      "entropy": 6.620915508270263,
      "epoch": 0.4500865551067513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987291780221088,
      "loss": 6.3958,
      "mean_token_accuracy": 0.15587402284145355,
      "num_tokens": 9897525.0,
      "step": 3900
    },
    {
      "entropy": 6.461075353622436,
      "epoch": 0.45066358915175997,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004987247947344887,
      "loss": 6.4155,
      "mean_token_accuracy": 0.16414239704608918,
      "num_tokens": 9909733.0,
      "step": 3905
    },
    {
      "entropy": 6.5133524417877195,
      "epoch": 0.4512406231967686,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004987204039219962,
      "loss": 6.4003,
      "mean_token_accuracy": 0.1509241595864296,
      "num_tokens": 9922503.0,
      "step": 3910
    },
    {
      "entropy": 6.562454795837402,
      "epoch": 0.45181765724177725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987160055847791,
      "loss": 6.4361,
      "mean_token_accuracy": 0.15687647461891174,
      "num_tokens": 9935517.0,
      "step": 3915
    },
    {
      "entropy": 6.568414878845215,
      "epoch": 0.45239469128678594,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004987115997229852,
      "loss": 6.4902,
      "mean_token_accuracy": 0.15789610743522645,
      "num_tokens": 9947663.0,
      "step": 3920
    },
    {
      "entropy": 6.452044486999512,
      "epoch": 0.4529717253317946,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004987071863367629,
      "loss": 6.2627,
      "mean_token_accuracy": 0.1615383207798004,
      "num_tokens": 9960924.0,
      "step": 3925
    },
    {
      "entropy": 6.513233327865601,
      "epoch": 0.4535487593768032,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987027654262604,
      "loss": 6.4018,
      "mean_token_accuracy": 0.15682899057865143,
      "num_tokens": 9973506.0,
      "step": 3930
    },
    {
      "entropy": 6.468669366836548,
      "epoch": 0.45412579342181186,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004986983369916264,
      "loss": 6.3917,
      "mean_token_accuracy": 0.16661963611841202,
      "num_tokens": 9988334.0,
      "step": 3935
    },
    {
      "entropy": 6.605206775665283,
      "epoch": 0.45470282746682056,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986939010330102,
      "loss": 6.4731,
      "mean_token_accuracy": 0.15531497150659562,
      "num_tokens": 10001295.0,
      "step": 3940
    },
    {
      "entropy": 6.533522653579712,
      "epoch": 0.4552798615118292,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986894575505606,
      "loss": 6.3616,
      "mean_token_accuracy": 0.15742329508066177,
      "num_tokens": 10012396.0,
      "step": 3945
    },
    {
      "entropy": 6.534685850143433,
      "epoch": 0.45585689555683784,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986850065444272,
      "loss": 6.4301,
      "mean_token_accuracy": 0.16010557413101195,
      "num_tokens": 10025918.0,
      "step": 3950
    },
    {
      "entropy": 6.627367115020752,
      "epoch": 0.45643392960184653,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986805480147598,
      "loss": 6.4338,
      "mean_token_accuracy": 0.15506452023983003,
      "num_tokens": 10038674.0,
      "step": 3955
    },
    {
      "entropy": 6.467591571807861,
      "epoch": 0.4570109636468552,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004986760819617082,
      "loss": 6.3963,
      "mean_token_accuracy": 0.162594798207283,
      "num_tokens": 10051623.0,
      "step": 3960
    },
    {
      "entropy": 6.561280107498169,
      "epoch": 0.4575879976918638,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004986716083854228,
      "loss": 6.4667,
      "mean_token_accuracy": 0.15451734513044357,
      "num_tokens": 10063731.0,
      "step": 3965
    },
    {
      "entropy": 6.623692226409912,
      "epoch": 0.45816503173687245,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986671272860538,
      "loss": 6.4732,
      "mean_token_accuracy": 0.15955658853054047,
      "num_tokens": 10075897.0,
      "step": 3970
    },
    {
      "entropy": 6.495250701904297,
      "epoch": 0.45874206578188115,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986626386637521,
      "loss": 6.4207,
      "mean_token_accuracy": 0.1547025889158249,
      "num_tokens": 10088605.0,
      "step": 3975
    },
    {
      "entropy": 6.545332717895508,
      "epoch": 0.4593190998268898,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004986581425186688,
      "loss": 6.4034,
      "mean_token_accuracy": 0.16402129381895064,
      "num_tokens": 10101758.0,
      "step": 3980
    },
    {
      "entropy": 6.557800436019898,
      "epoch": 0.4598961338718984,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986536388509548,
      "loss": 6.4477,
      "mean_token_accuracy": 0.15040433257818223,
      "num_tokens": 10113838.0,
      "step": 3985
    },
    {
      "entropy": 6.48162088394165,
      "epoch": 0.4604731679169071,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004986491276607618,
      "loss": 6.3992,
      "mean_token_accuracy": 0.1552007108926773,
      "num_tokens": 10127676.0,
      "step": 3990
    },
    {
      "entropy": 6.502557945251465,
      "epoch": 0.46105020196191576,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004986446089482416,
      "loss": 6.3374,
      "mean_token_accuracy": 0.1574241816997528,
      "num_tokens": 10139734.0,
      "step": 3995
    },
    {
      "entropy": 6.540197706222534,
      "epoch": 0.4616272360069244,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986400827135459,
      "loss": 6.452,
      "mean_token_accuracy": 0.15818918943405152,
      "num_tokens": 10151842.0,
      "step": 4000
    },
    {
      "entropy": 6.5954841613769535,
      "epoch": 0.46220427005193304,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986355489568272,
      "loss": 6.4817,
      "mean_token_accuracy": 0.15511930286884307,
      "num_tokens": 10163539.0,
      "step": 4005
    },
    {
      "entropy": 6.532063102722168,
      "epoch": 0.46278130409694174,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986310076782379,
      "loss": 6.4467,
      "mean_token_accuracy": 0.15533970296382904,
      "num_tokens": 10174760.0,
      "step": 4010
    },
    {
      "entropy": 6.498932600021362,
      "epoch": 0.4633583381419504,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986264588779307,
      "loss": 6.4642,
      "mean_token_accuracy": 0.15401545017957688,
      "num_tokens": 10186627.0,
      "step": 4015
    },
    {
      "entropy": 6.647648715972901,
      "epoch": 0.463935372186959,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0004986219025560586,
      "loss": 6.35,
      "mean_token_accuracy": 0.1703260600566864,
      "num_tokens": 10200084.0,
      "step": 4020
    },
    {
      "entropy": 6.472187328338623,
      "epoch": 0.4645124062319677,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498617338712775,
      "loss": 6.4214,
      "mean_token_accuracy": 0.16182194352149964,
      "num_tokens": 10212121.0,
      "step": 4025
    },
    {
      "entropy": 6.5356916904449465,
      "epoch": 0.46508944027697635,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004986127673482332,
      "loss": 6.4148,
      "mean_token_accuracy": 0.16143590807914734,
      "num_tokens": 10225138.0,
      "step": 4030
    },
    {
      "entropy": 6.6138585090637205,
      "epoch": 0.465666474321985,
      "grad_norm": 0.73046875,
      "learning_rate": 0.0004986081884625871,
      "loss": 6.4351,
      "mean_token_accuracy": 0.1563952624797821,
      "num_tokens": 10238309.0,
      "step": 4035
    },
    {
      "entropy": 6.528492736816406,
      "epoch": 0.46624350836699363,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986036020559906,
      "loss": 6.4045,
      "mean_token_accuracy": 0.15613240748643875,
      "num_tokens": 10251405.0,
      "step": 4040
    },
    {
      "entropy": 6.525904607772827,
      "epoch": 0.4668205424120023,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498599008128598,
      "loss": 6.3588,
      "mean_token_accuracy": 0.1662200689315796,
      "num_tokens": 10262870.0,
      "step": 4045
    },
    {
      "entropy": 6.4476752281188965,
      "epoch": 0.46739757645701097,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985944066805639,
      "loss": 6.4183,
      "mean_token_accuracy": 0.15107485949993132,
      "num_tokens": 10274283.0,
      "step": 4050
    },
    {
      "entropy": 6.554322624206543,
      "epoch": 0.4679746105020196,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498589797712043,
      "loss": 6.4333,
      "mean_token_accuracy": 0.15794023424386977,
      "num_tokens": 10287581.0,
      "step": 4055
    },
    {
      "entropy": 6.595016288757324,
      "epoch": 0.4685516445470283,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004985851812231903,
      "loss": 6.3567,
      "mean_token_accuracy": 0.16236607283353804,
      "num_tokens": 10299654.0,
      "step": 4060
    },
    {
      "entropy": 6.4879011631011965,
      "epoch": 0.46912867859203694,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985805572141611,
      "loss": 6.4447,
      "mean_token_accuracy": 0.15488010570406913,
      "num_tokens": 10312323.0,
      "step": 4065
    },
    {
      "entropy": 6.54975700378418,
      "epoch": 0.4697057126370456,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985759256851108,
      "loss": 6.2925,
      "mean_token_accuracy": 0.16682098954916,
      "num_tokens": 10325078.0,
      "step": 4070
    },
    {
      "entropy": 6.481452751159668,
      "epoch": 0.4702827466820542,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004985712866361954,
      "loss": 6.4389,
      "mean_token_accuracy": 0.15707961320877076,
      "num_tokens": 10339938.0,
      "step": 4075
    },
    {
      "entropy": 6.5055419445037845,
      "epoch": 0.4708597807270629,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985666400675709,
      "loss": 6.4518,
      "mean_token_accuracy": 0.1565823420882225,
      "num_tokens": 10352024.0,
      "step": 4080
    },
    {
      "entropy": 6.502655839920044,
      "epoch": 0.47143681477207156,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004985619859793934,
      "loss": 6.2503,
      "mean_token_accuracy": 0.17177620381116868,
      "num_tokens": 10365919.0,
      "step": 4085
    },
    {
      "entropy": 6.411081266403198,
      "epoch": 0.4720138488170802,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985573243718195,
      "loss": 6.3393,
      "mean_token_accuracy": 0.1644672930240631,
      "num_tokens": 10377488.0,
      "step": 4090
    },
    {
      "entropy": 6.475267219543457,
      "epoch": 0.4725908828620889,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004985526552450061,
      "loss": 6.3579,
      "mean_token_accuracy": 0.16348247230052948,
      "num_tokens": 10391187.0,
      "step": 4095
    },
    {
      "entropy": 6.440578937530518,
      "epoch": 0.47316791690709753,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985479785991103,
      "loss": 6.3374,
      "mean_token_accuracy": 0.16505655497312546,
      "num_tokens": 10403458.0,
      "step": 4100
    },
    {
      "entropy": 6.575636100769043,
      "epoch": 0.47374495095210617,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985432944342892,
      "loss": 6.5123,
      "mean_token_accuracy": 0.14661003053188323,
      "num_tokens": 10416533.0,
      "step": 4105
    },
    {
      "entropy": 6.5452169418334964,
      "epoch": 0.4743219849971148,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985386027507003,
      "loss": 6.3576,
      "mean_token_accuracy": 0.1599026843905449,
      "num_tokens": 10429593.0,
      "step": 4110
    },
    {
      "entropy": 6.452538537979126,
      "epoch": 0.4748990190421235,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985339035485018,
      "loss": 6.3891,
      "mean_token_accuracy": 0.16478680968284606,
      "num_tokens": 10440520.0,
      "step": 4115
    },
    {
      "entropy": 6.481285953521729,
      "epoch": 0.47547605308713214,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004985291968278513,
      "loss": 6.4028,
      "mean_token_accuracy": 0.1564889907836914,
      "num_tokens": 10453971.0,
      "step": 4120
    },
    {
      "entropy": 6.454246282577515,
      "epoch": 0.4760530871321408,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004985244825889074,
      "loss": 6.393,
      "mean_token_accuracy": 0.1587365560233593,
      "num_tokens": 10466033.0,
      "step": 4125
    },
    {
      "entropy": 6.5247338771820065,
      "epoch": 0.4766301211771495,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004985197608318284,
      "loss": 6.3609,
      "mean_token_accuracy": 0.1652022197842598,
      "num_tokens": 10478608.0,
      "step": 4130
    },
    {
      "entropy": 6.537705278396606,
      "epoch": 0.4772071552221581,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985150315567735,
      "loss": 6.4645,
      "mean_token_accuracy": 0.16325337141752244,
      "num_tokens": 10491223.0,
      "step": 4135
    },
    {
      "entropy": 6.584768152236938,
      "epoch": 0.47778418926716676,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004985102947639013,
      "loss": 6.506,
      "mean_token_accuracy": 0.15094189196825028,
      "num_tokens": 10504398.0,
      "step": 4140
    },
    {
      "entropy": 6.512157869338989,
      "epoch": 0.4783612233121754,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985055504533715,
      "loss": 6.3349,
      "mean_token_accuracy": 0.16145680993795394,
      "num_tokens": 10516998.0,
      "step": 4145
    },
    {
      "entropy": 6.511150979995728,
      "epoch": 0.4789382573571841,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985007986253435,
      "loss": 6.3307,
      "mean_token_accuracy": 0.15873684138059616,
      "num_tokens": 10529632.0,
      "step": 4150
    },
    {
      "entropy": 6.396101284027099,
      "epoch": 0.47951529140219273,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004984960392799772,
      "loss": 6.3296,
      "mean_token_accuracy": 0.16098663210868835,
      "num_tokens": 10542479.0,
      "step": 4155
    },
    {
      "entropy": 6.440076684951782,
      "epoch": 0.4800923254472014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004984912724174326,
      "loss": 6.2693,
      "mean_token_accuracy": 0.17203231304883956,
      "num_tokens": 10555322.0,
      "step": 4160
    },
    {
      "entropy": 6.5678126335144045,
      "epoch": 0.48066935949221,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004984864980378701,
      "loss": 6.4139,
      "mean_token_accuracy": 0.15598823130130768,
      "num_tokens": 10567757.0,
      "step": 4165
    },
    {
      "entropy": 6.547306394577026,
      "epoch": 0.4812463935372187,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004984817161414502,
      "loss": 6.4422,
      "mean_token_accuracy": 0.14962815642356872,
      "num_tokens": 10580820.0,
      "step": 4170
    },
    {
      "entropy": 6.538690710067749,
      "epoch": 0.48182342758222735,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004984769267283338,
      "loss": 6.4479,
      "mean_token_accuracy": 0.15779006332159043,
      "num_tokens": 10595690.0,
      "step": 4175
    },
    {
      "entropy": 6.494137716293335,
      "epoch": 0.482400461627236,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984721297986819,
      "loss": 6.346,
      "mean_token_accuracy": 0.15723106414079666,
      "num_tokens": 10608393.0,
      "step": 4180
    },
    {
      "entropy": 6.548528623580933,
      "epoch": 0.4829774956722447,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004984673253526561,
      "loss": 6.4367,
      "mean_token_accuracy": 0.15608918517827988,
      "num_tokens": 10621073.0,
      "step": 4185
    },
    {
      "entropy": 6.442586946487427,
      "epoch": 0.4835545297172533,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984625133904176,
      "loss": 6.3089,
      "mean_token_accuracy": 0.1682225674390793,
      "num_tokens": 10633945.0,
      "step": 4190
    },
    {
      "entropy": 6.481200122833252,
      "epoch": 0.48413156376226196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984576939121286,
      "loss": 6.4348,
      "mean_token_accuracy": 0.15343092679977416,
      "num_tokens": 10645727.0,
      "step": 4195
    },
    {
      "entropy": 6.521467208862305,
      "epoch": 0.4847085978072706,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984528669179511,
      "loss": 6.291,
      "mean_token_accuracy": 0.1592160165309906,
      "num_tokens": 10658488.0,
      "step": 4200
    },
    {
      "entropy": 6.476761770248413,
      "epoch": 0.4852856318522793,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984480324080472,
      "loss": 6.3897,
      "mean_token_accuracy": 0.15781337171792983,
      "num_tokens": 10671361.0,
      "step": 4205
    },
    {
      "entropy": 6.494213676452636,
      "epoch": 0.48586266589728794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049844319038258,
      "loss": 6.3209,
      "mean_token_accuracy": 0.1734808161854744,
      "num_tokens": 10684825.0,
      "step": 4210
    },
    {
      "entropy": 6.475206899642944,
      "epoch": 0.4864396999422966,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984383408417119,
      "loss": 6.4307,
      "mean_token_accuracy": 0.1594849243760109,
      "num_tokens": 10698627.0,
      "step": 4215
    },
    {
      "entropy": 6.523198556900025,
      "epoch": 0.4870167339873053,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984334837856064,
      "loss": 6.3561,
      "mean_token_accuracy": 0.15792025476694108,
      "num_tokens": 10712124.0,
      "step": 4220
    },
    {
      "entropy": 6.516054153442383,
      "epoch": 0.4875937680323139,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984286192144263,
      "loss": 6.393,
      "mean_token_accuracy": 0.15413977801799775,
      "num_tokens": 10723939.0,
      "step": 4225
    },
    {
      "entropy": 6.508268594741821,
      "epoch": 0.48817080207732255,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984237471283359,
      "loss": 6.3928,
      "mean_token_accuracy": 0.1572265714406967,
      "num_tokens": 10736795.0,
      "step": 4230
    },
    {
      "entropy": 6.4552051544189455,
      "epoch": 0.4887478361223312,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984188675274986,
      "loss": 6.2919,
      "mean_token_accuracy": 0.16994941532611846,
      "num_tokens": 10749905.0,
      "step": 4235
    },
    {
      "entropy": 6.466286897659302,
      "epoch": 0.4893248701673399,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004984139804120787,
      "loss": 6.4002,
      "mean_token_accuracy": 0.15417940616607667,
      "num_tokens": 10761852.0,
      "step": 4240
    },
    {
      "entropy": 6.501400518417358,
      "epoch": 0.4899019042123485,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984090857822406,
      "loss": 6.3811,
      "mean_token_accuracy": 0.16262035071849823,
      "num_tokens": 10774673.0,
      "step": 4245
    },
    {
      "entropy": 6.462345552444458,
      "epoch": 0.49047893825735717,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984041836381488,
      "loss": 6.3484,
      "mean_token_accuracy": 0.1620374873280525,
      "num_tokens": 10786892.0,
      "step": 4250
    },
    {
      "entropy": 6.481334590911866,
      "epoch": 0.49105597230236586,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983992739799682,
      "loss": 6.3739,
      "mean_token_accuracy": 0.1637142077088356,
      "num_tokens": 10798846.0,
      "step": 4255
    },
    {
      "entropy": 6.4820136547088625,
      "epoch": 0.4916330063473745,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004983943568078642,
      "loss": 6.2917,
      "mean_token_accuracy": 0.16547314524650575,
      "num_tokens": 10811893.0,
      "step": 4260
    },
    {
      "entropy": 6.472692060470581,
      "epoch": 0.49221004039238314,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983894321220017,
      "loss": 6.3439,
      "mean_token_accuracy": 0.16635758727788924,
      "num_tokens": 10825595.0,
      "step": 4265
    },
    {
      "entropy": 6.431785249710083,
      "epoch": 0.4927870744373918,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983844999225468,
      "loss": 6.3143,
      "mean_token_accuracy": 0.17111872732639313,
      "num_tokens": 10837546.0,
      "step": 4270
    },
    {
      "entropy": 6.51116623878479,
      "epoch": 0.4933641084824005,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004983795602096651,
      "loss": 6.4218,
      "mean_token_accuracy": 0.15393567085266113,
      "num_tokens": 10851234.0,
      "step": 4275
    },
    {
      "entropy": 6.4908205509185795,
      "epoch": 0.4939411425274091,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498374612983523,
      "loss": 6.375,
      "mean_token_accuracy": 0.1528395116329193,
      "num_tokens": 10864397.0,
      "step": 4280
    },
    {
      "entropy": 6.514799118041992,
      "epoch": 0.49451817657241776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983696582442866,
      "loss": 6.3765,
      "mean_token_accuracy": 0.15875952392816545,
      "num_tokens": 10877230.0,
      "step": 4285
    },
    {
      "entropy": 6.470470523834228,
      "epoch": 0.49509521061742645,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004983646959921227,
      "loss": 6.3415,
      "mean_token_accuracy": 0.1599888324737549,
      "num_tokens": 10890847.0,
      "step": 4290
    },
    {
      "entropy": 6.440327835083008,
      "epoch": 0.4956722446624351,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004983597262271984,
      "loss": 6.3097,
      "mean_token_accuracy": 0.16640594899654387,
      "num_tokens": 10903488.0,
      "step": 4295
    },
    {
      "entropy": 6.5825787544250485,
      "epoch": 0.49624927870744373,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983547489496804,
      "loss": 6.3807,
      "mean_token_accuracy": 0.15209589451551436,
      "num_tokens": 10914744.0,
      "step": 4300
    },
    {
      "entropy": 6.49207763671875,
      "epoch": 0.49682631275245237,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004983497641597365,
      "loss": 6.3837,
      "mean_token_accuracy": 0.15209848284721375,
      "num_tokens": 10928307.0,
      "step": 4305
    },
    {
      "entropy": 6.510096883773803,
      "epoch": 0.49740334679746107,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004983447718575342,
      "loss": 6.4199,
      "mean_token_accuracy": 0.15832170844078064,
      "num_tokens": 10939876.0,
      "step": 4310
    },
    {
      "entropy": 6.479329204559326,
      "epoch": 0.4979803808424697,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004983397720432414,
      "loss": 6.3041,
      "mean_token_accuracy": 0.16492047309875488,
      "num_tokens": 10952123.0,
      "step": 4315
    },
    {
      "entropy": 6.4766120433807375,
      "epoch": 0.49855741488747835,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004983347647170264,
      "loss": 6.3814,
      "mean_token_accuracy": 0.15579510778188704,
      "num_tokens": 10967374.0,
      "step": 4320
    },
    {
      "entropy": 6.49477186203003,
      "epoch": 0.49913444893248704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983297498790574,
      "loss": 6.4387,
      "mean_token_accuracy": 0.15156637281179428,
      "num_tokens": 10980761.0,
      "step": 4325
    },
    {
      "entropy": 6.530684614181519,
      "epoch": 0.4997114829774957,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983247275295034,
      "loss": 6.3672,
      "mean_token_accuracy": 0.15633219629526138,
      "num_tokens": 10994538.0,
      "step": 4330
    },
    {
      "entropy": 6.51979341506958,
      "epoch": 0.5002885170225043,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004983196976685329,
      "loss": 6.4175,
      "mean_token_accuracy": 0.1540020525455475,
      "num_tokens": 11006956.0,
      "step": 4335
    },
    {
      "entropy": 6.479969930648804,
      "epoch": 0.500865551067513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983146602963155,
      "loss": 6.3815,
      "mean_token_accuracy": 0.16485208570957183,
      "num_tokens": 11019202.0,
      "step": 4340
    },
    {
      "entropy": 6.568608665466309,
      "epoch": 0.5014425851125216,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983096154130203,
      "loss": 6.3646,
      "mean_token_accuracy": 0.16060912609100342,
      "num_tokens": 11031560.0,
      "step": 4345
    },
    {
      "entropy": 6.526084327697754,
      "epoch": 0.5020196191575302,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983045630188171,
      "loss": 6.4622,
      "mean_token_accuracy": 0.15379957109689713,
      "num_tokens": 11044537.0,
      "step": 4350
    },
    {
      "entropy": 6.451754522323609,
      "epoch": 0.502596653202539,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982995031138759,
      "loss": 6.3001,
      "mean_token_accuracy": 0.1613658294081688,
      "num_tokens": 11056642.0,
      "step": 4355
    },
    {
      "entropy": 6.453772687911988,
      "epoch": 0.5031736872475476,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004982944356983666,
      "loss": 6.2954,
      "mean_token_accuracy": 0.16220536679029465,
      "num_tokens": 11068642.0,
      "step": 4360
    },
    {
      "entropy": 6.558634853363037,
      "epoch": 0.5037507212925563,
      "grad_norm": 3.21875,
      "learning_rate": 0.0004982893607724602,
      "loss": 6.5895,
      "mean_token_accuracy": 0.15590605288743972,
      "num_tokens": 11082056.0,
      "step": 4365
    },
    {
      "entropy": 6.497712707519531,
      "epoch": 0.5043277553375649,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000498284278336327,
      "loss": 6.3903,
      "mean_token_accuracy": 0.15248484015464783,
      "num_tokens": 11093722.0,
      "step": 4370
    },
    {
      "entropy": 6.525355005264283,
      "epoch": 0.5049047893825735,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982791883901379,
      "loss": 6.3873,
      "mean_token_accuracy": 0.1526179663836956,
      "num_tokens": 11105949.0,
      "step": 4375
    },
    {
      "entropy": 6.4398966312408445,
      "epoch": 0.5054818234275822,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982740909340643,
      "loss": 6.2806,
      "mean_token_accuracy": 0.16731281131505965,
      "num_tokens": 11118074.0,
      "step": 4380
    },
    {
      "entropy": 6.574570846557617,
      "epoch": 0.5060588574725908,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982689859682775,
      "loss": 6.3591,
      "mean_token_accuracy": 0.15872727185487748,
      "num_tokens": 11130917.0,
      "step": 4385
    },
    {
      "entropy": 6.454862403869629,
      "epoch": 0.5066358915175996,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004982638734929492,
      "loss": 6.3971,
      "mean_token_accuracy": 0.16015870422124862,
      "num_tokens": 11142829.0,
      "step": 4390
    },
    {
      "entropy": 6.4587291240692135,
      "epoch": 0.5072129255626082,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982587535082515,
      "loss": 6.356,
      "mean_token_accuracy": 0.16462198942899703,
      "num_tokens": 11154606.0,
      "step": 4395
    },
    {
      "entropy": 6.567020273208618,
      "epoch": 0.5077899596076169,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982536260143565,
      "loss": 6.3357,
      "mean_token_accuracy": 0.15825513154268264,
      "num_tokens": 11167160.0,
      "step": 4400
    },
    {
      "entropy": 6.403166437149048,
      "epoch": 0.5083669936526255,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004982484910114367,
      "loss": 6.3295,
      "mean_token_accuracy": 0.16778065413236617,
      "num_tokens": 11178813.0,
      "step": 4405
    },
    {
      "entropy": 6.381605005264282,
      "epoch": 0.5089440276976341,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982433484996648,
      "loss": 6.2923,
      "mean_token_accuracy": 0.162590828537941,
      "num_tokens": 11192648.0,
      "step": 4410
    },
    {
      "entropy": 6.628140115737915,
      "epoch": 0.5095210617426428,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004982381984792138,
      "loss": 6.393,
      "mean_token_accuracy": 0.15693947672843933,
      "num_tokens": 11205519.0,
      "step": 4415
    },
    {
      "entropy": 6.461419773101807,
      "epoch": 0.5100980957876514,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004982330409502568,
      "loss": 6.3921,
      "mean_token_accuracy": 0.15413278788328172,
      "num_tokens": 11218362.0,
      "step": 4420
    },
    {
      "entropy": 6.442616987228393,
      "epoch": 0.5106751298326602,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982278759129674,
      "loss": 6.3803,
      "mean_token_accuracy": 0.15451293140649797,
      "num_tokens": 11231821.0,
      "step": 4425
    },
    {
      "entropy": 6.557499504089355,
      "epoch": 0.5112521638776688,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004982227033675194,
      "loss": 6.422,
      "mean_token_accuracy": 0.15312207788228988,
      "num_tokens": 11246300.0,
      "step": 4430
    },
    {
      "entropy": 6.465668153762818,
      "epoch": 0.5118291979226774,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982175233140865,
      "loss": 6.3175,
      "mean_token_accuracy": 0.1643589124083519,
      "num_tokens": 11258561.0,
      "step": 4435
    },
    {
      "entropy": 6.531848382949829,
      "epoch": 0.5124062319676861,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982123357528431,
      "loss": 6.3149,
      "mean_token_accuracy": 0.1639846608042717,
      "num_tokens": 11271719.0,
      "step": 4440
    },
    {
      "entropy": 6.302284049987793,
      "epoch": 0.5129832660126947,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982071406839636,
      "loss": 6.3153,
      "mean_token_accuracy": 0.16485968083143235,
      "num_tokens": 11284166.0,
      "step": 4445
    },
    {
      "entropy": 6.476738452911377,
      "epoch": 0.5135603000577034,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004982019381076229,
      "loss": 6.365,
      "mean_token_accuracy": 0.15987009257078172,
      "num_tokens": 11296140.0,
      "step": 4450
    },
    {
      "entropy": 6.515838384628296,
      "epoch": 0.514137334102712,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004981967280239958,
      "loss": 6.392,
      "mean_token_accuracy": 0.1580490365624428,
      "num_tokens": 11308680.0,
      "step": 4455
    },
    {
      "entropy": 6.520589065551758,
      "epoch": 0.5147143681477208,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981915104332578,
      "loss": 6.3734,
      "mean_token_accuracy": 0.15471306443214417,
      "num_tokens": 11320887.0,
      "step": 4460
    },
    {
      "entropy": 6.518787240982055,
      "epoch": 0.5152914021927294,
      "grad_norm": 0.828125,
      "learning_rate": 0.000498186285335584,
      "loss": 6.3706,
      "mean_token_accuracy": 0.15980480164289473,
      "num_tokens": 11333873.0,
      "step": 4465
    },
    {
      "entropy": 6.5042322158813475,
      "epoch": 0.515868436237738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981810527311505,
      "loss": 6.4526,
      "mean_token_accuracy": 0.15650095790624619,
      "num_tokens": 11345526.0,
      "step": 4470
    },
    {
      "entropy": 6.537562704086303,
      "epoch": 0.5164454702827467,
      "grad_norm": 0.80078125,
      "learning_rate": 0.000498175812620133,
      "loss": 6.3399,
      "mean_token_accuracy": 0.15564233660697938,
      "num_tokens": 11356840.0,
      "step": 4475
    },
    {
      "entropy": 6.456150674819947,
      "epoch": 0.5170225043277553,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981705650027081,
      "loss": 6.4184,
      "mean_token_accuracy": 0.15510667115449905,
      "num_tokens": 11369146.0,
      "step": 4480
    },
    {
      "entropy": 6.471530199050903,
      "epoch": 0.517599538372764,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498165309879052,
      "loss": 6.351,
      "mean_token_accuracy": 0.16200470179319382,
      "num_tokens": 11382399.0,
      "step": 4485
    },
    {
      "entropy": 6.437090301513672,
      "epoch": 0.5181765724177726,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004981600472493415,
      "loss": 6.2501,
      "mean_token_accuracy": 0.16678088307380676,
      "num_tokens": 11395602.0,
      "step": 4490
    },
    {
      "entropy": 6.418407583236695,
      "epoch": 0.5187536064627813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981547771137537,
      "loss": 6.3489,
      "mean_token_accuracy": 0.16000750213861464,
      "num_tokens": 11408530.0,
      "step": 4495
    },
    {
      "entropy": 6.519910526275635,
      "epoch": 0.51933064050779,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981494994724658,
      "loss": 6.3563,
      "mean_token_accuracy": 0.16234227418899536,
      "num_tokens": 11420150.0,
      "step": 4500
    },
    {
      "entropy": 6.446649980545044,
      "epoch": 0.5199076745527986,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981442143256554,
      "loss": 6.3945,
      "mean_token_accuracy": 0.15244318544864655,
      "num_tokens": 11432318.0,
      "step": 4505
    },
    {
      "entropy": 6.507375001907349,
      "epoch": 0.5204847085978073,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004981389216735001,
      "loss": 6.4357,
      "mean_token_accuracy": 0.1586089849472046,
      "num_tokens": 11444143.0,
      "step": 4510
    },
    {
      "entropy": 6.438950109481811,
      "epoch": 0.5210617426428159,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498133621516178,
      "loss": 6.33,
      "mean_token_accuracy": 0.16003450751304626,
      "num_tokens": 11455997.0,
      "step": 4515
    },
    {
      "entropy": 6.510848474502564,
      "epoch": 0.5216387766878245,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981283138538675,
      "loss": 6.4088,
      "mean_token_accuracy": 0.15459372103214264,
      "num_tokens": 11468271.0,
      "step": 4520
    },
    {
      "entropy": 6.529215717315674,
      "epoch": 0.5222158107328332,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498122998686747,
      "loss": 6.3398,
      "mean_token_accuracy": 0.15672484934329986,
      "num_tokens": 11479729.0,
      "step": 4525
    },
    {
      "entropy": 6.391436529159546,
      "epoch": 0.5227928447778419,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981176760149951,
      "loss": 6.2739,
      "mean_token_accuracy": 0.1627101868391037,
      "num_tokens": 11492261.0,
      "step": 4530
    },
    {
      "entropy": 6.450825357437134,
      "epoch": 0.5233698788228506,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004981123458387911,
      "loss": 6.3,
      "mean_token_accuracy": 0.1585073545575142,
      "num_tokens": 11504802.0,
      "step": 4535
    },
    {
      "entropy": 6.528818082809448,
      "epoch": 0.5239469128678592,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981070081583142,
      "loss": 6.3564,
      "mean_token_accuracy": 0.1574664145708084,
      "num_tokens": 11518105.0,
      "step": 4540
    },
    {
      "entropy": 6.427476978302002,
      "epoch": 0.5245239469128679,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000498101662973744,
      "loss": 6.3377,
      "mean_token_accuracy": 0.15979625582695006,
      "num_tokens": 11532085.0,
      "step": 4545
    },
    {
      "entropy": 6.439642763137817,
      "epoch": 0.5251009809578765,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00049809631028526,
      "loss": 6.2873,
      "mean_token_accuracy": 0.16680559366941453,
      "num_tokens": 11544328.0,
      "step": 4550
    },
    {
      "entropy": 6.438679456710815,
      "epoch": 0.5256780150028851,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980909500930424,
      "loss": 6.3784,
      "mean_token_accuracy": 0.1642422378063202,
      "num_tokens": 11556496.0,
      "step": 4555
    },
    {
      "entropy": 6.560665273666382,
      "epoch": 0.5262550490478938,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980855823972717,
      "loss": 6.4115,
      "mean_token_accuracy": 0.1573752909898758,
      "num_tokens": 11569330.0,
      "step": 4560
    },
    {
      "entropy": 6.473380136489868,
      "epoch": 0.5268320830929025,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498080207198128,
      "loss": 6.3165,
      "mean_token_accuracy": 0.16247887313365936,
      "num_tokens": 11581335.0,
      "step": 4565
    },
    {
      "entropy": 6.4558507919311525,
      "epoch": 0.5274091171379112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980748244957925,
      "loss": 6.2877,
      "mean_token_accuracy": 0.16184458136558533,
      "num_tokens": 11593130.0,
      "step": 4570
    },
    {
      "entropy": 6.335966634750366,
      "epoch": 0.5279861511829198,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004980694342904461,
      "loss": 6.3012,
      "mean_token_accuracy": 0.16489815562963486,
      "num_tokens": 11605140.0,
      "step": 4575
    },
    {
      "entropy": 6.467780017852784,
      "epoch": 0.5285631852279284,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004980640365822701,
      "loss": 6.3236,
      "mean_token_accuracy": 0.15909543633460999,
      "num_tokens": 11617361.0,
      "step": 4580
    },
    {
      "entropy": 6.439427661895752,
      "epoch": 0.5291402192729371,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498058631371446,
      "loss": 6.3211,
      "mean_token_accuracy": 0.16430423706769942,
      "num_tokens": 11630780.0,
      "step": 4585
    },
    {
      "entropy": 6.3723077297210695,
      "epoch": 0.5297172533179457,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980532186581555,
      "loss": 6.2942,
      "mean_token_accuracy": 0.16157945692539216,
      "num_tokens": 11643120.0,
      "step": 4590
    },
    {
      "entropy": 6.509933567047119,
      "epoch": 0.5302942873629544,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498047798442581,
      "loss": 6.3214,
      "mean_token_accuracy": 0.16044917702674866,
      "num_tokens": 11655849.0,
      "step": 4595
    },
    {
      "entropy": 6.421504974365234,
      "epoch": 0.5308713214079631,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004980423707249044,
      "loss": 6.324,
      "mean_token_accuracy": 0.15950872004032135,
      "num_tokens": 11669072.0,
      "step": 4600
    },
    {
      "entropy": 6.4032351016998295,
      "epoch": 0.5314483554529718,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004980369355053086,
      "loss": 6.265,
      "mean_token_accuracy": 0.16826080530881882,
      "num_tokens": 11680253.0,
      "step": 4605
    },
    {
      "entropy": 6.482820701599121,
      "epoch": 0.5320253894979804,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980314927839763,
      "loss": 6.2559,
      "mean_token_accuracy": 0.15586785078048707,
      "num_tokens": 11691465.0,
      "step": 4610
    },
    {
      "entropy": 6.44303879737854,
      "epoch": 0.532602423542989,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980260425610903,
      "loss": 6.394,
      "mean_token_accuracy": 0.16090328097343445,
      "num_tokens": 11703679.0,
      "step": 4615
    },
    {
      "entropy": 6.469963407516479,
      "epoch": 0.5331794575879977,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004980205848368343,
      "loss": 6.3535,
      "mean_token_accuracy": 0.1571087807416916,
      "num_tokens": 11716204.0,
      "step": 4620
    },
    {
      "entropy": 6.517374849319458,
      "epoch": 0.5337564916330063,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980151196113917,
      "loss": 6.339,
      "mean_token_accuracy": 0.16261640936136246,
      "num_tokens": 11728404.0,
      "step": 4625
    },
    {
      "entropy": 6.387979793548584,
      "epoch": 0.534333525678015,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004980096468849464,
      "loss": 6.2656,
      "mean_token_accuracy": 0.160543355345726,
      "num_tokens": 11740170.0,
      "step": 4630
    },
    {
      "entropy": 6.368899822235107,
      "epoch": 0.5349105597230237,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004980041666576823,
      "loss": 6.2578,
      "mean_token_accuracy": 0.16634061187505722,
      "num_tokens": 11753062.0,
      "step": 4635
    },
    {
      "entropy": 6.433922815322876,
      "epoch": 0.5354875937680323,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004979986789297837,
      "loss": 6.2715,
      "mean_token_accuracy": 0.16562960147857667,
      "num_tokens": 11765588.0,
      "step": 4640
    },
    {
      "entropy": 6.465123987197876,
      "epoch": 0.536064627813041,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979931837014355,
      "loss": 6.3848,
      "mean_token_accuracy": 0.16014475971460343,
      "num_tokens": 11777358.0,
      "step": 4645
    },
    {
      "entropy": 6.529675102233886,
      "epoch": 0.5366416618580496,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979876809728223,
      "loss": 6.4021,
      "mean_token_accuracy": 0.1560329467058182,
      "num_tokens": 11791572.0,
      "step": 4650
    },
    {
      "entropy": 6.474690055847168,
      "epoch": 0.5372186959030583,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979821707441292,
      "loss": 6.439,
      "mean_token_accuracy": 0.1557312175631523,
      "num_tokens": 11805198.0,
      "step": 4655
    },
    {
      "entropy": 6.5011818408966064,
      "epoch": 0.5377957299480669,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979766530155416,
      "loss": 6.3507,
      "mean_token_accuracy": 0.15437426120042802,
      "num_tokens": 11817882.0,
      "step": 4660
    },
    {
      "entropy": 6.450196981430054,
      "epoch": 0.5383727639930755,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979711277872449,
      "loss": 6.254,
      "mean_token_accuracy": 0.1714956507086754,
      "num_tokens": 11830334.0,
      "step": 4665
    },
    {
      "entropy": 6.36470775604248,
      "epoch": 0.5389497980380843,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004979655950594252,
      "loss": 6.3324,
      "mean_token_accuracy": 0.16521675139665604,
      "num_tokens": 11843057.0,
      "step": 4670
    },
    {
      "entropy": 6.521092224121094,
      "epoch": 0.5395268320830929,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004979600548322684,
      "loss": 6.4443,
      "mean_token_accuracy": 0.1515732452273369,
      "num_tokens": 11856086.0,
      "step": 4675
    },
    {
      "entropy": 6.502345943450928,
      "epoch": 0.5401038661281016,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000497954507105961,
      "loss": 6.3598,
      "mean_token_accuracy": 0.1554923489689827,
      "num_tokens": 11867799.0,
      "step": 4680
    },
    {
      "entropy": 6.406812906265259,
      "epoch": 0.5406809001731102,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979489518806893,
      "loss": 6.2653,
      "mean_token_accuracy": 0.169070328772068,
      "num_tokens": 11881488.0,
      "step": 4685
    },
    {
      "entropy": 6.489722061157226,
      "epoch": 0.5412579342181189,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979433891566405,
      "loss": 6.3691,
      "mean_token_accuracy": 0.15394651889801025,
      "num_tokens": 11893495.0,
      "step": 4690
    },
    {
      "entropy": 6.486820793151855,
      "epoch": 0.5418349682631275,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979378189340015,
      "loss": 6.3088,
      "mean_token_accuracy": 0.16032520085573196,
      "num_tokens": 11905456.0,
      "step": 4695
    },
    {
      "entropy": 6.447782754898071,
      "epoch": 0.5424120023081361,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004979322412129597,
      "loss": 6.3555,
      "mean_token_accuracy": 0.15736544877290726,
      "num_tokens": 11917987.0,
      "step": 4700
    },
    {
      "entropy": 6.472414350509643,
      "epoch": 0.5429890363531449,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979266559937028,
      "loss": 6.3375,
      "mean_token_accuracy": 0.16168057471513747,
      "num_tokens": 11930104.0,
      "step": 4705
    },
    {
      "entropy": 6.486454343795776,
      "epoch": 0.5435660703981535,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979210632764185,
      "loss": 6.3531,
      "mean_token_accuracy": 0.15208696722984313,
      "num_tokens": 11942841.0,
      "step": 4710
    },
    {
      "entropy": 6.467970561981201,
      "epoch": 0.5441431044431622,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004979154630612949,
      "loss": 6.3429,
      "mean_token_accuracy": 0.15887532234191895,
      "num_tokens": 11955509.0,
      "step": 4715
    },
    {
      "entropy": 6.456158542633057,
      "epoch": 0.5447201384881708,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004979098553485205,
      "loss": 6.3279,
      "mean_token_accuracy": 0.16368138641119004,
      "num_tokens": 11967389.0,
      "step": 4720
    },
    {
      "entropy": 6.5094503402709964,
      "epoch": 0.5452971725331794,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004979042401382838,
      "loss": 6.3774,
      "mean_token_accuracy": 0.1637259915471077,
      "num_tokens": 11980173.0,
      "step": 4725
    },
    {
      "entropy": 6.483466386795044,
      "epoch": 0.5458742065781881,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004978986174307737,
      "loss": 6.3284,
      "mean_token_accuracy": 0.16224620044231414,
      "num_tokens": 11993039.0,
      "step": 4730
    },
    {
      "entropy": 6.408599996566773,
      "epoch": 0.5464512406231967,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004978929872261794,
      "loss": 6.2337,
      "mean_token_accuracy": 0.16886009871959687,
      "num_tokens": 12005861.0,
      "step": 4735
    },
    {
      "entropy": 6.323904705047608,
      "epoch": 0.5470282746682055,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004978873495246901,
      "loss": 6.1946,
      "mean_token_accuracy": 0.1696503296494484,
      "num_tokens": 12018137.0,
      "step": 4740
    },
    {
      "entropy": 6.472019958496094,
      "epoch": 0.5476053087132141,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978817043264955,
      "loss": 6.3375,
      "mean_token_accuracy": 0.1588849812746048,
      "num_tokens": 12030086.0,
      "step": 4745
    },
    {
      "entropy": 6.40808687210083,
      "epoch": 0.5481823427582228,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004978760516317856,
      "loss": 6.3081,
      "mean_token_accuracy": 0.1678110256791115,
      "num_tokens": 12043166.0,
      "step": 4750
    },
    {
      "entropy": 6.395149755477905,
      "epoch": 0.5487593768032314,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004978703914407503,
      "loss": 6.264,
      "mean_token_accuracy": 0.16222834289073945,
      "num_tokens": 12055373.0,
      "step": 4755
    },
    {
      "entropy": 6.39506983757019,
      "epoch": 0.54933641084824,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978647237535802,
      "loss": 6.2642,
      "mean_token_accuracy": 0.16176026910543442,
      "num_tokens": 12066369.0,
      "step": 4760
    },
    {
      "entropy": 6.374369716644287,
      "epoch": 0.5499134448932487,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004978590485704657,
      "loss": 6.1943,
      "mean_token_accuracy": 0.16742823272943497,
      "num_tokens": 12077939.0,
      "step": 4765
    },
    {
      "entropy": 6.513216352462768,
      "epoch": 0.5504904789382573,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978533658915979,
      "loss": 6.3859,
      "mean_token_accuracy": 0.1533949315547943,
      "num_tokens": 12091102.0,
      "step": 4770
    },
    {
      "entropy": 6.490841245651245,
      "epoch": 0.5510675129832661,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004978476757171678,
      "loss": 6.3215,
      "mean_token_accuracy": 0.16209329962730407,
      "num_tokens": 12103144.0,
      "step": 4775
    },
    {
      "entropy": 6.436073350906372,
      "epoch": 0.5516445470282747,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978419780473668,
      "loss": 6.3979,
      "mean_token_accuracy": 0.16156178414821626,
      "num_tokens": 12116664.0,
      "step": 4780
    },
    {
      "entropy": 6.447421932220459,
      "epoch": 0.5522215810732833,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978362728823865,
      "loss": 6.3087,
      "mean_token_accuracy": 0.16004290282726288,
      "num_tokens": 12129656.0,
      "step": 4785
    },
    {
      "entropy": 6.387583160400391,
      "epoch": 0.552798615118292,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978305602224189,
      "loss": 6.2694,
      "mean_token_accuracy": 0.16280092149972916,
      "num_tokens": 12143221.0,
      "step": 4790
    },
    {
      "entropy": 6.481845951080322,
      "epoch": 0.5533756491633006,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004978248400676562,
      "loss": 6.2451,
      "mean_token_accuracy": 0.16497370451688767,
      "num_tokens": 12156409.0,
      "step": 4795
    },
    {
      "entropy": 6.324881172180175,
      "epoch": 0.5539526832083093,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004978191124182905,
      "loss": 6.1727,
      "mean_token_accuracy": 0.17471756488084794,
      "num_tokens": 12169589.0,
      "step": 4800
    },
    {
      "entropy": 6.380939769744873,
      "epoch": 0.5545297172533179,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004978133772745149,
      "loss": 6.3131,
      "mean_token_accuracy": 0.1579168915748596,
      "num_tokens": 12182692.0,
      "step": 4805
    },
    {
      "entropy": 6.419868421554566,
      "epoch": 0.5551067512983267,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004978076346365218,
      "loss": 6.2702,
      "mean_token_accuracy": 0.16413773596286774,
      "num_tokens": 12195892.0,
      "step": 4810
    },
    {
      "entropy": 6.485994005203247,
      "epoch": 0.5556837853433353,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004978018845045047,
      "loss": 6.2949,
      "mean_token_accuracy": 0.15944212526082993,
      "num_tokens": 12208974.0,
      "step": 4815
    },
    {
      "entropy": 6.459294605255127,
      "epoch": 0.5562608193883439,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977961268786568,
      "loss": 6.361,
      "mean_token_accuracy": 0.1640543147921562,
      "num_tokens": 12221342.0,
      "step": 4820
    },
    {
      "entropy": 6.406201362609863,
      "epoch": 0.5568378534333526,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977903617591719,
      "loss": 6.3377,
      "mean_token_accuracy": 0.16159975826740264,
      "num_tokens": 12233778.0,
      "step": 4825
    },
    {
      "entropy": 6.394937467575073,
      "epoch": 0.5574148874783612,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977845891462439,
      "loss": 6.1738,
      "mean_token_accuracy": 0.16942807286977768,
      "num_tokens": 12246001.0,
      "step": 4830
    },
    {
      "entropy": 6.368630838394165,
      "epoch": 0.5579919215233698,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977788090400668,
      "loss": 6.2319,
      "mean_token_accuracy": 0.16157334297895432,
      "num_tokens": 12258677.0,
      "step": 4835
    },
    {
      "entropy": 6.527727890014648,
      "epoch": 0.5585689555683785,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977730214408352,
      "loss": 6.3329,
      "mean_token_accuracy": 0.15864758640527726,
      "num_tokens": 12270786.0,
      "step": 4840
    },
    {
      "entropy": 6.39428677558899,
      "epoch": 0.5591459896133872,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004977672263487435,
      "loss": 6.3389,
      "mean_token_accuracy": 0.1648403987288475,
      "num_tokens": 12284439.0,
      "step": 4845
    },
    {
      "entropy": 6.443121910095215,
      "epoch": 0.5597230236583959,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000497761423763987,
      "loss": 6.3145,
      "mean_token_accuracy": 0.15776993781328202,
      "num_tokens": 12297005.0,
      "step": 4850
    },
    {
      "entropy": 6.48302993774414,
      "epoch": 0.5603000577034045,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977556136867606,
      "loss": 6.3379,
      "mean_token_accuracy": 0.15854468047618867,
      "num_tokens": 12308855.0,
      "step": 4855
    },
    {
      "entropy": 6.496091365814209,
      "epoch": 0.5608770917484132,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004977497961172598,
      "loss": 6.2969,
      "mean_token_accuracy": 0.16524532288312913,
      "num_tokens": 12321120.0,
      "step": 4860
    },
    {
      "entropy": 6.426670408248901,
      "epoch": 0.5614541257934218,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977439710556802,
      "loss": 6.3581,
      "mean_token_accuracy": 0.15703101307153702,
      "num_tokens": 12333793.0,
      "step": 4865
    },
    {
      "entropy": 6.395258378982544,
      "epoch": 0.5620311598384304,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004977381385022179,
      "loss": 6.2346,
      "mean_token_accuracy": 0.16402624398469925,
      "num_tokens": 12345998.0,
      "step": 4870
    },
    {
      "entropy": 6.420551776885986,
      "epoch": 0.5626081938834391,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004977322984570688,
      "loss": 6.2915,
      "mean_token_accuracy": 0.1633617177605629,
      "num_tokens": 12358580.0,
      "step": 4875
    },
    {
      "entropy": 6.50617299079895,
      "epoch": 0.5631852279284478,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004977264509204296,
      "loss": 6.3728,
      "mean_token_accuracy": 0.15802410542964934,
      "num_tokens": 12370867.0,
      "step": 4880
    },
    {
      "entropy": 6.460091161727905,
      "epoch": 0.5637622619734565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977205958924967,
      "loss": 6.335,
      "mean_token_accuracy": 0.16238923370838165,
      "num_tokens": 12384086.0,
      "step": 4885
    },
    {
      "entropy": 6.331658267974854,
      "epoch": 0.5643392960184651,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977147333734673,
      "loss": 6.1915,
      "mean_token_accuracy": 0.1692612126469612,
      "num_tokens": 12396852.0,
      "step": 4890
    },
    {
      "entropy": 6.493855905532837,
      "epoch": 0.5649163300634737,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004977088633635385,
      "loss": 6.2907,
      "mean_token_accuracy": 0.16344485133886338,
      "num_tokens": 12409924.0,
      "step": 4895
    },
    {
      "entropy": 6.301746273040772,
      "epoch": 0.5654933641084824,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977029858629076,
      "loss": 6.2559,
      "mean_token_accuracy": 0.17010487616062164,
      "num_tokens": 12422218.0,
      "step": 4900
    },
    {
      "entropy": 6.406967258453369,
      "epoch": 0.566070398153491,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004976971008717726,
      "loss": 6.2744,
      "mean_token_accuracy": 0.15997430682182312,
      "num_tokens": 12433522.0,
      "step": 4905
    },
    {
      "entropy": 6.589291667938232,
      "epoch": 0.5666474321984997,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976912083903309,
      "loss": 6.3465,
      "mean_token_accuracy": 0.15723779648542405,
      "num_tokens": 12445526.0,
      "step": 4910
    },
    {
      "entropy": 6.381515216827393,
      "epoch": 0.5672244662435084,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004976853084187814,
      "loss": 6.3013,
      "mean_token_accuracy": 0.16589126586914063,
      "num_tokens": 12458403.0,
      "step": 4915
    },
    {
      "entropy": 6.522457313537598,
      "epoch": 0.5678015002885171,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976794009573219,
      "loss": 6.3572,
      "mean_token_accuracy": 0.16034309267997743,
      "num_tokens": 12470437.0,
      "step": 4920
    },
    {
      "entropy": 6.491942453384399,
      "epoch": 0.5683785343335257,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004976734860061515,
      "loss": 6.2934,
      "mean_token_accuracy": 0.16653283834457397,
      "num_tokens": 12483345.0,
      "step": 4925
    },
    {
      "entropy": 6.37202935218811,
      "epoch": 0.5689555683785343,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976675635654688,
      "loss": 6.2938,
      "mean_token_accuracy": 0.16135128289461137,
      "num_tokens": 12495675.0,
      "step": 4930
    },
    {
      "entropy": 6.417981672286987,
      "epoch": 0.569532602423543,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976616336354733,
      "loss": 6.3121,
      "mean_token_accuracy": 0.16374748200178146,
      "num_tokens": 12507569.0,
      "step": 4935
    },
    {
      "entropy": 6.435359477996826,
      "epoch": 0.5701096364685516,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976556962163645,
      "loss": 6.371,
      "mean_token_accuracy": 0.1623230755329132,
      "num_tokens": 12521198.0,
      "step": 4940
    },
    {
      "entropy": 6.379960393905639,
      "epoch": 0.5706866705135603,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004976497513083419,
      "loss": 6.2657,
      "mean_token_accuracy": 0.1603717714548111,
      "num_tokens": 12533242.0,
      "step": 4945
    },
    {
      "entropy": 6.486459255218506,
      "epoch": 0.571263704558569,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004976437989116055,
      "loss": 6.3877,
      "mean_token_accuracy": 0.16001511961221696,
      "num_tokens": 12545918.0,
      "step": 4950
    },
    {
      "entropy": 6.4408200740814205,
      "epoch": 0.5718407386035776,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004976378390263554,
      "loss": 6.324,
      "mean_token_accuracy": 0.1614006757736206,
      "num_tokens": 12558090.0,
      "step": 4955
    },
    {
      "entropy": 6.4780820369720455,
      "epoch": 0.5724177726485863,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004976318716527924,
      "loss": 6.348,
      "mean_token_accuracy": 0.16186215579509736,
      "num_tokens": 12571453.0,
      "step": 4960
    },
    {
      "entropy": 6.449974870681762,
      "epoch": 0.5729948066935949,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004976258967911168,
      "loss": 6.2967,
      "mean_token_accuracy": 0.1604265719652176,
      "num_tokens": 12584762.0,
      "step": 4965
    },
    {
      "entropy": 6.457449626922608,
      "epoch": 0.5735718407386036,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004976199144415298,
      "loss": 6.2847,
      "mean_token_accuracy": 0.15467647165060044,
      "num_tokens": 12597224.0,
      "step": 4970
    },
    {
      "entropy": 6.4018861770629885,
      "epoch": 0.5741488747836122,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004976139246042325,
      "loss": 6.3598,
      "mean_token_accuracy": 0.15818336009979247,
      "num_tokens": 12609837.0,
      "step": 4975
    },
    {
      "entropy": 6.476587915420533,
      "epoch": 0.5747259088286208,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004976079272794265,
      "loss": 6.3834,
      "mean_token_accuracy": 0.15177921205759048,
      "num_tokens": 12622596.0,
      "step": 4980
    },
    {
      "entropy": 6.460304689407349,
      "epoch": 0.5753029428736296,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976019224673134,
      "loss": 6.2836,
      "mean_token_accuracy": 0.16508358716964722,
      "num_tokens": 12634721.0,
      "step": 4985
    },
    {
      "entropy": 6.443642854690552,
      "epoch": 0.5758799769186382,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975959101680953,
      "loss": 6.3139,
      "mean_token_accuracy": 0.16074737012386323,
      "num_tokens": 12646389.0,
      "step": 4990
    },
    {
      "entropy": 6.481823301315307,
      "epoch": 0.5764570109636469,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004975898903819742,
      "loss": 6.4046,
      "mean_token_accuracy": 0.15126152634620665,
      "num_tokens": 12661081.0,
      "step": 4995
    },
    {
      "entropy": 6.560647344589233,
      "epoch": 0.5770340450086555,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004975838631091527,
      "loss": 6.3346,
      "mean_token_accuracy": 0.15827033370733262,
      "num_tokens": 12672505.0,
      "step": 5000
    },
    {
      "entropy": 6.421548557281494,
      "epoch": 0.5776110790536642,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975778283498336,
      "loss": 6.3526,
      "mean_token_accuracy": 0.15744248777627945,
      "num_tokens": 12684805.0,
      "step": 5005
    },
    {
      "entropy": 6.38816728591919,
      "epoch": 0.5781881130986728,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975717861042198,
      "loss": 6.3018,
      "mean_token_accuracy": 0.15945205241441726,
      "num_tokens": 12698557.0,
      "step": 5010
    },
    {
      "entropy": 6.495122861862183,
      "epoch": 0.5787651471436814,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004975657363725146,
      "loss": 6.3206,
      "mean_token_accuracy": 0.15233609974384307,
      "num_tokens": 12712472.0,
      "step": 5015
    },
    {
      "entropy": 6.413561677932739,
      "epoch": 0.5793421811886902,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975596791549213,
      "loss": 6.1908,
      "mean_token_accuracy": 0.16952557265758514,
      "num_tokens": 12725698.0,
      "step": 5020
    },
    {
      "entropy": 6.426093578338623,
      "epoch": 0.5799192152336988,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975536144516437,
      "loss": 6.2959,
      "mean_token_accuracy": 0.1619711399078369,
      "num_tokens": 12738126.0,
      "step": 5025
    },
    {
      "entropy": 6.4965479373931885,
      "epoch": 0.5804962492787075,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497547542262886,
      "loss": 6.3113,
      "mean_token_accuracy": 0.1720854163169861,
      "num_tokens": 12751402.0,
      "step": 5030
    },
    {
      "entropy": 6.415817785263061,
      "epoch": 0.5810732833237161,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975414625888521,
      "loss": 6.3234,
      "mean_token_accuracy": 0.15630837082862853,
      "num_tokens": 12763957.0,
      "step": 5035
    },
    {
      "entropy": 6.4553731918334964,
      "epoch": 0.5816503173687247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975353754297468,
      "loss": 6.2747,
      "mean_token_accuracy": 0.16455962955951692,
      "num_tokens": 12776498.0,
      "step": 5040
    },
    {
      "entropy": 6.456435108184815,
      "epoch": 0.5822273514137334,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975292807857745,
      "loss": 6.2916,
      "mean_token_accuracy": 0.16458143144845963,
      "num_tokens": 12788498.0,
      "step": 5045
    },
    {
      "entropy": 6.342044496536255,
      "epoch": 0.582804385458742,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004975231786571406,
      "loss": 6.2133,
      "mean_token_accuracy": 0.16185703426599501,
      "num_tokens": 12799907.0,
      "step": 5050
    },
    {
      "entropy": 6.425001573562622,
      "epoch": 0.5833814195037508,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975170690440499,
      "loss": 6.327,
      "mean_token_accuracy": 0.1625050887465477,
      "num_tokens": 12812610.0,
      "step": 5055
    },
    {
      "entropy": 6.462145233154297,
      "epoch": 0.5839584535487594,
      "grad_norm": 0.75,
      "learning_rate": 0.0004975109519467083,
      "loss": 6.3213,
      "mean_token_accuracy": 0.1591153174638748,
      "num_tokens": 12825082.0,
      "step": 5060
    },
    {
      "entropy": 6.347224283218384,
      "epoch": 0.584535487593768,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004975048273653212,
      "loss": 6.248,
      "mean_token_accuracy": 0.16950058192014694,
      "num_tokens": 12837999.0,
      "step": 5065
    },
    {
      "entropy": 6.363939332962036,
      "epoch": 0.5851125216387767,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974986953000948,
      "loss": 6.1894,
      "mean_token_accuracy": 0.17064935117959976,
      "num_tokens": 12850020.0,
      "step": 5070
    },
    {
      "entropy": 6.394009160995483,
      "epoch": 0.5856895556837853,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004974925557512354,
      "loss": 6.2981,
      "mean_token_accuracy": 0.16221174597740173,
      "num_tokens": 12863818.0,
      "step": 5075
    },
    {
      "entropy": 6.442247009277343,
      "epoch": 0.586266589728794,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004974864087189493,
      "loss": 6.3011,
      "mean_token_accuracy": 0.16013864129781724,
      "num_tokens": 12876959.0,
      "step": 5080
    },
    {
      "entropy": 6.445587396621704,
      "epoch": 0.5868436237738026,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974802542034434,
      "loss": 6.2764,
      "mean_token_accuracy": 0.15875303149223327,
      "num_tokens": 12889817.0,
      "step": 5085
    },
    {
      "entropy": 6.396495199203491,
      "epoch": 0.5874206578188114,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974740922049246,
      "loss": 6.2503,
      "mean_token_accuracy": 0.1640445441007614,
      "num_tokens": 12901438.0,
      "step": 5090
    },
    {
      "entropy": 6.383800745010376,
      "epoch": 0.58799769186382,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974679227236004,
      "loss": 6.1984,
      "mean_token_accuracy": 0.1692484885454178,
      "num_tokens": 12914115.0,
      "step": 5095
    },
    {
      "entropy": 6.406459999084473,
      "epoch": 0.5885747259088286,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974617457596779,
      "loss": 6.2579,
      "mean_token_accuracy": 0.1679217040538788,
      "num_tokens": 12926822.0,
      "step": 5100
    },
    {
      "entropy": 6.37530369758606,
      "epoch": 0.5891517599538373,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004974555613133652,
      "loss": 6.2917,
      "mean_token_accuracy": 0.1603657677769661,
      "num_tokens": 12939515.0,
      "step": 5105
    },
    {
      "entropy": 6.258018732070923,
      "epoch": 0.5897287939988459,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974493693848702,
      "loss": 6.2179,
      "mean_token_accuracy": 0.18086731135845185,
      "num_tokens": 12952195.0,
      "step": 5110
    },
    {
      "entropy": 6.452794170379638,
      "epoch": 0.5903058280438546,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004974431699744011,
      "loss": 6.2358,
      "mean_token_accuracy": 0.16284567266702651,
      "num_tokens": 12964563.0,
      "step": 5115
    },
    {
      "entropy": 6.354326820373535,
      "epoch": 0.5908828620888632,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004974369630821665,
      "loss": 6.167,
      "mean_token_accuracy": 0.16809700131416322,
      "num_tokens": 12976564.0,
      "step": 5120
    },
    {
      "entropy": 6.21676197052002,
      "epoch": 0.5914598961338718,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974307487083752,
      "loss": 6.124,
      "mean_token_accuracy": 0.16871170401573182,
      "num_tokens": 12989757.0,
      "step": 5125
    },
    {
      "entropy": 6.430266237258911,
      "epoch": 0.5920369301788806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974245268532361,
      "loss": 6.3467,
      "mean_token_accuracy": 0.16050200611352922,
      "num_tokens": 13003846.0,
      "step": 5130
    },
    {
      "entropy": 6.493060302734375,
      "epoch": 0.5926139642238892,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004974182975169585,
      "loss": 6.3453,
      "mean_token_accuracy": 0.1525108903646469,
      "num_tokens": 13018389.0,
      "step": 5135
    },
    {
      "entropy": 6.47206563949585,
      "epoch": 0.5931909982688979,
      "grad_norm": 0.796875,
      "learning_rate": 0.000497412060699752,
      "loss": 6.3402,
      "mean_token_accuracy": 0.1555670529603958,
      "num_tokens": 13031893.0,
      "step": 5140
    },
    {
      "entropy": 6.444800662994385,
      "epoch": 0.5937680323139065,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004974058164018263,
      "loss": 6.2744,
      "mean_token_accuracy": 0.16890939027071,
      "num_tokens": 13045916.0,
      "step": 5145
    },
    {
      "entropy": 6.461399793624878,
      "epoch": 0.5943450663589152,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004973995646233914,
      "loss": 6.3126,
      "mean_token_accuracy": 0.1604936346411705,
      "num_tokens": 13058468.0,
      "step": 5150
    },
    {
      "entropy": 6.452783203125,
      "epoch": 0.5949221004039238,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973933053646576,
      "loss": 6.3328,
      "mean_token_accuracy": 0.15725631415843963,
      "num_tokens": 13070778.0,
      "step": 5155
    },
    {
      "entropy": 6.452268409729004,
      "epoch": 0.5954991344489324,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004973870386258355,
      "loss": 6.2174,
      "mean_token_accuracy": 0.1663954645395279,
      "num_tokens": 13082287.0,
      "step": 5160
    },
    {
      "entropy": 6.329747056961059,
      "epoch": 0.5960761684939412,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004973807644071357,
      "loss": 6.2168,
      "mean_token_accuracy": 0.1709250345826149,
      "num_tokens": 13095505.0,
      "step": 5165
    },
    {
      "entropy": 6.327900695800781,
      "epoch": 0.5966532025389498,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004973744827087695,
      "loss": 6.2446,
      "mean_token_accuracy": 0.16142298728227616,
      "num_tokens": 13108351.0,
      "step": 5170
    },
    {
      "entropy": 6.483879852294922,
      "epoch": 0.5972302365839585,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497368193530948,
      "loss": 6.303,
      "mean_token_accuracy": 0.16267842650413514,
      "num_tokens": 13120941.0,
      "step": 5175
    },
    {
      "entropy": 6.439163446426392,
      "epoch": 0.5978072706289671,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004973618968738828,
      "loss": 6.2475,
      "mean_token_accuracy": 0.16663485914468765,
      "num_tokens": 13134217.0,
      "step": 5180
    },
    {
      "entropy": 6.345459890365601,
      "epoch": 0.5983843046739757,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973555927377856,
      "loss": 6.2653,
      "mean_token_accuracy": 0.1663901150226593,
      "num_tokens": 13146245.0,
      "step": 5185
    },
    {
      "entropy": 6.438876342773438,
      "epoch": 0.5989613387189844,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973492811228685,
      "loss": 6.287,
      "mean_token_accuracy": 0.1630580872297287,
      "num_tokens": 13158498.0,
      "step": 5190
    },
    {
      "entropy": 6.429390382766724,
      "epoch": 0.599538372763993,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004973429620293438,
      "loss": 6.3184,
      "mean_token_accuracy": 0.15690135210752487,
      "num_tokens": 13170464.0,
      "step": 5195
    },
    {
      "entropy": 6.437797021865845,
      "epoch": 0.6001154068090018,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973366354574239,
      "loss": 6.3777,
      "mean_token_accuracy": 0.1602035418152809,
      "num_tokens": 13183315.0,
      "step": 5200
    },
    {
      "entropy": 6.518604564666748,
      "epoch": 0.6006924408540104,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004973303014073219,
      "loss": 6.405,
      "mean_token_accuracy": 0.15694308578968047,
      "num_tokens": 13196536.0,
      "step": 5205
    },
    {
      "entropy": 6.4371904850006105,
      "epoch": 0.601269474899019,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004973239598792504,
      "loss": 6.2495,
      "mean_token_accuracy": 0.16467582434415817,
      "num_tokens": 13209159.0,
      "step": 5210
    },
    {
      "entropy": 6.449504613876343,
      "epoch": 0.6018465089440277,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004973176108734232,
      "loss": 6.2826,
      "mean_token_accuracy": 0.1641786962747574,
      "num_tokens": 13221563.0,
      "step": 5215
    },
    {
      "entropy": 6.4400306224823,
      "epoch": 0.6024235429890363,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973112543900535,
      "loss": 6.3671,
      "mean_token_accuracy": 0.1555931881070137,
      "num_tokens": 13234209.0,
      "step": 5220
    },
    {
      "entropy": 6.4899622917175295,
      "epoch": 0.603000577034045,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973048904293551,
      "loss": 6.2561,
      "mean_token_accuracy": 0.15689144134521485,
      "num_tokens": 13246198.0,
      "step": 5225
    },
    {
      "entropy": 6.364943170547486,
      "epoch": 0.6035776110790536,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004972985189915422,
      "loss": 6.273,
      "mean_token_accuracy": 0.16756874471902847,
      "num_tokens": 13259149.0,
      "step": 5230
    },
    {
      "entropy": 6.379005336761475,
      "epoch": 0.6041546451240624,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497292140076829,
      "loss": 6.2695,
      "mean_token_accuracy": 0.1631223067641258,
      "num_tokens": 13271138.0,
      "step": 5235
    },
    {
      "entropy": 6.40664587020874,
      "epoch": 0.604731679169071,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972857536854301,
      "loss": 6.2714,
      "mean_token_accuracy": 0.15566522479057313,
      "num_tokens": 13283639.0,
      "step": 5240
    },
    {
      "entropy": 6.401790046691895,
      "epoch": 0.6053087132140796,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972793598175603,
      "loss": 6.3049,
      "mean_token_accuracy": 0.16139982044696807,
      "num_tokens": 13295322.0,
      "step": 5245
    },
    {
      "entropy": 6.381245756149292,
      "epoch": 0.6058857472590883,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972729584734347,
      "loss": 6.2634,
      "mean_token_accuracy": 0.16289384216070174,
      "num_tokens": 13308431.0,
      "step": 5250
    },
    {
      "entropy": 6.390211057662964,
      "epoch": 0.6064627813040969,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004972665496532685,
      "loss": 6.2001,
      "mean_token_accuracy": 0.16877303421497344,
      "num_tokens": 13321878.0,
      "step": 5255
    },
    {
      "entropy": 6.296879863739013,
      "epoch": 0.6070398153491056,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004972601333572774,
      "loss": 6.2828,
      "mean_token_accuracy": 0.16521066427230835,
      "num_tokens": 13334166.0,
      "step": 5260
    },
    {
      "entropy": 6.450336933135986,
      "epoch": 0.6076168493941142,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004972537095856769,
      "loss": 6.3535,
      "mean_token_accuracy": 0.16121897250413894,
      "num_tokens": 13347704.0,
      "step": 5265
    },
    {
      "entropy": 6.421620845794678,
      "epoch": 0.608193883439123,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972472783386834,
      "loss": 6.2252,
      "mean_token_accuracy": 0.1674308717250824,
      "num_tokens": 13359659.0,
      "step": 5270
    },
    {
      "entropy": 6.4067305564880375,
      "epoch": 0.6087709174841316,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972408396165132,
      "loss": 6.2504,
      "mean_token_accuracy": 0.16205257177352905,
      "num_tokens": 13371952.0,
      "step": 5275
    },
    {
      "entropy": 6.460360765457153,
      "epoch": 0.6093479515291402,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004972343934193826,
      "loss": 6.2829,
      "mean_token_accuracy": 0.16754993051290512,
      "num_tokens": 13385757.0,
      "step": 5280
    },
    {
      "entropy": 6.375572776794433,
      "epoch": 0.6099249855741489,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004972279397475086,
      "loss": 6.2449,
      "mean_token_accuracy": 0.1626361846923828,
      "num_tokens": 13397657.0,
      "step": 5285
    },
    {
      "entropy": 6.354529762268067,
      "epoch": 0.6105020196191575,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972214786011083,
      "loss": 6.2598,
      "mean_token_accuracy": 0.16665343195199966,
      "num_tokens": 13410882.0,
      "step": 5290
    },
    {
      "entropy": 6.385918283462525,
      "epoch": 0.6110790536641661,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000497215009980399,
      "loss": 6.2998,
      "mean_token_accuracy": 0.15810784995555877,
      "num_tokens": 13423310.0,
      "step": 5295
    },
    {
      "entropy": 6.431512498855591,
      "epoch": 0.6116560877091748,
      "grad_norm": 0.76171875,
      "learning_rate": 0.000497208533885598,
      "loss": 6.2736,
      "mean_token_accuracy": 0.1636954054236412,
      "num_tokens": 13437837.0,
      "step": 5300
    },
    {
      "entropy": 6.522774076461792,
      "epoch": 0.6122331217541835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972020503169235,
      "loss": 6.2688,
      "mean_token_accuracy": 0.1651971697807312,
      "num_tokens": 13450637.0,
      "step": 5305
    },
    {
      "entropy": 6.39950647354126,
      "epoch": 0.6128101557991922,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971955592745934,
      "loss": 6.3332,
      "mean_token_accuracy": 0.1568018838763237,
      "num_tokens": 13464429.0,
      "step": 5310
    },
    {
      "entropy": 6.408504819869995,
      "epoch": 0.6133871898442008,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497189060758826,
      "loss": 6.1998,
      "mean_token_accuracy": 0.1667480379343033,
      "num_tokens": 13476710.0,
      "step": 5315
    },
    {
      "entropy": 6.397567081451416,
      "epoch": 0.6139642238892095,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00049718255476984,
      "loss": 6.251,
      "mean_token_accuracy": 0.16203757673501967,
      "num_tokens": 13489828.0,
      "step": 5320
    },
    {
      "entropy": 6.430128335952759,
      "epoch": 0.6145412579342181,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971760413078539,
      "loss": 6.2457,
      "mean_token_accuracy": 0.16466025859117508,
      "num_tokens": 13503584.0,
      "step": 5325
    },
    {
      "entropy": 6.360280466079712,
      "epoch": 0.6151182919792267,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004971695203730872,
      "loss": 6.2378,
      "mean_token_accuracy": 0.16777390688657762,
      "num_tokens": 13514658.0,
      "step": 5330
    },
    {
      "entropy": 6.3734703540802,
      "epoch": 0.6156953260242354,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497162991965759,
      "loss": 6.2795,
      "mean_token_accuracy": 0.1756128638982773,
      "num_tokens": 13526424.0,
      "step": 5335
    },
    {
      "entropy": 6.390287590026856,
      "epoch": 0.6162723600692441,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004971564560860889,
      "loss": 6.2342,
      "mean_token_accuracy": 0.1635016083717346,
      "num_tokens": 13540515.0,
      "step": 5340
    },
    {
      "entropy": 6.433468246459961,
      "epoch": 0.6168493941142528,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004971499127342968,
      "loss": 6.3327,
      "mean_token_accuracy": 0.1582094430923462,
      "num_tokens": 13553634.0,
      "step": 5345
    },
    {
      "entropy": 6.42935209274292,
      "epoch": 0.6174264281592614,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971433619106027,
      "loss": 6.3363,
      "mean_token_accuracy": 0.16416244804859162,
      "num_tokens": 13565168.0,
      "step": 5350
    },
    {
      "entropy": 6.381501960754394,
      "epoch": 0.61800346220427,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497136803615227,
      "loss": 6.2739,
      "mean_token_accuracy": 0.16165704429149627,
      "num_tokens": 13578511.0,
      "step": 5355
    },
    {
      "entropy": 6.401495885848999,
      "epoch": 0.6185804962492787,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004971302378483902,
      "loss": 6.2496,
      "mean_token_accuracy": 0.1660989612340927,
      "num_tokens": 13590535.0,
      "step": 5360
    },
    {
      "entropy": 6.3950817584991455,
      "epoch": 0.6191575302942873,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004971236646103132,
      "loss": 6.2847,
      "mean_token_accuracy": 0.16551341712474824,
      "num_tokens": 13604549.0,
      "step": 5365
    },
    {
      "entropy": 6.423378038406372,
      "epoch": 0.619734564339296,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004971170839012171,
      "loss": 6.2773,
      "mean_token_accuracy": 0.158201265335083,
      "num_tokens": 13617658.0,
      "step": 5370
    },
    {
      "entropy": 6.378845930099487,
      "epoch": 0.6203115983843047,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971104957213233,
      "loss": 6.1836,
      "mean_token_accuracy": 0.1718725860118866,
      "num_tokens": 13630559.0,
      "step": 5375
    },
    {
      "entropy": 6.347926759719849,
      "epoch": 0.6208886324293134,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971039000708531,
      "loss": 6.2582,
      "mean_token_accuracy": 0.16711993515491486,
      "num_tokens": 13643018.0,
      "step": 5380
    },
    {
      "entropy": 6.414066219329834,
      "epoch": 0.621465666474322,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004970972969500286,
      "loss": 6.2407,
      "mean_token_accuracy": 0.16316382586956024,
      "num_tokens": 13655053.0,
      "step": 5385
    },
    {
      "entropy": 6.40781569480896,
      "epoch": 0.6220427005193306,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000497090686359072,
      "loss": 6.2107,
      "mean_token_accuracy": 0.16605478525161743,
      "num_tokens": 13667420.0,
      "step": 5390
    },
    {
      "entropy": 6.3122974872589115,
      "epoch": 0.6226197345643393,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004970840682982054,
      "loss": 6.1826,
      "mean_token_accuracy": 0.16799420565366746,
      "num_tokens": 13680662.0,
      "step": 5395
    },
    {
      "entropy": 6.402325630187988,
      "epoch": 0.6231967686093479,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004970774427676514,
      "loss": 6.2935,
      "mean_token_accuracy": 0.15851684659719467,
      "num_tokens": 13694720.0,
      "step": 5400
    },
    {
      "entropy": 6.4278172016143795,
      "epoch": 0.6237738026543566,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497070809767633,
      "loss": 6.212,
      "mean_token_accuracy": 0.1631224125623703,
      "num_tokens": 13706826.0,
      "step": 5405
    },
    {
      "entropy": 6.314861869812011,
      "epoch": 0.6243508366993653,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004970641692983731,
      "loss": 6.2796,
      "mean_token_accuracy": 0.1627788543701172,
      "num_tokens": 13719526.0,
      "step": 5410
    },
    {
      "entropy": 6.395055103302002,
      "epoch": 0.624927870744374,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970575213600954,
      "loss": 6.1683,
      "mean_token_accuracy": 0.17014200389385223,
      "num_tokens": 13731583.0,
      "step": 5415
    },
    {
      "entropy": 6.429858112335205,
      "epoch": 0.6255049047893826,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004970508659530231,
      "loss": 6.2841,
      "mean_token_accuracy": 0.16601747125387192,
      "num_tokens": 13743690.0,
      "step": 5420
    },
    {
      "entropy": 6.385574388504028,
      "epoch": 0.6260819388343912,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970442030773802,
      "loss": 6.2359,
      "mean_token_accuracy": 0.1637325868010521,
      "num_tokens": 13756264.0,
      "step": 5425
    },
    {
      "entropy": 6.392513990402222,
      "epoch": 0.6266589728793999,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970375327333909,
      "loss": 6.2579,
      "mean_token_accuracy": 0.1582666516304016,
      "num_tokens": 13768089.0,
      "step": 5430
    },
    {
      "entropy": 6.420020866394043,
      "epoch": 0.6272360069244085,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004970308549212796,
      "loss": 6.2961,
      "mean_token_accuracy": 0.16104890406131744,
      "num_tokens": 13780671.0,
      "step": 5435
    },
    {
      "entropy": 6.4573148727417,
      "epoch": 0.6278130409694171,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004970241696412705,
      "loss": 6.2876,
      "mean_token_accuracy": 0.160064260661602,
      "num_tokens": 13793678.0,
      "step": 5440
    },
    {
      "entropy": 6.423211479187012,
      "epoch": 0.6283900750144259,
      "grad_norm": 0.78125,
      "learning_rate": 0.000497017476893589,
      "loss": 6.2226,
      "mean_token_accuracy": 0.16376707404851915,
      "num_tokens": 13807001.0,
      "step": 5445
    },
    {
      "entropy": 6.402011871337891,
      "epoch": 0.6289671090594345,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004970107766784598,
      "loss": 6.1794,
      "mean_token_accuracy": 0.16806947141885759,
      "num_tokens": 13820227.0,
      "step": 5450
    },
    {
      "entropy": 6.405494213104248,
      "epoch": 0.6295441431044432,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004970040689961084,
      "loss": 6.2841,
      "mean_token_accuracy": 0.16475358456373215,
      "num_tokens": 13831663.0,
      "step": 5455
    },
    {
      "entropy": 6.29299693107605,
      "epoch": 0.6301211771494518,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004969973538467605,
      "loss": 6.1821,
      "mean_token_accuracy": 0.16974693536758423,
      "num_tokens": 13845173.0,
      "step": 5460
    },
    {
      "entropy": 6.43713493347168,
      "epoch": 0.6306982111944605,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969906312306419,
      "loss": 6.231,
      "mean_token_accuracy": 0.16185085326433182,
      "num_tokens": 13857551.0,
      "step": 5465
    },
    {
      "entropy": 6.28310809135437,
      "epoch": 0.6312752452394691,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004969839011479786,
      "loss": 6.1619,
      "mean_token_accuracy": 0.17454650551080703,
      "num_tokens": 13870238.0,
      "step": 5470
    },
    {
      "entropy": 6.3346006870269775,
      "epoch": 0.6318522792844777,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000496977163598997,
      "loss": 6.2471,
      "mean_token_accuracy": 0.1618565484881401,
      "num_tokens": 13883145.0,
      "step": 5475
    },
    {
      "entropy": 6.452572059631348,
      "epoch": 0.6324293133294865,
      "grad_norm": 0.875,
      "learning_rate": 0.0004969704185839239,
      "loss": 6.2915,
      "mean_token_accuracy": 0.15788978338241577,
      "num_tokens": 13895777.0,
      "step": 5480
    },
    {
      "entropy": 6.3593542098999025,
      "epoch": 0.6330063473744951,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004969636661029859,
      "loss": 6.1848,
      "mean_token_accuracy": 0.17119636833667756,
      "num_tokens": 13907641.0,
      "step": 5485
    },
    {
      "entropy": 6.476277828216553,
      "epoch": 0.6335833814195038,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969569061564103,
      "loss": 6.2872,
      "mean_token_accuracy": 0.1563932254910469,
      "num_tokens": 13921576.0,
      "step": 5490
    },
    {
      "entropy": 6.4062957763671875,
      "epoch": 0.6341604154645124,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969501387444245,
      "loss": 6.2387,
      "mean_token_accuracy": 0.16197298020124434,
      "num_tokens": 13935158.0,
      "step": 5495
    },
    {
      "entropy": 6.29954342842102,
      "epoch": 0.634737449509521,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004969433638672559,
      "loss": 6.1988,
      "mean_token_accuracy": 0.16927525103092195,
      "num_tokens": 13947803.0,
      "step": 5500
    },
    {
      "entropy": 6.441785907745361,
      "epoch": 0.6353144835545297,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969365815251325,
      "loss": 6.2764,
      "mean_token_accuracy": 0.16404346376657486,
      "num_tokens": 13960573.0,
      "step": 5505
    },
    {
      "entropy": 6.359479236602783,
      "epoch": 0.6358915175995383,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969297917182825,
      "loss": 6.15,
      "mean_token_accuracy": 0.16838484555482863,
      "num_tokens": 13970994.0,
      "step": 5510
    },
    {
      "entropy": 6.370149993896485,
      "epoch": 0.6364685516445471,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004969229944469342,
      "loss": 6.3613,
      "mean_token_accuracy": 0.15786928087472915,
      "num_tokens": 13983617.0,
      "step": 5515
    },
    {
      "entropy": 6.4495750904083256,
      "epoch": 0.6370455856895557,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004969161897113162,
      "loss": 6.2431,
      "mean_token_accuracy": 0.16852736622095107,
      "num_tokens": 13995779.0,
      "step": 5520
    },
    {
      "entropy": 6.342259693145752,
      "epoch": 0.6376226197345644,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004969093775116574,
      "loss": 6.2784,
      "mean_token_accuracy": 0.1653564378619194,
      "num_tokens": 14008650.0,
      "step": 5525
    },
    {
      "entropy": 6.3147563457489015,
      "epoch": 0.638199653779573,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004969025578481869,
      "loss": 6.1887,
      "mean_token_accuracy": 0.16722843796014786,
      "num_tokens": 14021787.0,
      "step": 5530
    },
    {
      "entropy": 6.5476946353912355,
      "epoch": 0.6387766878245816,
      "grad_norm": 0.84375,
      "learning_rate": 0.000496895730721134,
      "loss": 6.3405,
      "mean_token_accuracy": 0.15578458309173585,
      "num_tokens": 14035036.0,
      "step": 5535
    },
    {
      "entropy": 6.4367437839508055,
      "epoch": 0.6393537218695903,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004968888961307286,
      "loss": 6.2784,
      "mean_token_accuracy": 0.16260789930820466,
      "num_tokens": 14048589.0,
      "step": 5540
    },
    {
      "entropy": 6.335751581192016,
      "epoch": 0.6399307559145989,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968820540772003,
      "loss": 6.1649,
      "mean_token_accuracy": 0.18534766435623168,
      "num_tokens": 14061007.0,
      "step": 5545
    },
    {
      "entropy": 6.402452230453491,
      "epoch": 0.6405077899596077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968752045607794,
      "loss": 6.2971,
      "mean_token_accuracy": 0.16098586469888687,
      "num_tokens": 14074165.0,
      "step": 5550
    },
    {
      "entropy": 6.505826616287232,
      "epoch": 0.6410848240046163,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004968683475816961,
      "loss": 6.355,
      "mean_token_accuracy": 0.15098087936639787,
      "num_tokens": 14086261.0,
      "step": 5555
    },
    {
      "entropy": 6.418167352676392,
      "epoch": 0.641661858049625,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004968614831401811,
      "loss": 6.1924,
      "mean_token_accuracy": 0.17131647616624832,
      "num_tokens": 14098458.0,
      "step": 5560
    },
    {
      "entropy": 6.348795795440674,
      "epoch": 0.6422388920946336,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004968546112364654,
      "loss": 6.3645,
      "mean_token_accuracy": 0.1578898921608925,
      "num_tokens": 14110754.0,
      "step": 5565
    },
    {
      "entropy": 6.420198011398315,
      "epoch": 0.6428159261396422,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049684773187078,
      "loss": 6.249,
      "mean_token_accuracy": 0.1605192869901657,
      "num_tokens": 14123361.0,
      "step": 5570
    },
    {
      "entropy": 6.400240230560303,
      "epoch": 0.6433929601846509,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968408450433565,
      "loss": 6.2097,
      "mean_token_accuracy": 0.17002205550670624,
      "num_tokens": 14135576.0,
      "step": 5575
    },
    {
      "entropy": 6.366382312774658,
      "epoch": 0.6439699942296595,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004968339507544263,
      "loss": 6.2722,
      "mean_token_accuracy": 0.16116065233945848,
      "num_tokens": 14147879.0,
      "step": 5580
    },
    {
      "entropy": 6.426327133178711,
      "epoch": 0.6445470282746683,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968270490042212,
      "loss": 6.3084,
      "mean_token_accuracy": 0.16116396337747574,
      "num_tokens": 14159449.0,
      "step": 5585
    },
    {
      "entropy": 6.321964263916016,
      "epoch": 0.6451240623196769,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004968201397929737,
      "loss": 6.1803,
      "mean_token_accuracy": 0.16564835458993912,
      "num_tokens": 14172547.0,
      "step": 5590
    },
    {
      "entropy": 6.398483514785767,
      "epoch": 0.6457010963646855,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004968132231209158,
      "loss": 6.2574,
      "mean_token_accuracy": 0.1655385360121727,
      "num_tokens": 14184670.0,
      "step": 5595
    },
    {
      "entropy": 6.39441933631897,
      "epoch": 0.6462781304096942,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004968062989882803,
      "loss": 6.2306,
      "mean_token_accuracy": 0.16546582281589509,
      "num_tokens": 14198048.0,
      "step": 5600
    },
    {
      "entropy": 6.324096584320069,
      "epoch": 0.6468551644547028,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967993673953003,
      "loss": 6.1823,
      "mean_token_accuracy": 0.1676321193575859,
      "num_tokens": 14209470.0,
      "step": 5605
    },
    {
      "entropy": 6.376352262496948,
      "epoch": 0.6474321984997115,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967924283422087,
      "loss": 6.1919,
      "mean_token_accuracy": 0.1653391182422638,
      "num_tokens": 14221731.0,
      "step": 5610
    },
    {
      "entropy": 6.435838794708252,
      "epoch": 0.6480092325447201,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000496785481829239,
      "loss": 6.2543,
      "mean_token_accuracy": 0.15581413954496384,
      "num_tokens": 14235125.0,
      "step": 5615
    },
    {
      "entropy": 6.372358417510986,
      "epoch": 0.6485862665897288,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004967785278566245,
      "loss": 6.2227,
      "mean_token_accuracy": 0.1686449021100998,
      "num_tokens": 14248941.0,
      "step": 5620
    },
    {
      "entropy": 6.3675590515136715,
      "epoch": 0.6491633006347375,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004967715664245997,
      "loss": 6.1475,
      "mean_token_accuracy": 0.17231432795524598,
      "num_tokens": 14260955.0,
      "step": 5625
    },
    {
      "entropy": 6.30697455406189,
      "epoch": 0.6497403346797461,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967645975333983,
      "loss": 6.1607,
      "mean_token_accuracy": 0.17377639561891556,
      "num_tokens": 14274123.0,
      "step": 5630
    },
    {
      "entropy": 6.441181182861328,
      "epoch": 0.6503173687247548,
      "grad_norm": 0.875,
      "learning_rate": 0.0004967576211832548,
      "loss": 6.2351,
      "mean_token_accuracy": 0.17109169363975524,
      "num_tokens": 14286920.0,
      "step": 5635
    },
    {
      "entropy": 6.349151277542115,
      "epoch": 0.6508944027697634,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967506373744039,
      "loss": 6.2187,
      "mean_token_accuracy": 0.16517532244324684,
      "num_tokens": 14299297.0,
      "step": 5640
    },
    {
      "entropy": 6.364994144439697,
      "epoch": 0.651471436814772,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004967436461070805,
      "loss": 6.215,
      "mean_token_accuracy": 0.16879084408283235,
      "num_tokens": 14312107.0,
      "step": 5645
    },
    {
      "entropy": 6.362016820907593,
      "epoch": 0.6520484708597807,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0004967366473815196,
      "loss": 6.216,
      "mean_token_accuracy": 0.16388487368822097,
      "num_tokens": 14325777.0,
      "step": 5650
    },
    {
      "entropy": 6.399789810180664,
      "epoch": 0.6526255049047894,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004967296411979568,
      "loss": 6.224,
      "mean_token_accuracy": 0.16334544569253923,
      "num_tokens": 14337617.0,
      "step": 5655
    },
    {
      "entropy": 6.287733173370361,
      "epoch": 0.6532025389497981,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004967226275566275,
      "loss": 6.166,
      "mean_token_accuracy": 0.17259837239980697,
      "num_tokens": 14349756.0,
      "step": 5660
    },
    {
      "entropy": 6.386903429031372,
      "epoch": 0.6537795729948067,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496715606457768,
      "loss": 6.2819,
      "mean_token_accuracy": 0.16123751401901246,
      "num_tokens": 14363358.0,
      "step": 5665
    },
    {
      "entropy": 6.398137712478638,
      "epoch": 0.6543566070398154,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004967085779016142,
      "loss": 6.2254,
      "mean_token_accuracy": 0.1677267700433731,
      "num_tokens": 14375510.0,
      "step": 5670
    },
    {
      "entropy": 6.38195309638977,
      "epoch": 0.654933641084824,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967015418884022,
      "loss": 6.239,
      "mean_token_accuracy": 0.16811206191778183,
      "num_tokens": 14389523.0,
      "step": 5675
    },
    {
      "entropy": 6.312136507034301,
      "epoch": 0.6555106751298326,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966944984183692,
      "loss": 6.2165,
      "mean_token_accuracy": 0.1704944595694542,
      "num_tokens": 14400457.0,
      "step": 5680
    },
    {
      "entropy": 6.264360666275024,
      "epoch": 0.6560877091748413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966874474917518,
      "loss": 6.1747,
      "mean_token_accuracy": 0.17271694391965867,
      "num_tokens": 14413088.0,
      "step": 5685
    },
    {
      "entropy": 6.444734716415406,
      "epoch": 0.65666474321985,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966803891087873,
      "loss": 6.2309,
      "mean_token_accuracy": 0.16306148767471312,
      "num_tokens": 14425816.0,
      "step": 5690
    },
    {
      "entropy": 6.303487730026245,
      "epoch": 0.6572417772648587,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496673323269713,
      "loss": 6.1156,
      "mean_token_accuracy": 0.17381228059530257,
      "num_tokens": 14438739.0,
      "step": 5695
    },
    {
      "entropy": 6.31482458114624,
      "epoch": 0.6578188113098673,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966662499747666,
      "loss": 6.2582,
      "mean_token_accuracy": 0.16371830701828002,
      "num_tokens": 14451543.0,
      "step": 5700
    },
    {
      "entropy": 6.4650774002075195,
      "epoch": 0.6583958453548759,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966591692241859,
      "loss": 6.2895,
      "mean_token_accuracy": 0.1613236203789711,
      "num_tokens": 14464969.0,
      "step": 5705
    },
    {
      "entropy": 6.410737085342407,
      "epoch": 0.6589728793998846,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966520810182092,
      "loss": 6.2467,
      "mean_token_accuracy": 0.16270333379507065,
      "num_tokens": 14477622.0,
      "step": 5710
    },
    {
      "entropy": 6.424910831451416,
      "epoch": 0.6595499134448932,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966449853570749,
      "loss": 6.2512,
      "mean_token_accuracy": 0.16616804748773575,
      "num_tokens": 14490306.0,
      "step": 5715
    },
    {
      "entropy": 6.296383047103882,
      "epoch": 0.6601269474899019,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966378822410215,
      "loss": 6.1068,
      "mean_token_accuracy": 0.18176421523094177,
      "num_tokens": 14501701.0,
      "step": 5720
    },
    {
      "entropy": 6.349727487564087,
      "epoch": 0.6607039815349106,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004966307716702881,
      "loss": 6.2303,
      "mean_token_accuracy": 0.16906958520412446,
      "num_tokens": 14515700.0,
      "step": 5725
    },
    {
      "entropy": 6.392299318313599,
      "epoch": 0.6612810155799193,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004966236536451138,
      "loss": 6.2194,
      "mean_token_accuracy": 0.1620727926492691,
      "num_tokens": 14528583.0,
      "step": 5730
    },
    {
      "entropy": 6.321354389190674,
      "epoch": 0.6618580496249279,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000496616528165738,
      "loss": 6.1655,
      "mean_token_accuracy": 0.17074524164199828,
      "num_tokens": 14541896.0,
      "step": 5735
    },
    {
      "entropy": 6.323581647872925,
      "epoch": 0.6624350836699365,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966093952324003,
      "loss": 6.0875,
      "mean_token_accuracy": 0.1751362219452858,
      "num_tokens": 14556948.0,
      "step": 5740
    },
    {
      "entropy": 6.376759576797485,
      "epoch": 0.6630121177149452,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966022548453406,
      "loss": 6.2559,
      "mean_token_accuracy": 0.16236974000930787,
      "num_tokens": 14568802.0,
      "step": 5745
    },
    {
      "entropy": 6.376679134368897,
      "epoch": 0.6635891517599538,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004965951070047993,
      "loss": 6.2158,
      "mean_token_accuracy": 0.1644376128911972,
      "num_tokens": 14582664.0,
      "step": 5750
    },
    {
      "entropy": 6.3371021270751955,
      "epoch": 0.6641661858049625,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965879517110166,
      "loss": 6.3063,
      "mean_token_accuracy": 0.16026860177516938,
      "num_tokens": 14596805.0,
      "step": 5755
    },
    {
      "entropy": 6.394929265975952,
      "epoch": 0.6647432198499712,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965807889642333,
      "loss": 6.2369,
      "mean_token_accuracy": 0.16370510458946227,
      "num_tokens": 14610532.0,
      "step": 5760
    },
    {
      "entropy": 6.368929100036621,
      "epoch": 0.6653202538949798,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049657361876469,
      "loss": 6.2615,
      "mean_token_accuracy": 0.16059991121292114,
      "num_tokens": 14622928.0,
      "step": 5765
    },
    {
      "entropy": 6.305924272537231,
      "epoch": 0.6658972879399885,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004965664411126282,
      "loss": 6.1307,
      "mean_token_accuracy": 0.16946146190166472,
      "num_tokens": 14634410.0,
      "step": 5770
    },
    {
      "entropy": 6.247943449020386,
      "epoch": 0.6664743219849971,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004965592560082894,
      "loss": 6.1508,
      "mean_token_accuracy": 0.16843760758638382,
      "num_tokens": 14647571.0,
      "step": 5775
    },
    {
      "entropy": 6.257874631881714,
      "epoch": 0.6670513560300058,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004965520634519149,
      "loss": 6.0601,
      "mean_token_accuracy": 0.18189057409763337,
      "num_tokens": 14660656.0,
      "step": 5780
    },
    {
      "entropy": 6.397317409515381,
      "epoch": 0.6676283900750144,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965448634437468,
      "loss": 6.2544,
      "mean_token_accuracy": 0.16246428936719895,
      "num_tokens": 14672267.0,
      "step": 5785
    },
    {
      "entropy": 6.36127347946167,
      "epoch": 0.668205424120023,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004965376559840272,
      "loss": 6.2441,
      "mean_token_accuracy": 0.15803717374801635,
      "num_tokens": 14684590.0,
      "step": 5790
    },
    {
      "entropy": 6.369046449661255,
      "epoch": 0.6687824581650318,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004965304410729986,
      "loss": 6.2395,
      "mean_token_accuracy": 0.1638997569680214,
      "num_tokens": 14697299.0,
      "step": 5795
    },
    {
      "entropy": 6.288568830490112,
      "epoch": 0.6693594922100404,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004965232187109037,
      "loss": 6.227,
      "mean_token_accuracy": 0.1662815257906914,
      "num_tokens": 14709328.0,
      "step": 5800
    },
    {
      "entropy": 6.3782798767089846,
      "epoch": 0.6699365262550491,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004965159888979854,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16616736054420472,
      "num_tokens": 14721786.0,
      "step": 5805
    },
    {
      "entropy": 6.3899914741516115,
      "epoch": 0.6705135603000577,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965087516344869,
      "loss": 6.2694,
      "mean_token_accuracy": 0.1632115736603737,
      "num_tokens": 14734120.0,
      "step": 5810
    },
    {
      "entropy": 6.351525115966797,
      "epoch": 0.6710905943450663,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965015069206515,
      "loss": 6.2735,
      "mean_token_accuracy": 0.16400251686573028,
      "num_tokens": 14746159.0,
      "step": 5815
    },
    {
      "entropy": 6.435773468017578,
      "epoch": 0.671667628390075,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496494254756723,
      "loss": 6.2072,
      "mean_token_accuracy": 0.1680669292807579,
      "num_tokens": 14758132.0,
      "step": 5820
    },
    {
      "entropy": 6.281584072113037,
      "epoch": 0.6722446624350836,
      "grad_norm": 0.75,
      "learning_rate": 0.0004964869951429451,
      "loss": 6.0942,
      "mean_token_accuracy": 0.17387358695268632,
      "num_tokens": 14770934.0,
      "step": 5825
    },
    {
      "entropy": 6.296800994873047,
      "epoch": 0.6728216964800924,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004964797280795622,
      "loss": 6.2519,
      "mean_token_accuracy": 0.16522246748209,
      "num_tokens": 14785084.0,
      "step": 5830
    },
    {
      "entropy": 6.402192687988281,
      "epoch": 0.673398730525101,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004964724535668188,
      "loss": 6.3061,
      "mean_token_accuracy": 0.17040158063173294,
      "num_tokens": 14798553.0,
      "step": 5835
    },
    {
      "entropy": 6.397847509384155,
      "epoch": 0.6739757645701097,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964651716049593,
      "loss": 6.2497,
      "mean_token_accuracy": 0.16112502813339233,
      "num_tokens": 14810239.0,
      "step": 5840
    },
    {
      "entropy": 6.421293783187866,
      "epoch": 0.6745527986151183,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004964578821942288,
      "loss": 6.2363,
      "mean_token_accuracy": 0.16685911267995834,
      "num_tokens": 14822704.0,
      "step": 5845
    },
    {
      "entropy": 6.395439910888672,
      "epoch": 0.6751298326601269,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004964505853348724,
      "loss": 6.2621,
      "mean_token_accuracy": 0.16632062196731567,
      "num_tokens": 14836913.0,
      "step": 5850
    },
    {
      "entropy": 6.315466976165771,
      "epoch": 0.6757068667051356,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004964432810271356,
      "loss": 6.2172,
      "mean_token_accuracy": 0.16847725361585617,
      "num_tokens": 14849697.0,
      "step": 5855
    },
    {
      "entropy": 6.328092336654663,
      "epoch": 0.6762839007501442,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004964359692712641,
      "loss": 6.173,
      "mean_token_accuracy": 0.1680448904633522,
      "num_tokens": 14861719.0,
      "step": 5860
    },
    {
      "entropy": 6.370050001144409,
      "epoch": 0.676860934795153,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004964286500675037,
      "loss": 6.1903,
      "mean_token_accuracy": 0.169232539832592,
      "num_tokens": 14873786.0,
      "step": 5865
    },
    {
      "entropy": 6.322551298141479,
      "epoch": 0.6774379688401616,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964213234161007,
      "loss": 6.1336,
      "mean_token_accuracy": 0.16598559767007828,
      "num_tokens": 14888952.0,
      "step": 5870
    },
    {
      "entropy": 6.440288496017456,
      "epoch": 0.6780150028851702,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004964139893173014,
      "loss": 6.2969,
      "mean_token_accuracy": 0.16432846188545228,
      "num_tokens": 14903039.0,
      "step": 5875
    },
    {
      "entropy": 6.347249984741211,
      "epoch": 0.6785920369301789,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964066477713525,
      "loss": 6.2428,
      "mean_token_accuracy": 0.15882514864206315,
      "num_tokens": 14915949.0,
      "step": 5880
    },
    {
      "entropy": 6.389187002182007,
      "epoch": 0.6791690709751875,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963992987785011,
      "loss": 6.309,
      "mean_token_accuracy": 0.16167141646146774,
      "num_tokens": 14928724.0,
      "step": 5885
    },
    {
      "entropy": 6.41411247253418,
      "epoch": 0.6797461050201962,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963919423389942,
      "loss": 6.2041,
      "mean_token_accuracy": 0.16787551641464232,
      "num_tokens": 14941759.0,
      "step": 5890
    },
    {
      "entropy": 6.321781730651855,
      "epoch": 0.6803231390652048,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963845784530794,
      "loss": 6.2576,
      "mean_token_accuracy": 0.16255476772785188,
      "num_tokens": 14954373.0,
      "step": 5895
    },
    {
      "entropy": 6.33484411239624,
      "epoch": 0.6809001731102136,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000496377207121004,
      "loss": 6.1296,
      "mean_token_accuracy": 0.17248332351446152,
      "num_tokens": 14966195.0,
      "step": 5900
    },
    {
      "entropy": 6.3584531307220455,
      "epoch": 0.6814772071552222,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004963698283430164,
      "loss": 6.229,
      "mean_token_accuracy": 0.16396655291318893,
      "num_tokens": 14979594.0,
      "step": 5905
    },
    {
      "entropy": 6.394085788726807,
      "epoch": 0.6820542412002308,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004963624421193646,
      "loss": 6.2991,
      "mean_token_accuracy": 0.16592729836702347,
      "num_tokens": 14993215.0,
      "step": 5910
    },
    {
      "entropy": 6.433213758468628,
      "epoch": 0.6826312752452395,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000496355048450297,
      "loss": 6.2525,
      "mean_token_accuracy": 0.1618594080209732,
      "num_tokens": 15006241.0,
      "step": 5915
    },
    {
      "entropy": 6.315669107437134,
      "epoch": 0.6832083092902481,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004963476473360623,
      "loss": 6.2196,
      "mean_token_accuracy": 0.1624471828341484,
      "num_tokens": 15018750.0,
      "step": 5920
    },
    {
      "entropy": 6.291262006759643,
      "epoch": 0.6837853433352568,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004963402387769094,
      "loss": 6.171,
      "mean_token_accuracy": 0.16801615953445434,
      "num_tokens": 15031065.0,
      "step": 5925
    },
    {
      "entropy": 6.385349893569947,
      "epoch": 0.6843623773802654,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004963328227730876,
      "loss": 6.1904,
      "mean_token_accuracy": 0.16905016154050828,
      "num_tokens": 15043129.0,
      "step": 5930
    },
    {
      "entropy": 6.352339839935302,
      "epoch": 0.684939411425274,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963253993248461,
      "loss": 6.2007,
      "mean_token_accuracy": 0.16816207468509675,
      "num_tokens": 15057531.0,
      "step": 5935
    },
    {
      "entropy": 6.342672061920166,
      "epoch": 0.6855164454702828,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004963179684324349,
      "loss": 6.2527,
      "mean_token_accuracy": 0.16352169662714006,
      "num_tokens": 15070105.0,
      "step": 5940
    },
    {
      "entropy": 6.41286940574646,
      "epoch": 0.6860934795152914,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963105300961036,
      "loss": 6.2282,
      "mean_token_accuracy": 0.15723423808813095,
      "num_tokens": 15082482.0,
      "step": 5945
    },
    {
      "entropy": 6.369506120681763,
      "epoch": 0.6866705135603001,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004963030843161026,
      "loss": 6.2664,
      "mean_token_accuracy": 0.15877436846494675,
      "num_tokens": 15095564.0,
      "step": 5950
    },
    {
      "entropy": 6.4327515125274655,
      "epoch": 0.6872475476053087,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004962956310926823,
      "loss": 6.2567,
      "mean_token_accuracy": 0.15945251435041427,
      "num_tokens": 15108948.0,
      "step": 5955
    },
    {
      "entropy": 6.3614583015441895,
      "epoch": 0.6878245816503173,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004962881704260934,
      "loss": 6.1733,
      "mean_token_accuracy": 0.16766972094774246,
      "num_tokens": 15121537.0,
      "step": 5960
    },
    {
      "entropy": 6.282821607589722,
      "epoch": 0.688401615695326,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004962807023165868,
      "loss": 6.1881,
      "mean_token_accuracy": 0.17159215807914735,
      "num_tokens": 15134657.0,
      "step": 5965
    },
    {
      "entropy": 6.340293598175049,
      "epoch": 0.6889786497403346,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004962732267644138,
      "loss": 6.1509,
      "mean_token_accuracy": 0.17035603374242783,
      "num_tokens": 15146301.0,
      "step": 5970
    },
    {
      "entropy": 6.32009449005127,
      "epoch": 0.6895556837853434,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004962657437698254,
      "loss": 6.2484,
      "mean_token_accuracy": 0.16358525454998016,
      "num_tokens": 15157886.0,
      "step": 5975
    },
    {
      "entropy": 6.3822681427001955,
      "epoch": 0.690132717830352,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004962582533330739,
      "loss": 6.1996,
      "mean_token_accuracy": 0.1663409873843193,
      "num_tokens": 15170457.0,
      "step": 5980
    },
    {
      "entropy": 6.350669097900391,
      "epoch": 0.6907097518753607,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004962507554544109,
      "loss": 6.2533,
      "mean_token_accuracy": 0.16517189443111419,
      "num_tokens": 15182504.0,
      "step": 5985
    },
    {
      "entropy": 6.3586745262146,
      "epoch": 0.6912867859203693,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962432501340886,
      "loss": 6.2435,
      "mean_token_accuracy": 0.1614229127764702,
      "num_tokens": 15194809.0,
      "step": 5990
    },
    {
      "entropy": 6.367116832733155,
      "epoch": 0.6918638199653779,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004962357373723596,
      "loss": 6.1836,
      "mean_token_accuracy": 0.1651404932141304,
      "num_tokens": 15207185.0,
      "step": 5995
    },
    {
      "entropy": 6.353478956222534,
      "epoch": 0.6924408540103866,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004962282171694763,
      "loss": 6.2033,
      "mean_token_accuracy": 0.16734539121389388,
      "num_tokens": 15220406.0,
      "step": 6000
    },
    {
      "epoch": 0.6924408540103866,
      "eval_entropy": 6.138688796956939,
      "eval_loss": 6.260073184967041,
      "eval_mean_token_accuracy": 0.16800092866038918,
      "eval_num_tokens": 15220406.0,
      "eval_runtime": 17.4454,
      "eval_samples_per_second": 1612.801,
      "eval_steps_per_second": 201.6,
      "step": 6000
    },
    {
      "entropy": 6.357237100601196,
      "epoch": 0.6930178880553952,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004962206895256919,
      "loss": 6.2433,
      "mean_token_accuracy": 0.1609627515077591,
      "num_tokens": 15231698.0,
      "step": 6005
    },
    {
      "entropy": 6.401620817184448,
      "epoch": 0.693594922100404,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004962131544412595,
      "loss": 6.2855,
      "mean_token_accuracy": 0.15989433825016022,
      "num_tokens": 15243462.0,
      "step": 6010
    },
    {
      "entropy": 6.4744181632995605,
      "epoch": 0.6941719561454126,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004962056119164325,
      "loss": 6.2977,
      "mean_token_accuracy": 0.1585573062300682,
      "num_tokens": 15257311.0,
      "step": 6015
    },
    {
      "entropy": 6.325224685668945,
      "epoch": 0.6947489901904212,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004961980619514646,
      "loss": 6.2524,
      "mean_token_accuracy": 0.1612841933965683,
      "num_tokens": 15270186.0,
      "step": 6020
    },
    {
      "entropy": 6.312858438491821,
      "epoch": 0.6953260242354299,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961905045466098,
      "loss": 6.2041,
      "mean_token_accuracy": 0.16199923306703568,
      "num_tokens": 15282573.0,
      "step": 6025
    },
    {
      "entropy": 6.426106882095337,
      "epoch": 0.6959030582804385,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0004961829397021222,
      "loss": 6.2437,
      "mean_token_accuracy": 0.1614992082118988,
      "num_tokens": 15296721.0,
      "step": 6030
    },
    {
      "entropy": 6.397392511367798,
      "epoch": 0.6964800923254472,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004961753674182564,
      "loss": 6.2139,
      "mean_token_accuracy": 0.1720279037952423,
      "num_tokens": 15308458.0,
      "step": 6035
    },
    {
      "entropy": 6.355577421188355,
      "epoch": 0.6970571263704558,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004961677876952669,
      "loss": 6.2178,
      "mean_token_accuracy": 0.1608239158987999,
      "num_tokens": 15320402.0,
      "step": 6040
    },
    {
      "entropy": 6.395686626434326,
      "epoch": 0.6976341604154646,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004961602005334087,
      "loss": 6.2689,
      "mean_token_accuracy": 0.16339593529701232,
      "num_tokens": 15333196.0,
      "step": 6045
    },
    {
      "entropy": 6.311012077331543,
      "epoch": 0.6982111944604732,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000496152605932937,
      "loss": 6.1947,
      "mean_token_accuracy": 0.16391687095165253,
      "num_tokens": 15345216.0,
      "step": 6050
    },
    {
      "entropy": 6.369989585876465,
      "epoch": 0.6987882285054818,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004961450038941074,
      "loss": 6.2674,
      "mean_token_accuracy": 0.16239935606718064,
      "num_tokens": 15357772.0,
      "step": 6055
    },
    {
      "entropy": 6.377553176879883,
      "epoch": 0.6993652625504905,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961373944171754,
      "loss": 6.1635,
      "mean_token_accuracy": 0.16831500232219695,
      "num_tokens": 15371110.0,
      "step": 6060
    },
    {
      "entropy": 6.268352222442627,
      "epoch": 0.6999422965954991,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961297775023968,
      "loss": 6.1729,
      "mean_token_accuracy": 0.16793664395809174,
      "num_tokens": 15383731.0,
      "step": 6065
    },
    {
      "entropy": 6.407302713394165,
      "epoch": 0.7005193306405078,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496122153150028,
      "loss": 6.2077,
      "mean_token_accuracy": 0.1660628244280815,
      "num_tokens": 15395849.0,
      "step": 6070
    },
    {
      "entropy": 6.384264755249023,
      "epoch": 0.7010963646855164,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004961145213603255,
      "loss": 6.1689,
      "mean_token_accuracy": 0.16858903616666793,
      "num_tokens": 15408552.0,
      "step": 6075
    },
    {
      "entropy": 6.338636875152588,
      "epoch": 0.7016733987305251,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000496106882133546,
      "loss": 6.2405,
      "mean_token_accuracy": 0.16671737134456635,
      "num_tokens": 15421940.0,
      "step": 6080
    },
    {
      "entropy": 6.291077041625977,
      "epoch": 0.7022504327755338,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004960992354699463,
      "loss": 6.2009,
      "mean_token_accuracy": 0.16258899569511415,
      "num_tokens": 15434232.0,
      "step": 6085
    },
    {
      "entropy": 6.414277458190918,
      "epoch": 0.7028274668205424,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004960915813697836,
      "loss": 6.2098,
      "mean_token_accuracy": 0.15860261023044586,
      "num_tokens": 15448052.0,
      "step": 6090
    },
    {
      "entropy": 6.392745780944824,
      "epoch": 0.7034045008655511,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004960839198333154,
      "loss": 6.1144,
      "mean_token_accuracy": 0.17049459218978882,
      "num_tokens": 15459141.0,
      "step": 6095
    },
    {
      "entropy": 6.240172338485718,
      "epoch": 0.7039815349105597,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004960762508607994,
      "loss": 6.0878,
      "mean_token_accuracy": 0.1716955214738846,
      "num_tokens": 15471705.0,
      "step": 6100
    },
    {
      "entropy": 6.309173583984375,
      "epoch": 0.7045585689555683,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004960685744524934,
      "loss": 6.1684,
      "mean_token_accuracy": 0.16721562147140503,
      "num_tokens": 15484821.0,
      "step": 6105
    },
    {
      "entropy": 6.386078023910523,
      "epoch": 0.705135603000577,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004960608906086558,
      "loss": 6.2208,
      "mean_token_accuracy": 0.16064056158065795,
      "num_tokens": 15497157.0,
      "step": 6110
    },
    {
      "entropy": 6.34077033996582,
      "epoch": 0.7057126370455857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496053199329545,
      "loss": 6.2138,
      "mean_token_accuracy": 0.16658349484205245,
      "num_tokens": 15508919.0,
      "step": 6115
    },
    {
      "entropy": 6.398518896102905,
      "epoch": 0.7062896710905944,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004960455006154196,
      "loss": 6.2674,
      "mean_token_accuracy": 0.16163421869277955,
      "num_tokens": 15521235.0,
      "step": 6120
    },
    {
      "entropy": 6.367052125930786,
      "epoch": 0.706866705135603,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004960377944665386,
      "loss": 6.2191,
      "mean_token_accuracy": 0.16501127034425736,
      "num_tokens": 15534278.0,
      "step": 6125
    },
    {
      "entropy": 6.294226884841919,
      "epoch": 0.7074437391806117,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004960300808831611,
      "loss": 6.182,
      "mean_token_accuracy": 0.1696803465485573,
      "num_tokens": 15548034.0,
      "step": 6130
    },
    {
      "entropy": 6.381236171722412,
      "epoch": 0.7080207732256203,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960223598655467,
      "loss": 6.1555,
      "mean_token_accuracy": 0.17080835551023482,
      "num_tokens": 15559425.0,
      "step": 6135
    },
    {
      "entropy": 6.433323335647583,
      "epoch": 0.7085978072706289,
      "grad_norm": 0.875,
      "learning_rate": 0.000496014631413955,
      "loss": 6.2619,
      "mean_token_accuracy": 0.1681198939681053,
      "num_tokens": 15570634.0,
      "step": 6140
    },
    {
      "entropy": 6.356976652145386,
      "epoch": 0.7091748413156376,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004960068955286459,
      "loss": 6.2906,
      "mean_token_accuracy": 0.16518030315637589,
      "num_tokens": 15582710.0,
      "step": 6145
    },
    {
      "entropy": 6.378293657302857,
      "epoch": 0.7097518753606463,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004959991522098797,
      "loss": 6.227,
      "mean_token_accuracy": 0.15896626859903334,
      "num_tokens": 15594798.0,
      "step": 6150
    },
    {
      "entropy": 6.364326047897339,
      "epoch": 0.710328909405655,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959914014579168,
      "loss": 6.1423,
      "mean_token_accuracy": 0.17040738612413406,
      "num_tokens": 15606482.0,
      "step": 6155
    },
    {
      "entropy": 6.359493207931519,
      "epoch": 0.7109059434506636,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004959836432730178,
      "loss": 6.1753,
      "mean_token_accuracy": 0.17176218777894975,
      "num_tokens": 15620178.0,
      "step": 6160
    },
    {
      "entropy": 6.3298866748809814,
      "epoch": 0.7114829774956722,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004959758776554438,
      "loss": 6.1991,
      "mean_token_accuracy": 0.16317809075117112,
      "num_tokens": 15632469.0,
      "step": 6165
    },
    {
      "entropy": 6.403311443328858,
      "epoch": 0.7120600115406809,
      "grad_norm": 0.7578125,
      "learning_rate": 0.000495968104605456,
      "loss": 6.2258,
      "mean_token_accuracy": 0.1647837668657303,
      "num_tokens": 15645610.0,
      "step": 6170
    },
    {
      "entropy": 6.398124265670776,
      "epoch": 0.7126370455856895,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004959603241233157,
      "loss": 6.2062,
      "mean_token_accuracy": 0.16843665093183519,
      "num_tokens": 15657716.0,
      "step": 6175
    },
    {
      "entropy": 6.339931488037109,
      "epoch": 0.7132140796306982,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959525362092846,
      "loss": 6.1924,
      "mean_token_accuracy": 0.16136354058980942,
      "num_tokens": 15670258.0,
      "step": 6180
    },
    {
      "entropy": 6.348710298538208,
      "epoch": 0.7137911136757069,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959447408636247,
      "loss": 6.1538,
      "mean_token_accuracy": 0.16794612407684326,
      "num_tokens": 15681196.0,
      "step": 6185
    },
    {
      "entropy": 6.416023778915405,
      "epoch": 0.7143681477207156,
      "grad_norm": 0.75,
      "learning_rate": 0.0004959369380865983,
      "loss": 6.2747,
      "mean_token_accuracy": 0.16282055974006654,
      "num_tokens": 15694591.0,
      "step": 6190
    },
    {
      "entropy": 6.293649387359619,
      "epoch": 0.7149451817657242,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959291278784676,
      "loss": 6.1255,
      "mean_token_accuracy": 0.17489954084157944,
      "num_tokens": 15708272.0,
      "step": 6195
    },
    {
      "entropy": 6.3302257537841795,
      "epoch": 0.7155222158107328,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959213102394954,
      "loss": 6.1724,
      "mean_token_accuracy": 0.172370782494545,
      "num_tokens": 15720136.0,
      "step": 6200
    },
    {
      "entropy": 6.36483941078186,
      "epoch": 0.7160992498557415,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004959134851699449,
      "loss": 6.1565,
      "mean_token_accuracy": 0.16822995990514755,
      "num_tokens": 15732043.0,
      "step": 6205
    },
    {
      "entropy": 6.360851812362671,
      "epoch": 0.7166762839007501,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959056526700788,
      "loss": 6.2137,
      "mean_token_accuracy": 0.16269091218709947,
      "num_tokens": 15743471.0,
      "step": 6210
    },
    {
      "entropy": 6.3516560077667235,
      "epoch": 0.7172533179457588,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004958978127401609,
      "loss": 6.2061,
      "mean_token_accuracy": 0.16105854660272598,
      "num_tokens": 15756846.0,
      "step": 6215
    },
    {
      "entropy": 6.33987021446228,
      "epoch": 0.7178303519907675,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004958899653804548,
      "loss": 6.2435,
      "mean_token_accuracy": 0.16606948673725128,
      "num_tokens": 15769993.0,
      "step": 6220
    },
    {
      "entropy": 6.313188934326172,
      "epoch": 0.7184073860357761,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958821105912246,
      "loss": 6.2005,
      "mean_token_accuracy": 0.16074076294898987,
      "num_tokens": 15782081.0,
      "step": 6225
    },
    {
      "entropy": 6.374634170532227,
      "epoch": 0.7189844200807848,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000495874248372734,
      "loss": 6.2334,
      "mean_token_accuracy": 0.16276226937770844,
      "num_tokens": 15793987.0,
      "step": 6230
    },
    {
      "entropy": 6.337182664871216,
      "epoch": 0.7195614541257934,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000495866378725248,
      "loss": 6.1536,
      "mean_token_accuracy": 0.17333306670188903,
      "num_tokens": 15805749.0,
      "step": 6235
    },
    {
      "entropy": 6.28895583152771,
      "epoch": 0.7201384881708021,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000495858501649031,
      "loss": 6.1259,
      "mean_token_accuracy": 0.16905360966920852,
      "num_tokens": 15818402.0,
      "step": 6240
    },
    {
      "entropy": 6.330921030044555,
      "epoch": 0.7207155222158107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000495850617144348,
      "loss": 6.1038,
      "mean_token_accuracy": 0.17487937062978745,
      "num_tokens": 15830783.0,
      "step": 6245
    },
    {
      "entropy": 6.27276611328125,
      "epoch": 0.7212925562608193,
      "grad_norm": 0.75,
      "learning_rate": 0.0004958427252114643,
      "loss": 6.1121,
      "mean_token_accuracy": 0.17082489728927613,
      "num_tokens": 15843231.0,
      "step": 6250
    },
    {
      "entropy": 6.2428213596344,
      "epoch": 0.7218695903058281,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000495834825850645,
      "loss": 6.1356,
      "mean_token_accuracy": 0.18071018308401107,
      "num_tokens": 15855770.0,
      "step": 6255
    },
    {
      "entropy": 6.374913215637207,
      "epoch": 0.7224466243508367,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004958269190621562,
      "loss": 6.217,
      "mean_token_accuracy": 0.16166453212499618,
      "num_tokens": 15867979.0,
      "step": 6260
    },
    {
      "entropy": 6.3626570224761965,
      "epoch": 0.7230236583958454,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958190048462637,
      "loss": 6.2482,
      "mean_token_accuracy": 0.16734726279973983,
      "num_tokens": 15880878.0,
      "step": 6265
    },
    {
      "entropy": 6.319369268417359,
      "epoch": 0.723600692440854,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004958110832032336,
      "loss": 6.1765,
      "mean_token_accuracy": 0.17133153825998307,
      "num_tokens": 15893599.0,
      "step": 6270
    },
    {
      "entropy": 6.27418155670166,
      "epoch": 0.7241777264858626,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004958031541333322,
      "loss": 6.1537,
      "mean_token_accuracy": 0.1716024547815323,
      "num_tokens": 15906566.0,
      "step": 6275
    },
    {
      "entropy": 6.375898981094361,
      "epoch": 0.7247547605308713,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004957952176368267,
      "loss": 6.1673,
      "mean_token_accuracy": 0.1673166960477829,
      "num_tokens": 15920269.0,
      "step": 6280
    },
    {
      "entropy": 6.341583919525147,
      "epoch": 0.7253317945758799,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004957872737139836,
      "loss": 6.2241,
      "mean_token_accuracy": 0.16209050118923188,
      "num_tokens": 15932559.0,
      "step": 6285
    },
    {
      "entropy": 6.335490465164185,
      "epoch": 0.7259088286208887,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004957793223650702,
      "loss": 6.1512,
      "mean_token_accuracy": 0.1637295514345169,
      "num_tokens": 15945554.0,
      "step": 6290
    },
    {
      "entropy": 6.312412452697754,
      "epoch": 0.7264858626658973,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000495771363590354,
      "loss": 6.0815,
      "mean_token_accuracy": 0.17415805906057358,
      "num_tokens": 15958012.0,
      "step": 6295
    },
    {
      "entropy": 6.353401374816895,
      "epoch": 0.727062896710906,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004957633973901027,
      "loss": 6.1846,
      "mean_token_accuracy": 0.16412369459867476,
      "num_tokens": 15971508.0,
      "step": 6300
    },
    {
      "entropy": 6.341835260391235,
      "epoch": 0.7276399307559146,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004957554237645841,
      "loss": 6.1554,
      "mean_token_accuracy": 0.16791654974222184,
      "num_tokens": 15983445.0,
      "step": 6305
    },
    {
      "entropy": 6.353136157989502,
      "epoch": 0.7282169648009232,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957474427140665,
      "loss": 6.1746,
      "mean_token_accuracy": 0.16680117100477218,
      "num_tokens": 15995936.0,
      "step": 6310
    },
    {
      "entropy": 6.2642029285430905,
      "epoch": 0.7287939988459319,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004957394542388182,
      "loss": 6.1553,
      "mean_token_accuracy": 0.17148637622594834,
      "num_tokens": 16008192.0,
      "step": 6315
    },
    {
      "entropy": 6.396103096008301,
      "epoch": 0.7293710328909405,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004957314583391082,
      "loss": 6.212,
      "mean_token_accuracy": 0.16379689127206803,
      "num_tokens": 16020585.0,
      "step": 6320
    },
    {
      "entropy": 6.400233268737793,
      "epoch": 0.7299480669359493,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004957234550152052,
      "loss": 6.2545,
      "mean_token_accuracy": 0.1623969241976738,
      "num_tokens": 16032627.0,
      "step": 6325
    },
    {
      "entropy": 6.392393779754639,
      "epoch": 0.7305251009809579,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004957154442673784,
      "loss": 6.1956,
      "mean_token_accuracy": 0.1705213487148285,
      "num_tokens": 16047196.0,
      "step": 6330
    },
    {
      "entropy": 6.317262935638428,
      "epoch": 0.7311021350259665,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004957074260958972,
      "loss": 6.1473,
      "mean_token_accuracy": 0.166448113322258,
      "num_tokens": 16059875.0,
      "step": 6335
    },
    {
      "entropy": 6.392600727081299,
      "epoch": 0.7316791690709752,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956994005010315,
      "loss": 6.2114,
      "mean_token_accuracy": 0.1701406717300415,
      "num_tokens": 16071991.0,
      "step": 6340
    },
    {
      "entropy": 6.372545003890991,
      "epoch": 0.7322562031159838,
      "grad_norm": 0.8125,
      "learning_rate": 0.000495691367483051,
      "loss": 6.2525,
      "mean_token_accuracy": 0.15873659104108812,
      "num_tokens": 16085628.0,
      "step": 6345
    },
    {
      "entropy": 6.351456260681152,
      "epoch": 0.7328332371609925,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004956833270422259,
      "loss": 6.2165,
      "mean_token_accuracy": 0.16475701630115508,
      "num_tokens": 16098020.0,
      "step": 6350
    },
    {
      "entropy": 6.362127590179443,
      "epoch": 0.7334102712060011,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004956752791788269,
      "loss": 6.1462,
      "mean_token_accuracy": 0.16634501665830612,
      "num_tokens": 16109965.0,
      "step": 6355
    },
    {
      "entropy": 6.373689317703247,
      "epoch": 0.7339873052510099,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004956672238931244,
      "loss": 6.216,
      "mean_token_accuracy": 0.16246710792183877,
      "num_tokens": 16121891.0,
      "step": 6360
    },
    {
      "entropy": 6.40177731513977,
      "epoch": 0.7345643392960185,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004956591611853894,
      "loss": 6.1647,
      "mean_token_accuracy": 0.16610097885131836,
      "num_tokens": 16134181.0,
      "step": 6365
    },
    {
      "entropy": 6.373047256469727,
      "epoch": 0.7351413733410271,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004956510910558931,
      "loss": 6.223,
      "mean_token_accuracy": 0.1654479667544365,
      "num_tokens": 16146685.0,
      "step": 6370
    },
    {
      "entropy": 6.26585111618042,
      "epoch": 0.7357184073860358,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004956430135049071,
      "loss": 6.1736,
      "mean_token_accuracy": 0.1662848874926567,
      "num_tokens": 16159618.0,
      "step": 6375
    },
    {
      "entropy": 6.392060470581055,
      "epoch": 0.7362954414310444,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004956349285327028,
      "loss": 6.1032,
      "mean_token_accuracy": 0.17501070201396943,
      "num_tokens": 16172162.0,
      "step": 6380
    },
    {
      "entropy": 6.330788850784302,
      "epoch": 0.7368724754760531,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956268361395523,
      "loss": 6.1916,
      "mean_token_accuracy": 0.16330249309539796,
      "num_tokens": 16184663.0,
      "step": 6385
    },
    {
      "entropy": 6.341895961761475,
      "epoch": 0.7374495095210617,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004956187363257278,
      "loss": 6.1909,
      "mean_token_accuracy": 0.16805584877729415,
      "num_tokens": 16196459.0,
      "step": 6390
    },
    {
      "entropy": 6.408393335342407,
      "epoch": 0.7380265435660704,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956106290915017,
      "loss": 6.2057,
      "mean_token_accuracy": 0.1656257450580597,
      "num_tokens": 16208008.0,
      "step": 6395
    },
    {
      "entropy": 6.361274671554566,
      "epoch": 0.7386035776110791,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004956025144371467,
      "loss": 6.2381,
      "mean_token_accuracy": 0.16646528393030166,
      "num_tokens": 16221844.0,
      "step": 6400
    },
    {
      "entropy": 6.344065237045288,
      "epoch": 0.7391806116560877,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004955943923629356,
      "loss": 6.2567,
      "mean_token_accuracy": 0.16411988735198973,
      "num_tokens": 16234773.0,
      "step": 6405
    },
    {
      "entropy": 6.274953842163086,
      "epoch": 0.7397576457010964,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955862628691417,
      "loss": 6.1179,
      "mean_token_accuracy": 0.173089899122715,
      "num_tokens": 16247464.0,
      "step": 6410
    },
    {
      "entropy": 6.447964525222778,
      "epoch": 0.740334679746105,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004955781259560386,
      "loss": 6.2003,
      "mean_token_accuracy": 0.1572584852576256,
      "num_tokens": 16259947.0,
      "step": 6415
    },
    {
      "entropy": 6.422522401809692,
      "epoch": 0.7409117137911136,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004955699816238995,
      "loss": 6.3076,
      "mean_token_accuracy": 0.1625203400850296,
      "num_tokens": 16272280.0,
      "step": 6420
    },
    {
      "entropy": 6.374295711517334,
      "epoch": 0.7414887478361223,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955618298729988,
      "loss": 6.1882,
      "mean_token_accuracy": 0.16504510641098022,
      "num_tokens": 16284476.0,
      "step": 6425
    },
    {
      "entropy": 6.312162399291992,
      "epoch": 0.742065781881131,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004955536707036105,
      "loss": 6.1115,
      "mean_token_accuracy": 0.16799768358469008,
      "num_tokens": 16296877.0,
      "step": 6430
    },
    {
      "entropy": 6.271811151504517,
      "epoch": 0.7426428159261397,
      "grad_norm": 0.74609375,
      "learning_rate": 0.000495545504116009,
      "loss": 6.1095,
      "mean_token_accuracy": 0.1740468755364418,
      "num_tokens": 16310360.0,
      "step": 6435
    },
    {
      "entropy": 6.435547924041748,
      "epoch": 0.7432198499711483,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004955373301104691,
      "loss": 6.27,
      "mean_token_accuracy": 0.15865596383810043,
      "num_tokens": 16322965.0,
      "step": 6440
    },
    {
      "entropy": 6.338442611694336,
      "epoch": 0.743796884016157,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004955291486872657,
      "loss": 6.2202,
      "mean_token_accuracy": 0.1622794598340988,
      "num_tokens": 16337756.0,
      "step": 6445
    },
    {
      "entropy": 6.373736047744751,
      "epoch": 0.7443739180611656,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004955209598466738,
      "loss": 6.1571,
      "mean_token_accuracy": 0.16572255790233612,
      "num_tokens": 16350400.0,
      "step": 6450
    },
    {
      "entropy": 6.42721381187439,
      "epoch": 0.7449509521061742,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000495512763588969,
      "loss": 6.2663,
      "mean_token_accuracy": 0.15915481746196747,
      "num_tokens": 16363655.0,
      "step": 6455
    },
    {
      "entropy": 6.382620811462402,
      "epoch": 0.7455279861511829,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004955045599144269,
      "loss": 6.1398,
      "mean_token_accuracy": 0.17301644384860992,
      "num_tokens": 16375580.0,
      "step": 6460
    },
    {
      "entropy": 6.357531595230102,
      "epoch": 0.7461050201961916,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004954963488233235,
      "loss": 6.2241,
      "mean_token_accuracy": 0.16709440499544143,
      "num_tokens": 16388447.0,
      "step": 6465
    },
    {
      "entropy": 6.347971868515015,
      "epoch": 0.7466820542412003,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000495488130315935,
      "loss": 6.1107,
      "mean_token_accuracy": 0.1755620315670967,
      "num_tokens": 16401082.0,
      "step": 6470
    },
    {
      "entropy": 6.280410528182983,
      "epoch": 0.7472590882862089,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004954799043925377,
      "loss": 6.148,
      "mean_token_accuracy": 0.16513099819421767,
      "num_tokens": 16415108.0,
      "step": 6475
    },
    {
      "entropy": 6.3362969875335695,
      "epoch": 0.7478361223312175,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004954716710534082,
      "loss": 6.0848,
      "mean_token_accuracy": 0.17339474856853485,
      "num_tokens": 16427562.0,
      "step": 6480
    },
    {
      "entropy": 6.31677737236023,
      "epoch": 0.7484131563762262,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004954634302988237,
      "loss": 6.1379,
      "mean_token_accuracy": 0.16748676151037217,
      "num_tokens": 16441282.0,
      "step": 6485
    },
    {
      "entropy": 6.341112232208252,
      "epoch": 0.7489901904212348,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004954551821290612,
      "loss": 6.1689,
      "mean_token_accuracy": 0.16641539484262466,
      "num_tokens": 16453433.0,
      "step": 6490
    },
    {
      "entropy": 6.341647577285767,
      "epoch": 0.7495672244662435,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004954469265443981,
      "loss": 6.1824,
      "mean_token_accuracy": 0.16345324069261552,
      "num_tokens": 16466216.0,
      "step": 6495
    },
    {
      "entropy": 6.353080177307129,
      "epoch": 0.7501442585112522,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004954386635451123,
      "loss": 6.1845,
      "mean_token_accuracy": 0.1658550977706909,
      "num_tokens": 16479046.0,
      "step": 6500
    },
    {
      "entropy": 6.29761233329773,
      "epoch": 0.7507212925562609,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954303931314814,
      "loss": 6.1422,
      "mean_token_accuracy": 0.1719191774725914,
      "num_tokens": 16491407.0,
      "step": 6505
    },
    {
      "entropy": 6.368516111373902,
      "epoch": 0.7512983266012695,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004954221153037837,
      "loss": 6.1414,
      "mean_token_accuracy": 0.1638320803642273,
      "num_tokens": 16504726.0,
      "step": 6510
    },
    {
      "entropy": 6.40753402709961,
      "epoch": 0.7518753606462781,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004954138300622978,
      "loss": 6.2249,
      "mean_token_accuracy": 0.16406668722629547,
      "num_tokens": 16517835.0,
      "step": 6515
    },
    {
      "entropy": 6.360999631881714,
      "epoch": 0.7524523946912868,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000495405537407302,
      "loss": 6.2334,
      "mean_token_accuracy": 0.1658797487616539,
      "num_tokens": 16532018.0,
      "step": 6520
    },
    {
      "entropy": 6.297806024551392,
      "epoch": 0.7530294287362954,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004953972373390755,
      "loss": 6.1618,
      "mean_token_accuracy": 0.17094711363315582,
      "num_tokens": 16546211.0,
      "step": 6525
    },
    {
      "entropy": 6.406245517730713,
      "epoch": 0.753606462781304,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953889298578975,
      "loss": 6.2121,
      "mean_token_accuracy": 0.16255346834659576,
      "num_tokens": 16559788.0,
      "step": 6530
    },
    {
      "entropy": 6.26466932296753,
      "epoch": 0.7541834968263128,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953806149640473,
      "loss": 6.0753,
      "mean_token_accuracy": 0.17620715349912644,
      "num_tokens": 16573694.0,
      "step": 6535
    },
    {
      "entropy": 6.351998710632325,
      "epoch": 0.7547605308713214,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953722926578045,
      "loss": 6.1705,
      "mean_token_accuracy": 0.16607727259397506,
      "num_tokens": 16585506.0,
      "step": 6540
    },
    {
      "entropy": 6.285213470458984,
      "epoch": 0.7553375649163301,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953639629394492,
      "loss": 6.2271,
      "mean_token_accuracy": 0.1645355686545372,
      "num_tokens": 16598991.0,
      "step": 6545
    },
    {
      "entropy": 6.40970253944397,
      "epoch": 0.7559145989613387,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004953556258092613,
      "loss": 6.227,
      "mean_token_accuracy": 0.1658835232257843,
      "num_tokens": 16612025.0,
      "step": 6550
    },
    {
      "entropy": 6.334483432769775,
      "epoch": 0.7564916330063474,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953472812675216,
      "loss": 6.1364,
      "mean_token_accuracy": 0.17058963626623153,
      "num_tokens": 16624509.0,
      "step": 6555
    },
    {
      "entropy": 6.275288057327271,
      "epoch": 0.757068667051356,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004953389293145104,
      "loss": 6.1552,
      "mean_token_accuracy": 0.16298312842845916,
      "num_tokens": 16636920.0,
      "step": 6560
    },
    {
      "entropy": 6.296596527099609,
      "epoch": 0.7576457010963646,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004953305699505088,
      "loss": 6.0867,
      "mean_token_accuracy": 0.17344891875982285,
      "num_tokens": 16649381.0,
      "step": 6565
    },
    {
      "entropy": 6.368966484069825,
      "epoch": 0.7582227351413734,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953222031757979,
      "loss": 6.1939,
      "mean_token_accuracy": 0.15837449729442596,
      "num_tokens": 16662273.0,
      "step": 6570
    },
    {
      "entropy": 6.372677612304687,
      "epoch": 0.758799769186382,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004953138289906592,
      "loss": 6.1871,
      "mean_token_accuracy": 0.17127318382263185,
      "num_tokens": 16675184.0,
      "step": 6575
    },
    {
      "entropy": 6.309050750732422,
      "epoch": 0.7593768032313907,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004953054473953742,
      "loss": 6.1602,
      "mean_token_accuracy": 0.1698559433221817,
      "num_tokens": 16687658.0,
      "step": 6580
    },
    {
      "entropy": 6.365961122512817,
      "epoch": 0.7599538372763993,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004952970583902251,
      "loss": 6.1433,
      "mean_token_accuracy": 0.1670696035027504,
      "num_tokens": 16699806.0,
      "step": 6585
    },
    {
      "entropy": 6.3373250484466555,
      "epoch": 0.760530871321408,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004952886619754937,
      "loss": 6.1879,
      "mean_token_accuracy": 0.17124574780464172,
      "num_tokens": 16712647.0,
      "step": 6590
    },
    {
      "entropy": 6.343024015426636,
      "epoch": 0.7611079053664166,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004952802581514625,
      "loss": 6.1422,
      "mean_token_accuracy": 0.16678574532270432,
      "num_tokens": 16725438.0,
      "step": 6595
    },
    {
      "entropy": 6.330097246170044,
      "epoch": 0.7616849394114252,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004952718469184144,
      "loss": 6.1548,
      "mean_token_accuracy": 0.17337974458932875,
      "num_tokens": 16737346.0,
      "step": 6600
    },
    {
      "entropy": 6.348900222778321,
      "epoch": 0.762261973456434,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004952634282766322,
      "loss": 6.1497,
      "mean_token_accuracy": 0.17005545794963836,
      "num_tokens": 16749327.0,
      "step": 6605
    },
    {
      "entropy": 6.330654907226562,
      "epoch": 0.7628390075014426,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004952550022263988,
      "loss": 6.0736,
      "mean_token_accuracy": 0.17179016321897506,
      "num_tokens": 16761323.0,
      "step": 6610
    },
    {
      "entropy": 6.349520587921143,
      "epoch": 0.7634160415464513,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004952465687679979,
      "loss": 6.254,
      "mean_token_accuracy": 0.16122461408376693,
      "num_tokens": 16774688.0,
      "step": 6615
    },
    {
      "entropy": 6.431204414367675,
      "epoch": 0.7639930755914599,
      "grad_norm": 0.77734375,
      "learning_rate": 0.000495238127901713,
      "loss": 6.2608,
      "mean_token_accuracy": 0.16095422506332396,
      "num_tokens": 16787538.0,
      "step": 6620
    },
    {
      "entropy": 6.314582109451294,
      "epoch": 0.7645701096364685,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952296796278282,
      "loss": 6.1283,
      "mean_token_accuracy": 0.17241402566432953,
      "num_tokens": 16798642.0,
      "step": 6625
    },
    {
      "entropy": 6.355658340454101,
      "epoch": 0.7651471436814772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952212239466274,
      "loss": 6.2645,
      "mean_token_accuracy": 0.1689286157488823,
      "num_tokens": 16810785.0,
      "step": 6630
    },
    {
      "entropy": 6.409767293930054,
      "epoch": 0.7657241777264858,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952127608583953,
      "loss": 6.2347,
      "mean_token_accuracy": 0.16363934576511383,
      "num_tokens": 16823978.0,
      "step": 6635
    },
    {
      "entropy": 6.358902454376221,
      "epoch": 0.7663012117714946,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952042903634162,
      "loss": 6.1651,
      "mean_token_accuracy": 0.17277514338493347,
      "num_tokens": 16836878.0,
      "step": 6640
    },
    {
      "entropy": 6.344875907897949,
      "epoch": 0.7668782458165032,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004951958124619752,
      "loss": 6.1771,
      "mean_token_accuracy": 0.17383748292922974,
      "num_tokens": 16849849.0,
      "step": 6645
    },
    {
      "entropy": 6.34850664138794,
      "epoch": 0.7674552798615119,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004951873271543573,
      "loss": 6.192,
      "mean_token_accuracy": 0.16859251111745835,
      "num_tokens": 16862022.0,
      "step": 6650
    },
    {
      "entropy": 6.438310003280639,
      "epoch": 0.7680323139065205,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951788344408483,
      "loss": 6.2359,
      "mean_token_accuracy": 0.1643815368413925,
      "num_tokens": 16875421.0,
      "step": 6655
    },
    {
      "entropy": 6.322825193405151,
      "epoch": 0.7686093479515291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004951703343217335,
      "loss": 6.1067,
      "mean_token_accuracy": 0.1841158390045166,
      "num_tokens": 16889520.0,
      "step": 6660
    },
    {
      "entropy": 6.2558657169342045,
      "epoch": 0.7691863819965378,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004951618267972987,
      "loss": 6.1407,
      "mean_token_accuracy": 0.1638886660337448,
      "num_tokens": 16901449.0,
      "step": 6665
    },
    {
      "entropy": 6.402849817276001,
      "epoch": 0.7697634160415464,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951533118678304,
      "loss": 6.2012,
      "mean_token_accuracy": 0.16906618624925612,
      "num_tokens": 16914742.0,
      "step": 6670
    },
    {
      "entropy": 6.444267988204956,
      "epoch": 0.7703404500865552,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951447895336147,
      "loss": 6.1846,
      "mean_token_accuracy": 0.16533467471599578,
      "num_tokens": 16926645.0,
      "step": 6675
    },
    {
      "entropy": 6.32177734375,
      "epoch": 0.7709174841315638,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004951362597949384,
      "loss": 6.1355,
      "mean_token_accuracy": 0.17496459186077118,
      "num_tokens": 16939499.0,
      "step": 6680
    },
    {
      "entropy": 6.331139373779297,
      "epoch": 0.7714945181765724,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004951277226520883,
      "loss": 6.1351,
      "mean_token_accuracy": 0.17526398450136185,
      "num_tokens": 16951811.0,
      "step": 6685
    },
    {
      "entropy": 6.298669147491455,
      "epoch": 0.7720715522215811,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004951191781053517,
      "loss": 6.1556,
      "mean_token_accuracy": 0.1653261348605156,
      "num_tokens": 16963126.0,
      "step": 6690
    },
    {
      "entropy": 6.357400894165039,
      "epoch": 0.7726485862665897,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004951106261550157,
      "loss": 6.1722,
      "mean_token_accuracy": 0.16687628626823425,
      "num_tokens": 16976853.0,
      "step": 6695
    },
    {
      "entropy": 6.332295179367065,
      "epoch": 0.7732256203115984,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004951020668013683,
      "loss": 6.1921,
      "mean_token_accuracy": 0.16649820208549498,
      "num_tokens": 16988720.0,
      "step": 6700
    },
    {
      "entropy": 6.289232921600342,
      "epoch": 0.773802654356607,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950935000446972,
      "loss": 6.0953,
      "mean_token_accuracy": 0.17749709784984588,
      "num_tokens": 17000880.0,
      "step": 6705
    },
    {
      "entropy": 6.218468427658081,
      "epoch": 0.7743796884016156,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004950849258852905,
      "loss": 5.975,
      "mean_token_accuracy": 0.18584768772125243,
      "num_tokens": 17012540.0,
      "step": 6710
    },
    {
      "entropy": 6.276014614105224,
      "epoch": 0.7749567224466244,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004950763443234366,
      "loss": 6.1089,
      "mean_token_accuracy": 0.17434907257556914,
      "num_tokens": 17024163.0,
      "step": 6715
    },
    {
      "entropy": 6.373161888122558,
      "epoch": 0.775533756491633,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004950677553594243,
      "loss": 6.1703,
      "mean_token_accuracy": 0.16294383704662324,
      "num_tokens": 17037020.0,
      "step": 6720
    },
    {
      "entropy": 6.388097143173217,
      "epoch": 0.7761107905366417,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950591589935422,
      "loss": 6.1987,
      "mean_token_accuracy": 0.16396149396896362,
      "num_tokens": 17050132.0,
      "step": 6725
    },
    {
      "entropy": 6.318562936782837,
      "epoch": 0.7766878245816503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950505552260798,
      "loss": 6.131,
      "mean_token_accuracy": 0.1698935002088547,
      "num_tokens": 17063299.0,
      "step": 6730
    },
    {
      "entropy": 6.272518539428711,
      "epoch": 0.777264858626659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950419440573261,
      "loss": 6.1875,
      "mean_token_accuracy": 0.16989706754684447,
      "num_tokens": 17076312.0,
      "step": 6735
    },
    {
      "entropy": 6.288555431365967,
      "epoch": 0.7778418926716676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000495033325487571,
      "loss": 6.1588,
      "mean_token_accuracy": 0.16710376292467116,
      "num_tokens": 17089040.0,
      "step": 6740
    },
    {
      "entropy": 6.420559310913086,
      "epoch": 0.7784189267166762,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004950246995171042,
      "loss": 6.2479,
      "mean_token_accuracy": 0.16628091037273407,
      "num_tokens": 17102806.0,
      "step": 6745
    },
    {
      "entropy": 6.293579626083374,
      "epoch": 0.778995960761685,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000495016066146216,
      "loss": 6.0873,
      "mean_token_accuracy": 0.17714409083127974,
      "num_tokens": 17114782.0,
      "step": 6750
    },
    {
      "entropy": 6.290708208084107,
      "epoch": 0.7795729948066936,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004950074253751968,
      "loss": 6.1313,
      "mean_token_accuracy": 0.17414961010217667,
      "num_tokens": 17127386.0,
      "step": 6755
    },
    {
      "entropy": 6.3170980453491214,
      "epoch": 0.7801500288517023,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494998777204337,
      "loss": 6.1204,
      "mean_token_accuracy": 0.17070301622152328,
      "num_tokens": 17139244.0,
      "step": 6760
    },
    {
      "entropy": 6.375906133651734,
      "epoch": 0.7807270628967109,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949901216339276,
      "loss": 6.238,
      "mean_token_accuracy": 0.16011236011981964,
      "num_tokens": 17153136.0,
      "step": 6765
    },
    {
      "entropy": 6.350481653213501,
      "epoch": 0.7813040969417195,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949814586642598,
      "loss": 6.1064,
      "mean_token_accuracy": 0.16930470615625381,
      "num_tokens": 17166420.0,
      "step": 6770
    },
    {
      "entropy": 6.382876300811768,
      "epoch": 0.7818811309867282,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000494972788295625,
      "loss": 6.2473,
      "mean_token_accuracy": 0.1569085404276848,
      "num_tokens": 17180796.0,
      "step": 6775
    },
    {
      "entropy": 6.390883731842041,
      "epoch": 0.7824581650317368,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004949641105283144,
      "loss": 6.1946,
      "mean_token_accuracy": 0.16438286751508713,
      "num_tokens": 17192275.0,
      "step": 6780
    },
    {
      "entropy": 6.299509429931641,
      "epoch": 0.7830351990767456,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004949554253626205,
      "loss": 6.1623,
      "mean_token_accuracy": 0.16779853105545045,
      "num_tokens": 17203651.0,
      "step": 6785
    },
    {
      "entropy": 6.4158844470977785,
      "epoch": 0.7836122331217542,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004949467327988351,
      "loss": 6.2109,
      "mean_token_accuracy": 0.1698895961046219,
      "num_tokens": 17216521.0,
      "step": 6790
    },
    {
      "entropy": 6.3239830493927,
      "epoch": 0.7841892671667628,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004949380328372505,
      "loss": 6.1397,
      "mean_token_accuracy": 0.17289299368858338,
      "num_tokens": 17228435.0,
      "step": 6795
    },
    {
      "entropy": 6.310701036453247,
      "epoch": 0.7847663012117715,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004949293254781595,
      "loss": 6.0864,
      "mean_token_accuracy": 0.1724423810839653,
      "num_tokens": 17240746.0,
      "step": 6800
    },
    {
      "entropy": 6.214282369613647,
      "epoch": 0.7853433352567801,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004949206107218547,
      "loss": 6.0893,
      "mean_token_accuracy": 0.17622702717781066,
      "num_tokens": 17254089.0,
      "step": 6805
    },
    {
      "entropy": 6.303280830383301,
      "epoch": 0.7859203693017888,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004949118885686296,
      "loss": 6.1072,
      "mean_token_accuracy": 0.17078900188207627,
      "num_tokens": 17266389.0,
      "step": 6810
    },
    {
      "entropy": 6.357765245437622,
      "epoch": 0.7864974033467974,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004949031590187774,
      "loss": 6.1259,
      "mean_token_accuracy": 0.16939014941453934,
      "num_tokens": 17279049.0,
      "step": 6815
    },
    {
      "entropy": 6.336352300643921,
      "epoch": 0.7870744373918062,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004948944220725915,
      "loss": 6.1955,
      "mean_token_accuracy": 0.16172609031200408,
      "num_tokens": 17291134.0,
      "step": 6820
    },
    {
      "entropy": 6.3349395275115965,
      "epoch": 0.7876514714368148,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000494885677730366,
      "loss": 6.0774,
      "mean_token_accuracy": 0.17785610109567643,
      "num_tokens": 17304728.0,
      "step": 6825
    },
    {
      "entropy": 6.35065221786499,
      "epoch": 0.7882285054818234,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000494876925992395,
      "loss": 6.1665,
      "mean_token_accuracy": 0.1673291355371475,
      "num_tokens": 17317392.0,
      "step": 6830
    },
    {
      "entropy": 6.2910408020019535,
      "epoch": 0.7888055395268321,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004948681668589728,
      "loss": 6.1304,
      "mean_token_accuracy": 0.17116876244544982,
      "num_tokens": 17329639.0,
      "step": 6835
    },
    {
      "entropy": 6.320131015777588,
      "epoch": 0.7893825735718407,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494859400330394,
      "loss": 6.1502,
      "mean_token_accuracy": 0.17031224071979523,
      "num_tokens": 17341246.0,
      "step": 6840
    },
    {
      "entropy": 6.345110988616943,
      "epoch": 0.7899596076168494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948506264069535,
      "loss": 6.1563,
      "mean_token_accuracy": 0.1687989354133606,
      "num_tokens": 17353699.0,
      "step": 6845
    },
    {
      "entropy": 6.347292232513428,
      "epoch": 0.790536641661858,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004948418450889464,
      "loss": 6.1477,
      "mean_token_accuracy": 0.1734479546546936,
      "num_tokens": 17365950.0,
      "step": 6850
    },
    {
      "entropy": 6.2546172618865965,
      "epoch": 0.7911136757068667,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000494833056376668,
      "loss": 6.0821,
      "mean_token_accuracy": 0.17194298207759856,
      "num_tokens": 17378654.0,
      "step": 6855
    },
    {
      "entropy": 6.282538604736328,
      "epoch": 0.7916907097518754,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004948242602704139,
      "loss": 6.1543,
      "mean_token_accuracy": 0.17186086624860764,
      "num_tokens": 17391354.0,
      "step": 6860
    },
    {
      "entropy": 6.336555099487304,
      "epoch": 0.792267743796884,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004948154567704801,
      "loss": 6.1623,
      "mean_token_accuracy": 0.17139652073383332,
      "num_tokens": 17403019.0,
      "step": 6865
    },
    {
      "entropy": 6.305984163284302,
      "epoch": 0.7928447778418927,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948066458771625,
      "loss": 6.1636,
      "mean_token_accuracy": 0.17206017225980758,
      "num_tokens": 17415886.0,
      "step": 6870
    },
    {
      "entropy": 6.384555387496948,
      "epoch": 0.7934218118869013,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004947978275907577,
      "loss": 6.272,
      "mean_token_accuracy": 0.1580319046974182,
      "num_tokens": 17429734.0,
      "step": 6875
    },
    {
      "entropy": 6.387654590606689,
      "epoch": 0.79399884593191,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494789001911562,
      "loss": 6.1033,
      "mean_token_accuracy": 0.16667921841144562,
      "num_tokens": 17441749.0,
      "step": 6880
    },
    {
      "entropy": 6.392041492462158,
      "epoch": 0.7945758799769186,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004947801688398725,
      "loss": 6.1919,
      "mean_token_accuracy": 0.16391475200653077,
      "num_tokens": 17453626.0,
      "step": 6885
    },
    {
      "entropy": 6.289467096328735,
      "epoch": 0.7951529140219273,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004947713283759862,
      "loss": 6.1422,
      "mean_token_accuracy": 0.17084853053092958,
      "num_tokens": 17465038.0,
      "step": 6890
    },
    {
      "entropy": 6.32017035484314,
      "epoch": 0.795729948066936,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004947624805202003,
      "loss": 6.2005,
      "mean_token_accuracy": 0.15855661481618882,
      "num_tokens": 17478319.0,
      "step": 6895
    },
    {
      "entropy": 6.3642435550689695,
      "epoch": 0.7963069821119446,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947536252728126,
      "loss": 6.0826,
      "mean_token_accuracy": 0.17405088394880294,
      "num_tokens": 17491534.0,
      "step": 6900
    },
    {
      "entropy": 6.256607437133789,
      "epoch": 0.7968840161569533,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004947447626341209,
      "loss": 6.1176,
      "mean_token_accuracy": 0.16702562868595122,
      "num_tokens": 17504957.0,
      "step": 6905
    },
    {
      "entropy": 6.329520893096924,
      "epoch": 0.7974610502019619,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004947358926044232,
      "loss": 6.1307,
      "mean_token_accuracy": 0.1736439809203148,
      "num_tokens": 17518265.0,
      "step": 6910
    },
    {
      "entropy": 6.285739421844482,
      "epoch": 0.7980380842469705,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947270151840179,
      "loss": 6.1042,
      "mean_token_accuracy": 0.17833034843206405,
      "num_tokens": 17530588.0,
      "step": 6915
    },
    {
      "entropy": 6.200070571899414,
      "epoch": 0.7986151182919792,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004947181303732038,
      "loss": 6.0645,
      "mean_token_accuracy": 0.1750258818268776,
      "num_tokens": 17545429.0,
      "step": 6920
    },
    {
      "entropy": 6.368551015853882,
      "epoch": 0.7991921523369879,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947092381722793,
      "loss": 6.1731,
      "mean_token_accuracy": 0.16941929012537002,
      "num_tokens": 17558437.0,
      "step": 6925
    },
    {
      "entropy": 6.3232728958129885,
      "epoch": 0.7997691863819966,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004947003385815438,
      "loss": 6.122,
      "mean_token_accuracy": 0.16462525725364685,
      "num_tokens": 17569863.0,
      "step": 6930
    },
    {
      "entropy": 6.32643575668335,
      "epoch": 0.8003462204270052,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946914316012964,
      "loss": 6.202,
      "mean_token_accuracy": 0.16439114809036254,
      "num_tokens": 17581987.0,
      "step": 6935
    },
    {
      "entropy": 6.300495767593384,
      "epoch": 0.8009232544720138,
      "grad_norm": 0.765625,
      "learning_rate": 0.000494682517231837,
      "loss": 6.1986,
      "mean_token_accuracy": 0.16613231599330902,
      "num_tokens": 17594064.0,
      "step": 6940
    },
    {
      "entropy": 6.307906436920166,
      "epoch": 0.8015002885170225,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004946735954734652,
      "loss": 6.0737,
      "mean_token_accuracy": 0.16881750375032425,
      "num_tokens": 17606557.0,
      "step": 6945
    },
    {
      "entropy": 6.345771360397339,
      "epoch": 0.8020773225620311,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946646663264811,
      "loss": 6.1838,
      "mean_token_accuracy": 0.16938254088163376,
      "num_tokens": 17620095.0,
      "step": 6950
    },
    {
      "entropy": 6.280622100830078,
      "epoch": 0.8026543566070398,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946557297911852,
      "loss": 6.0541,
      "mean_token_accuracy": 0.17692890465259553,
      "num_tokens": 17633006.0,
      "step": 6955
    },
    {
      "entropy": 6.313730478286743,
      "epoch": 0.8032313906520485,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004946467858678777,
      "loss": 6.1475,
      "mean_token_accuracy": 0.164541295170784,
      "num_tokens": 17646863.0,
      "step": 6960
    },
    {
      "entropy": 6.352252578735351,
      "epoch": 0.8038084246970572,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004946378345568597,
      "loss": 6.1487,
      "mean_token_accuracy": 0.16321584284305574,
      "num_tokens": 17659917.0,
      "step": 6965
    },
    {
      "entropy": 6.360219669342041,
      "epoch": 0.8043854587420658,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004946288758584324,
      "loss": 6.1667,
      "mean_token_accuracy": 0.16748237907886504,
      "num_tokens": 17672867.0,
      "step": 6970
    },
    {
      "entropy": 6.3579973697662355,
      "epoch": 0.8049624927870744,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946199097728968,
      "loss": 6.1704,
      "mean_token_accuracy": 0.17079340666532516,
      "num_tokens": 17684600.0,
      "step": 6975
    },
    {
      "entropy": 6.339913034439087,
      "epoch": 0.8055395268320831,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004946109363005548,
      "loss": 6.1072,
      "mean_token_accuracy": 0.17065697759389878,
      "num_tokens": 17697023.0,
      "step": 6980
    },
    {
      "entropy": 6.322485399246216,
      "epoch": 0.8061165608770917,
      "grad_norm": 0.828125,
      "learning_rate": 0.000494601955441708,
      "loss": 6.1687,
      "mean_token_accuracy": 0.16757836192846298,
      "num_tokens": 17711592.0,
      "step": 6985
    },
    {
      "entropy": 6.416100358963012,
      "epoch": 0.8066935949221004,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945929671966585,
      "loss": 6.1653,
      "mean_token_accuracy": 0.1698947876691818,
      "num_tokens": 17723039.0,
      "step": 6990
    },
    {
      "entropy": 6.306565809249878,
      "epoch": 0.8072706289671091,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004945839715657085,
      "loss": 6.1177,
      "mean_token_accuracy": 0.16805417239665985,
      "num_tokens": 17736851.0,
      "step": 6995
    },
    {
      "entropy": 6.391094207763672,
      "epoch": 0.8078476630121177,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004945749685491607,
      "loss": 6.1555,
      "mean_token_accuracy": 0.1691310867667198,
      "num_tokens": 17749483.0,
      "step": 7000
    },
    {
      "entropy": 6.339906883239746,
      "epoch": 0.8084246970571264,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004945659581473181,
      "loss": 6.1816,
      "mean_token_accuracy": 0.16059267669916152,
      "num_tokens": 17760463.0,
      "step": 7005
    },
    {
      "entropy": 6.434919738769532,
      "epoch": 0.809001731102135,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945569403604833,
      "loss": 6.2524,
      "mean_token_accuracy": 0.16289447844028473,
      "num_tokens": 17772557.0,
      "step": 7010
    },
    {
      "entropy": 6.290898752212525,
      "epoch": 0.8095787651471437,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004945479151889601,
      "loss": 6.1405,
      "mean_token_accuracy": 0.17295335680246354,
      "num_tokens": 17785399.0,
      "step": 7015
    },
    {
      "entropy": 6.355238962173462,
      "epoch": 0.8101557991921523,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004945388826330517,
      "loss": 6.1667,
      "mean_token_accuracy": 0.16519918888807297,
      "num_tokens": 17798766.0,
      "step": 7020
    },
    {
      "entropy": 6.384207105636596,
      "epoch": 0.8107328332371609,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004945298426930621,
      "loss": 6.1533,
      "mean_token_accuracy": 0.16614916175603867,
      "num_tokens": 17811726.0,
      "step": 7025
    },
    {
      "entropy": 6.365244054794312,
      "epoch": 0.8113098672821697,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945207953692952,
      "loss": 6.0994,
      "mean_token_accuracy": 0.17219052463769913,
      "num_tokens": 17824142.0,
      "step": 7030
    },
    {
      "entropy": 6.322742366790772,
      "epoch": 0.8118869013271783,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004945117406620556,
      "loss": 6.1241,
      "mean_token_accuracy": 0.1732164204120636,
      "num_tokens": 17837497.0,
      "step": 7035
    },
    {
      "entropy": 6.291933012008667,
      "epoch": 0.812463935372187,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945026785716474,
      "loss": 6.1596,
      "mean_token_accuracy": 0.16709503829479216,
      "num_tokens": 17849186.0,
      "step": 7040
    },
    {
      "entropy": 6.336631917953492,
      "epoch": 0.8130409694171956,
      "grad_norm": 0.875,
      "learning_rate": 0.0004944936090983757,
      "loss": 6.182,
      "mean_token_accuracy": 0.16816715002059937,
      "num_tokens": 17861460.0,
      "step": 7045
    },
    {
      "entropy": 6.32755856513977,
      "epoch": 0.8136180034622043,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004944845322425455,
      "loss": 6.0981,
      "mean_token_accuracy": 0.17414549440145494,
      "num_tokens": 17874031.0,
      "step": 7050
    },
    {
      "entropy": 6.277894449234009,
      "epoch": 0.8141950375072129,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004944754480044621,
      "loss": 6.1176,
      "mean_token_accuracy": 0.16790224760770797,
      "num_tokens": 17886623.0,
      "step": 7055
    },
    {
      "entropy": 6.33378119468689,
      "epoch": 0.8147720715522215,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944663563844311,
      "loss": 6.1167,
      "mean_token_accuracy": 0.16459549963474274,
      "num_tokens": 17899888.0,
      "step": 7060
    },
    {
      "entropy": 6.300360584259034,
      "epoch": 0.8153491055972303,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004944572573827581,
      "loss": 6.1425,
      "mean_token_accuracy": 0.17327246218919753,
      "num_tokens": 17912089.0,
      "step": 7065
    },
    {
      "entropy": 6.34687385559082,
      "epoch": 0.8159261396422389,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004944481509997494,
      "loss": 6.2222,
      "mean_token_accuracy": 0.16654876470565796,
      "num_tokens": 17923790.0,
      "step": 7070
    },
    {
      "entropy": 6.400413751602173,
      "epoch": 0.8165031736872476,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000494439037235711,
      "loss": 6.154,
      "mean_token_accuracy": 0.17073314338922502,
      "num_tokens": 17936537.0,
      "step": 7075
    },
    {
      "entropy": 6.303481578826904,
      "epoch": 0.8170802077322562,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004944299160909495,
      "loss": 6.1222,
      "mean_token_accuracy": 0.1658458635210991,
      "num_tokens": 17948230.0,
      "step": 7080
    },
    {
      "entropy": 6.333654165267944,
      "epoch": 0.8176572417772648,
      "grad_norm": 0.796875,
      "learning_rate": 0.000494420787565772,
      "loss": 6.1076,
      "mean_token_accuracy": 0.17086146622896195,
      "num_tokens": 17960927.0,
      "step": 7085
    },
    {
      "entropy": 6.279893064498902,
      "epoch": 0.8182342758222735,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004944116516604852,
      "loss": 6.0779,
      "mean_token_accuracy": 0.17715319991111755,
      "num_tokens": 17974392.0,
      "step": 7090
    },
    {
      "entropy": 6.219158601760864,
      "epoch": 0.8188113098672821,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004944025083753965,
      "loss": 6.0373,
      "mean_token_accuracy": 0.17926018238067626,
      "num_tokens": 17987845.0,
      "step": 7095
    },
    {
      "entropy": 6.329027509689331,
      "epoch": 0.8193883439122909,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004943933577108133,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16645580977201463,
      "num_tokens": 18000610.0,
      "step": 7100
    },
    {
      "entropy": 6.299327182769775,
      "epoch": 0.8199653779572995,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004943841996670436,
      "loss": 6.0993,
      "mean_token_accuracy": 0.1676725208759308,
      "num_tokens": 18014322.0,
      "step": 7105
    },
    {
      "entropy": 6.305058240890503,
      "epoch": 0.8205424120023082,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004943750342443953,
      "loss": 6.1453,
      "mean_token_accuracy": 0.16816196143627166,
      "num_tokens": 18028072.0,
      "step": 7110
    },
    {
      "entropy": 6.335258913040161,
      "epoch": 0.8211194460473168,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004943658614431767,
      "loss": 6.1131,
      "mean_token_accuracy": 0.167927385866642,
      "num_tokens": 18039629.0,
      "step": 7115
    },
    {
      "entropy": 6.383225202560425,
      "epoch": 0.8216964800923254,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004943566812636963,
      "loss": 6.1576,
      "mean_token_accuracy": 0.17481892108917235,
      "num_tokens": 18052879.0,
      "step": 7120
    },
    {
      "entropy": 6.307858800888061,
      "epoch": 0.8222735141373341,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000494347493706263,
      "loss": 6.1765,
      "mean_token_accuracy": 0.1710774213075638,
      "num_tokens": 18065037.0,
      "step": 7125
    },
    {
      "entropy": 6.304836845397949,
      "epoch": 0.8228505481823427,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004943382987711856,
      "loss": 6.1087,
      "mean_token_accuracy": 0.17677446603775024,
      "num_tokens": 18078595.0,
      "step": 7130
    },
    {
      "entropy": 6.215980768203735,
      "epoch": 0.8234275822273515,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004943290964587734,
      "loss": 5.9584,
      "mean_token_accuracy": 0.1821457639336586,
      "num_tokens": 18089790.0,
      "step": 7135
    },
    {
      "entropy": 6.2496058464050295,
      "epoch": 0.8240046162723601,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004943198867693361,
      "loss": 6.0432,
      "mean_token_accuracy": 0.1753086805343628,
      "num_tokens": 18101343.0,
      "step": 7140
    },
    {
      "entropy": 6.350470972061157,
      "epoch": 0.8245816503173687,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004943106697031833,
      "loss": 6.1909,
      "mean_token_accuracy": 0.16834626346826553,
      "num_tokens": 18113124.0,
      "step": 7145
    },
    {
      "entropy": 6.328378009796142,
      "epoch": 0.8251586843623774,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004943014452606251,
      "loss": 6.145,
      "mean_token_accuracy": 0.17023891508579253,
      "num_tokens": 18125322.0,
      "step": 7150
    },
    {
      "entropy": 6.3475916385650635,
      "epoch": 0.825735718407386,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004942922134419717,
      "loss": 6.1932,
      "mean_token_accuracy": 0.16659340858459473,
      "num_tokens": 18137975.0,
      "step": 7155
    },
    {
      "entropy": 6.376634693145752,
      "epoch": 0.8263127524523947,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004942829742475336,
      "loss": 6.1566,
      "mean_token_accuracy": 0.16527699828147888,
      "num_tokens": 18151149.0,
      "step": 7160
    },
    {
      "entropy": 6.35319972038269,
      "epoch": 0.8268897864974033,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942737276776217,
      "loss": 6.1414,
      "mean_token_accuracy": 0.16284787356853486,
      "num_tokens": 18164451.0,
      "step": 7165
    },
    {
      "entropy": 6.271916341781616,
      "epoch": 0.827466820542412,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004942644737325468,
      "loss": 6.1039,
      "mean_token_accuracy": 0.16889770179986954,
      "num_tokens": 18177148.0,
      "step": 7170
    },
    {
      "entropy": 6.3122375965118405,
      "epoch": 0.8280438545874207,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004942552124126202,
      "loss": 6.1291,
      "mean_token_accuracy": 0.16649234741926194,
      "num_tokens": 18189428.0,
      "step": 7175
    },
    {
      "entropy": 6.342439794540406,
      "epoch": 0.8286208886324293,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004942459437181535,
      "loss": 6.0994,
      "mean_token_accuracy": 0.16976441591978073,
      "num_tokens": 18201644.0,
      "step": 7180
    },
    {
      "entropy": 6.326513719558716,
      "epoch": 0.829197922677438,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942366676494584,
      "loss": 6.1313,
      "mean_token_accuracy": 0.1665779948234558,
      "num_tokens": 18213366.0,
      "step": 7185
    },
    {
      "entropy": 6.272335004806519,
      "epoch": 0.8297749567224466,
      "grad_norm": 0.875,
      "learning_rate": 0.0004942273842068469,
      "loss": 6.1424,
      "mean_token_accuracy": 0.17365220338106155,
      "num_tokens": 18226767.0,
      "step": 7190
    },
    {
      "entropy": 6.336211061477661,
      "epoch": 0.8303519907674553,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004942180933906311,
      "loss": 6.0673,
      "mean_token_accuracy": 0.1732963427901268,
      "num_tokens": 18239250.0,
      "step": 7195
    },
    {
      "entropy": 6.232381200790405,
      "epoch": 0.8309290248124639,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004942087952011237,
      "loss": 6.1186,
      "mean_token_accuracy": 0.1715511605143547,
      "num_tokens": 18252053.0,
      "step": 7200
    },
    {
      "entropy": 6.319523096084595,
      "epoch": 0.8315060588574726,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004941994896386374,
      "loss": 6.1577,
      "mean_token_accuracy": 0.1603941097855568,
      "num_tokens": 18264769.0,
      "step": 7205
    },
    {
      "entropy": 6.390555953979492,
      "epoch": 0.8320830929024813,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004941901767034851,
      "loss": 6.1828,
      "mean_token_accuracy": 0.16464990973472596,
      "num_tokens": 18278185.0,
      "step": 7210
    },
    {
      "entropy": 6.283919048309326,
      "epoch": 0.8326601269474899,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941808563959802,
      "loss": 6.0769,
      "mean_token_accuracy": 0.17388446182012557,
      "num_tokens": 18290997.0,
      "step": 7215
    },
    {
      "entropy": 6.226381444931031,
      "epoch": 0.8332371609924986,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004941715287164359,
      "loss": 6.0793,
      "mean_token_accuracy": 0.17051244676113128,
      "num_tokens": 18303735.0,
      "step": 7220
    },
    {
      "entropy": 6.224194622039795,
      "epoch": 0.8338141950375072,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941621936651661,
      "loss": 6.0361,
      "mean_token_accuracy": 0.1779409721493721,
      "num_tokens": 18316061.0,
      "step": 7225
    },
    {
      "entropy": 6.325501871109009,
      "epoch": 0.8343912290825158,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004941528512424849,
      "loss": 6.0813,
      "mean_token_accuracy": 0.1724706158041954,
      "num_tokens": 18329478.0,
      "step": 7230
    },
    {
      "entropy": 6.277244186401367,
      "epoch": 0.8349682631275245,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941435014487064,
      "loss": 6.0834,
      "mean_token_accuracy": 0.1694391682744026,
      "num_tokens": 18341790.0,
      "step": 7235
    },
    {
      "entropy": 6.257685041427612,
      "epoch": 0.8355452971725332,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000494134144284145,
      "loss": 6.0609,
      "mean_token_accuracy": 0.17542690485715867,
      "num_tokens": 18353807.0,
      "step": 7240
    },
    {
      "entropy": 6.326960563659668,
      "epoch": 0.8361223312175419,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004941247797491156,
      "loss": 6.138,
      "mean_token_accuracy": 0.16580530852079392,
      "num_tokens": 18366118.0,
      "step": 7245
    },
    {
      "entropy": 6.224525594711304,
      "epoch": 0.8366993652625505,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004941154078439329,
      "loss": 6.0992,
      "mean_token_accuracy": 0.17626895606517792,
      "num_tokens": 18380542.0,
      "step": 7250
    },
    {
      "entropy": 6.379341077804566,
      "epoch": 0.8372763993075591,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004941060285689126,
      "loss": 6.1702,
      "mean_token_accuracy": 0.1661502555012703,
      "num_tokens": 18393976.0,
      "step": 7255
    },
    {
      "entropy": 6.262735748291016,
      "epoch": 0.8378534333525678,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004940966419243695,
      "loss": 6.1118,
      "mean_token_accuracy": 0.1704573854804039,
      "num_tokens": 18405770.0,
      "step": 7260
    },
    {
      "entropy": 6.328343677520752,
      "epoch": 0.8384304673975764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049408724791062,
      "loss": 6.1824,
      "mean_token_accuracy": 0.16913258135318757,
      "num_tokens": 18417980.0,
      "step": 7265
    },
    {
      "entropy": 6.394333171844482,
      "epoch": 0.8390075014425851,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004940778465279797,
      "loss": 6.0769,
      "mean_token_accuracy": 0.17640269845724105,
      "num_tokens": 18429511.0,
      "step": 7270
    },
    {
      "entropy": 6.282315731048584,
      "epoch": 0.8395845354875938,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940684377767647,
      "loss": 6.1463,
      "mean_token_accuracy": 0.1683313712477684,
      "num_tokens": 18443176.0,
      "step": 7275
    },
    {
      "entropy": 6.29357419013977,
      "epoch": 0.8401615695326025,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004940590216572916,
      "loss": 6.102,
      "mean_token_accuracy": 0.17451467961072922,
      "num_tokens": 18454664.0,
      "step": 7280
    },
    {
      "entropy": 6.3182899951934814,
      "epoch": 0.8407386035776111,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004940495981698772,
      "loss": 6.0923,
      "mean_token_accuracy": 0.17001918852329254,
      "num_tokens": 18465919.0,
      "step": 7285
    },
    {
      "entropy": 6.255602073669434,
      "epoch": 0.8413156376226197,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940401673148384,
      "loss": 6.0128,
      "mean_token_accuracy": 0.17237745225429535,
      "num_tokens": 18478847.0,
      "step": 7290
    },
    {
      "entropy": 6.330014038085937,
      "epoch": 0.8418926716676284,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940307290924923,
      "loss": 6.1426,
      "mean_token_accuracy": 0.16671836823225022,
      "num_tokens": 18490932.0,
      "step": 7295
    },
    {
      "entropy": 6.347289514541626,
      "epoch": 0.842469705712637,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004940212835031565,
      "loss": 6.1418,
      "mean_token_accuracy": 0.1622636690735817,
      "num_tokens": 18504333.0,
      "step": 7300
    },
    {
      "entropy": 6.292515897750855,
      "epoch": 0.8430467397576457,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004940118305471486,
      "loss": 6.1209,
      "mean_token_accuracy": 0.16780552119016648,
      "num_tokens": 18516021.0,
      "step": 7305
    },
    {
      "entropy": 6.254858493804932,
      "epoch": 0.8436237738026544,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004940023702247866,
      "loss": 6.0307,
      "mean_token_accuracy": 0.17692680954933165,
      "num_tokens": 18529281.0,
      "step": 7310
    },
    {
      "entropy": 6.250281906127929,
      "epoch": 0.844200807847663,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004939929025363886,
      "loss": 6.0826,
      "mean_token_accuracy": 0.1771777018904686,
      "num_tokens": 18541605.0,
      "step": 7315
    },
    {
      "entropy": 6.3663187503814695,
      "epoch": 0.8447778418926717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004939834274822731,
      "loss": 6.1244,
      "mean_token_accuracy": 0.16943843960762023,
      "num_tokens": 18553171.0,
      "step": 7320
    },
    {
      "entropy": 6.315127515792847,
      "epoch": 0.8453548759376803,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004939739450627588,
      "loss": 5.9897,
      "mean_token_accuracy": 0.1719767913222313,
      "num_tokens": 18565347.0,
      "step": 7325
    },
    {
      "entropy": 6.293874216079712,
      "epoch": 0.845931909982689,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939644552781647,
      "loss": 6.15,
      "mean_token_accuracy": 0.16545775383710862,
      "num_tokens": 18578102.0,
      "step": 7330
    },
    {
      "entropy": 6.291524267196655,
      "epoch": 0.8465089440276976,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939549581288099,
      "loss": 6.1244,
      "mean_token_accuracy": 0.1646455481648445,
      "num_tokens": 18590184.0,
      "step": 7335
    },
    {
      "entropy": 6.323913669586181,
      "epoch": 0.8470859780727062,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004939454536150138,
      "loss": 6.1435,
      "mean_token_accuracy": 0.16949068158864974,
      "num_tokens": 18603026.0,
      "step": 7340
    },
    {
      "entropy": 6.321570491790771,
      "epoch": 0.847663012117715,
      "grad_norm": 0.828125,
      "learning_rate": 0.000493935941737096,
      "loss": 6.1356,
      "mean_token_accuracy": 0.16862865388393403,
      "num_tokens": 18616841.0,
      "step": 7345
    },
    {
      "entropy": 6.315411138534546,
      "epoch": 0.8482400461627236,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004939264224953768,
      "loss": 6.0754,
      "mean_token_accuracy": 0.17875832915306092,
      "num_tokens": 18628777.0,
      "step": 7350
    },
    {
      "entropy": 6.31576189994812,
      "epoch": 0.8488170802077323,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000493916895890176,
      "loss": 6.1262,
      "mean_token_accuracy": 0.16413204818964006,
      "num_tokens": 18641626.0,
      "step": 7355
    },
    {
      "entropy": 6.406159973144531,
      "epoch": 0.8493941142527409,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493907361921814,
      "loss": 6.0959,
      "mean_token_accuracy": 0.1713714763522148,
      "num_tokens": 18653576.0,
      "step": 7360
    },
    {
      "entropy": 6.246872854232788,
      "epoch": 0.8499711482977496,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004938978205906118,
      "loss": 6.1322,
      "mean_token_accuracy": 0.17479570508003234,
      "num_tokens": 18666134.0,
      "step": 7365
    },
    {
      "entropy": 6.259529781341553,
      "epoch": 0.8505481823427582,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004938882718968901,
      "loss": 6.101,
      "mean_token_accuracy": 0.16042630970478058,
      "num_tokens": 18680029.0,
      "step": 7370
    },
    {
      "entropy": 6.365555238723755,
      "epoch": 0.8511252163877668,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004938787158409702,
      "loss": 6.1189,
      "mean_token_accuracy": 0.1694151431322098,
      "num_tokens": 18693136.0,
      "step": 7375
    },
    {
      "entropy": 6.323048782348633,
      "epoch": 0.8517022504327756,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004938691524231733,
      "loss": 6.171,
      "mean_token_accuracy": 0.16397368311882018,
      "num_tokens": 18706613.0,
      "step": 7380
    },
    {
      "entropy": 6.361038017272949,
      "epoch": 0.8522792844777842,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004938595816438212,
      "loss": 6.1415,
      "mean_token_accuracy": 0.16203895807266236,
      "num_tokens": 18719541.0,
      "step": 7385
    },
    {
      "entropy": 6.3449992656707765,
      "epoch": 0.8528563185227929,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938500035032358,
      "loss": 6.1301,
      "mean_token_accuracy": 0.17314599007368087,
      "num_tokens": 18730992.0,
      "step": 7390
    },
    {
      "entropy": 6.300184297561645,
      "epoch": 0.8534333525678015,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004938404180017392,
      "loss": 6.0712,
      "mean_token_accuracy": 0.1704453006386757,
      "num_tokens": 18743945.0,
      "step": 7395
    },
    {
      "entropy": 6.241794729232788,
      "epoch": 0.8540103866128101,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004938308251396539,
      "loss": 6.1287,
      "mean_token_accuracy": 0.1737273007631302,
      "num_tokens": 18756656.0,
      "step": 7400
    },
    {
      "entropy": 6.32796049118042,
      "epoch": 0.8545874206578188,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004938212249173024,
      "loss": 6.0837,
      "mean_token_accuracy": 0.17052091658115387,
      "num_tokens": 18769224.0,
      "step": 7405
    },
    {
      "entropy": 6.310837602615356,
      "epoch": 0.8551644547028274,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004938116173350078,
      "loss": 6.0749,
      "mean_token_accuracy": 0.17881231009960175,
      "num_tokens": 18782162.0,
      "step": 7410
    },
    {
      "entropy": 6.312178421020508,
      "epoch": 0.8557414887478362,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004938020023930932,
      "loss": 6.138,
      "mean_token_accuracy": 0.1700222000479698,
      "num_tokens": 18793014.0,
      "step": 7415
    },
    {
      "entropy": 6.400294160842895,
      "epoch": 0.8563185227928448,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004937923800918817,
      "loss": 6.2059,
      "mean_token_accuracy": 0.16382661759853362,
      "num_tokens": 18804879.0,
      "step": 7420
    },
    {
      "entropy": 6.3230022430419925,
      "epoch": 0.8568955568378535,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004937827504316974,
      "loss": 6.0589,
      "mean_token_accuracy": 0.17049174308776854,
      "num_tokens": 18817301.0,
      "step": 7425
    },
    {
      "entropy": 6.240580892562866,
      "epoch": 0.8574725908828621,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937731134128639,
      "loss": 6.0844,
      "mean_token_accuracy": 0.16979705542325974,
      "num_tokens": 18828424.0,
      "step": 7430
    },
    {
      "entropy": 6.309745979309082,
      "epoch": 0.8580496249278707,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937634690357054,
      "loss": 6.1167,
      "mean_token_accuracy": 0.17033765614032745,
      "num_tokens": 18840526.0,
      "step": 7435
    },
    {
      "entropy": 6.403766441345215,
      "epoch": 0.8586266589728794,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004937538173005462,
      "loss": 6.2014,
      "mean_token_accuracy": 0.16364138275384904,
      "num_tokens": 18854557.0,
      "step": 7440
    },
    {
      "entropy": 6.298587036132813,
      "epoch": 0.859203693017888,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937441582077111,
      "loss": 6.1573,
      "mean_token_accuracy": 0.16764754205942153,
      "num_tokens": 18868469.0,
      "step": 7445
    },
    {
      "entropy": 6.30604887008667,
      "epoch": 0.8597807270628968,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004937344917575249,
      "loss": 6.1008,
      "mean_token_accuracy": 0.16838963329792023,
      "num_tokens": 18880507.0,
      "step": 7450
    },
    {
      "entropy": 6.389565324783325,
      "epoch": 0.8603577611079054,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004937248179503127,
      "loss": 6.1776,
      "mean_token_accuracy": 0.16353571712970733,
      "num_tokens": 18892666.0,
      "step": 7455
    },
    {
      "entropy": 6.332041501998901,
      "epoch": 0.860934795152914,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004937151367863998,
      "loss": 6.1592,
      "mean_token_accuracy": 0.16899376958608628,
      "num_tokens": 18905022.0,
      "step": 7460
    },
    {
      "entropy": 6.33961443901062,
      "epoch": 0.8615118291979227,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000493705448266112,
      "loss": 6.1847,
      "mean_token_accuracy": 0.16367289870977403,
      "num_tokens": 18918279.0,
      "step": 7465
    },
    {
      "entropy": 6.329358911514282,
      "epoch": 0.8620888632429313,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936957523897752,
      "loss": 6.0615,
      "mean_token_accuracy": 0.16951747089624405,
      "num_tokens": 18930494.0,
      "step": 7470
    },
    {
      "entropy": 6.245223951339722,
      "epoch": 0.86266589728794,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004936860491577153,
      "loss": 6.0396,
      "mean_token_accuracy": 0.17186590582132338,
      "num_tokens": 18943918.0,
      "step": 7475
    },
    {
      "entropy": 6.25893611907959,
      "epoch": 0.8632429313329486,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004936763385702588,
      "loss": 6.1412,
      "mean_token_accuracy": 0.1655966654419899,
      "num_tokens": 18957170.0,
      "step": 7480
    },
    {
      "entropy": 6.309596061706543,
      "epoch": 0.8638199653779572,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936666206277322,
      "loss": 6.0708,
      "mean_token_accuracy": 0.1582779198884964,
      "num_tokens": 18970106.0,
      "step": 7485
    },
    {
      "entropy": 6.365174388885498,
      "epoch": 0.864396999422966,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004936568953304624,
      "loss": 6.147,
      "mean_token_accuracy": 0.168244831264019,
      "num_tokens": 18982797.0,
      "step": 7490
    },
    {
      "entropy": 6.329259634017944,
      "epoch": 0.8649740334679746,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004936471626787766,
      "loss": 6.2214,
      "mean_token_accuracy": 0.16954808235168456,
      "num_tokens": 18997122.0,
      "step": 7495
    },
    {
      "entropy": 6.344443130493164,
      "epoch": 0.8655510675129833,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004936374226730022,
      "loss": 6.1836,
      "mean_token_accuracy": 0.16055750995874404,
      "num_tokens": 19009345.0,
      "step": 7500
    },
    {
      "entropy": 6.354634952545166,
      "epoch": 0.8661281015579919,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936276753134666,
      "loss": 6.1688,
      "mean_token_accuracy": 0.16521380990743637,
      "num_tokens": 19021585.0,
      "step": 7505
    },
    {
      "entropy": 6.329277658462525,
      "epoch": 0.8667051356030006,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004936179206004976,
      "loss": 6.1355,
      "mean_token_accuracy": 0.16339420676231384,
      "num_tokens": 19034701.0,
      "step": 7510
    },
    {
      "entropy": 6.2735615253448485,
      "epoch": 0.8672821696480092,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936081585344236,
      "loss": 6.1192,
      "mean_token_accuracy": 0.17345345616340638,
      "num_tokens": 19047810.0,
      "step": 7515
    },
    {
      "entropy": 6.287371873855591,
      "epoch": 0.8678592036930178,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004935983891155727,
      "loss": 6.1019,
      "mean_token_accuracy": 0.173904687166214,
      "num_tokens": 19059546.0,
      "step": 7520
    },
    {
      "entropy": 6.357184267044067,
      "epoch": 0.8684362377380266,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004935886123442737,
      "loss": 6.1704,
      "mean_token_accuracy": 0.16851410567760466,
      "num_tokens": 19072302.0,
      "step": 7525
    },
    {
      "entropy": 6.344837951660156,
      "epoch": 0.8690132717830352,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004935788282208551,
      "loss": 6.1661,
      "mean_token_accuracy": 0.17020163536071778,
      "num_tokens": 19086283.0,
      "step": 7530
    },
    {
      "entropy": 6.306544637680053,
      "epoch": 0.8695903058280439,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004935690367456464,
      "loss": 6.0819,
      "mean_token_accuracy": 0.17379076927900314,
      "num_tokens": 19098586.0,
      "step": 7535
    },
    {
      "entropy": 6.294129991531372,
      "epoch": 0.8701673398730525,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004935592379189766,
      "loss": 6.0785,
      "mean_token_accuracy": 0.16863109916448593,
      "num_tokens": 19110795.0,
      "step": 7540
    },
    {
      "entropy": 6.309462785720825,
      "epoch": 0.8707443739180611,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935494317411754,
      "loss": 6.0386,
      "mean_token_accuracy": 0.17708732634782792,
      "num_tokens": 19123391.0,
      "step": 7545
    },
    {
      "entropy": 6.297340679168701,
      "epoch": 0.8713214079630698,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004935396182125726,
      "loss": 6.1017,
      "mean_token_accuracy": 0.17120068967342378,
      "num_tokens": 19135436.0,
      "step": 7550
    },
    {
      "entropy": 6.2987633228302,
      "epoch": 0.8718984420080784,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935297973334983,
      "loss": 6.0731,
      "mean_token_accuracy": 0.17533822655677794,
      "num_tokens": 19147210.0,
      "step": 7555
    },
    {
      "entropy": 6.354048061370849,
      "epoch": 0.8724754760530872,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935199691042828,
      "loss": 6.2243,
      "mean_token_accuracy": 0.1696576878428459,
      "num_tokens": 19161309.0,
      "step": 7560
    },
    {
      "entropy": 6.324620389938355,
      "epoch": 0.8730525100980958,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004935101335252568,
      "loss": 6.1642,
      "mean_token_accuracy": 0.16651286780834199,
      "num_tokens": 19174073.0,
      "step": 7565
    },
    {
      "entropy": 6.283979034423828,
      "epoch": 0.8736295441431045,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004935002905967509,
      "loss": 6.0653,
      "mean_token_accuracy": 0.17390892058610916,
      "num_tokens": 19187120.0,
      "step": 7570
    },
    {
      "entropy": 6.339527654647827,
      "epoch": 0.8742065781881131,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004934904403190963,
      "loss": 6.0223,
      "mean_token_accuracy": 0.17759824991226197,
      "num_tokens": 19201195.0,
      "step": 7575
    },
    {
      "entropy": 6.268510246276856,
      "epoch": 0.8747836122331217,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004934805826926242,
      "loss": 6.1503,
      "mean_token_accuracy": 0.16540689021348953,
      "num_tokens": 19214050.0,
      "step": 7580
    },
    {
      "entropy": 6.3610124588012695,
      "epoch": 0.8753606462781304,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004934707177176663,
      "loss": 6.1416,
      "mean_token_accuracy": 0.16823814809322357,
      "num_tokens": 19226678.0,
      "step": 7585
    },
    {
      "entropy": 6.340522146224975,
      "epoch": 0.875937680323139,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934608453945543,
      "loss": 6.1657,
      "mean_token_accuracy": 0.16496469229459762,
      "num_tokens": 19239038.0,
      "step": 7590
    },
    {
      "entropy": 6.214799928665161,
      "epoch": 0.8765147143681478,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004934509657236203,
      "loss": 6.0617,
      "mean_token_accuracy": 0.17199939340353013,
      "num_tokens": 19252934.0,
      "step": 7595
    },
    {
      "entropy": 6.388493824005127,
      "epoch": 0.8770917484131564,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004934410787051965,
      "loss": 6.1181,
      "mean_token_accuracy": 0.16989358365535737,
      "num_tokens": 19267646.0,
      "step": 7600
    },
    {
      "entropy": 6.334185314178467,
      "epoch": 0.877668782458165,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934311843396157,
      "loss": 6.108,
      "mean_token_accuracy": 0.17217393964529037,
      "num_tokens": 19279529.0,
      "step": 7605
    },
    {
      "entropy": 6.333045196533203,
      "epoch": 0.8782458165031737,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934212826272104,
      "loss": 6.1585,
      "mean_token_accuracy": 0.168229903280735,
      "num_tokens": 19291315.0,
      "step": 7610
    },
    {
      "entropy": 6.311127710342407,
      "epoch": 0.8788228505481823,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004934113735683137,
      "loss": 5.9887,
      "mean_token_accuracy": 0.18041695654392242,
      "num_tokens": 19303763.0,
      "step": 7615
    },
    {
      "entropy": 6.293818998336792,
      "epoch": 0.879399884593191,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000493401457163259,
      "loss": 6.1382,
      "mean_token_accuracy": 0.16354042440652847,
      "num_tokens": 19316400.0,
      "step": 7620
    },
    {
      "entropy": 6.28227686882019,
      "epoch": 0.8799769186381996,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004933915334123798,
      "loss": 6.0448,
      "mean_token_accuracy": 0.1758949264883995,
      "num_tokens": 19328227.0,
      "step": 7625
    },
    {
      "entropy": 6.260040950775147,
      "epoch": 0.8805539526832084,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004933816023160099,
      "loss": 5.9935,
      "mean_token_accuracy": 0.1776716187596321,
      "num_tokens": 19341211.0,
      "step": 7630
    },
    {
      "entropy": 6.282184839248657,
      "epoch": 0.881130986728217,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004933716638744832,
      "loss": 6.1425,
      "mean_token_accuracy": 0.17052342146635055,
      "num_tokens": 19353165.0,
      "step": 7635
    },
    {
      "entropy": 6.338343238830566,
      "epoch": 0.8817080207732256,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000493361718088134,
      "loss": 6.2006,
      "mean_token_accuracy": 0.16703000366687776,
      "num_tokens": 19365601.0,
      "step": 7640
    },
    {
      "entropy": 6.329648160934449,
      "epoch": 0.8822850548182343,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000493351764957297,
      "loss": 6.1175,
      "mean_token_accuracy": 0.1692799523472786,
      "num_tokens": 19377828.0,
      "step": 7645
    },
    {
      "entropy": 6.417537355422974,
      "epoch": 0.8828620888632429,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004933418044823068,
      "loss": 6.1337,
      "mean_token_accuracy": 0.1705181673169136,
      "num_tokens": 19390019.0,
      "step": 7650
    },
    {
      "entropy": 6.286002779006958,
      "epoch": 0.8834391229082516,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933318366634984,
      "loss": 6.0974,
      "mean_token_accuracy": 0.16971160471439362,
      "num_tokens": 19402179.0,
      "step": 7655
    },
    {
      "entropy": 6.289387607574463,
      "epoch": 0.8840161569532602,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933218615012072,
      "loss": 6.0488,
      "mean_token_accuracy": 0.16546063423156737,
      "num_tokens": 19412794.0,
      "step": 7660
    },
    {
      "entropy": 6.326491498947144,
      "epoch": 0.8845931909982689,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933118789957687,
      "loss": 6.1542,
      "mean_token_accuracy": 0.16174625530838965,
      "num_tokens": 19425932.0,
      "step": 7665
    },
    {
      "entropy": 6.367192554473877,
      "epoch": 0.8851702250432776,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004933018891475186,
      "loss": 6.0879,
      "mean_token_accuracy": 0.16910957992076875,
      "num_tokens": 19437543.0,
      "step": 7670
    },
    {
      "entropy": 6.3427753925323485,
      "epoch": 0.8857472590882862,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004932918919567927,
      "loss": 6.0655,
      "mean_token_accuracy": 0.17346468269824983,
      "num_tokens": 19450122.0,
      "step": 7675
    },
    {
      "entropy": 6.238670110702515,
      "epoch": 0.8863242931332949,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004932818874239275,
      "loss": 6.0811,
      "mean_token_accuracy": 0.16923540085554123,
      "num_tokens": 19464553.0,
      "step": 7680
    },
    {
      "entropy": 6.229796695709228,
      "epoch": 0.8869013271783035,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004932718755492595,
      "loss": 6.0449,
      "mean_token_accuracy": 0.17348248213529588,
      "num_tokens": 19477956.0,
      "step": 7685
    },
    {
      "entropy": 6.258459901809692,
      "epoch": 0.8874783612233121,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004932618563331254,
      "loss": 6.0766,
      "mean_token_accuracy": 0.166024249792099,
      "num_tokens": 19491610.0,
      "step": 7690
    },
    {
      "entropy": 6.231938743591309,
      "epoch": 0.8880553952683208,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004932518297758622,
      "loss": 5.9824,
      "mean_token_accuracy": 0.1762412056326866,
      "num_tokens": 19503612.0,
      "step": 7695
    },
    {
      "entropy": 6.255193471908569,
      "epoch": 0.8886324293133295,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004932417958778071,
      "loss": 6.0661,
      "mean_token_accuracy": 0.17482839226722718,
      "num_tokens": 19516956.0,
      "step": 7700
    },
    {
      "entropy": 6.198970603942871,
      "epoch": 0.8892094633583382,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004932317546392976,
      "loss": 6.0484,
      "mean_token_accuracy": 0.17286549359560013,
      "num_tokens": 19528545.0,
      "step": 7705
    },
    {
      "entropy": 6.335839414596558,
      "epoch": 0.8897864974033468,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932217060606714,
      "loss": 5.9887,
      "mean_token_accuracy": 0.17639979273080825,
      "num_tokens": 19539920.0,
      "step": 7710
    },
    {
      "entropy": 6.273063850402832,
      "epoch": 0.8903635314483554,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004932116501422665,
      "loss": 6.1135,
      "mean_token_accuracy": 0.16828238517045974,
      "num_tokens": 19551975.0,
      "step": 7715
    },
    {
      "entropy": 6.319854402542115,
      "epoch": 0.8909405654933641,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004932015868844211,
      "loss": 6.1178,
      "mean_token_accuracy": 0.1692993938922882,
      "num_tokens": 19565952.0,
      "step": 7720
    },
    {
      "entropy": 6.274467420578003,
      "epoch": 0.8915175995383727,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004931915162874738,
      "loss": 6.0669,
      "mean_token_accuracy": 0.17231060862541198,
      "num_tokens": 19578400.0,
      "step": 7725
    },
    {
      "entropy": 6.239807081222534,
      "epoch": 0.8920946335833814,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931814383517632,
      "loss": 6.0447,
      "mean_token_accuracy": 0.17400905042886733,
      "num_tokens": 19592237.0,
      "step": 7730
    },
    {
      "entropy": 6.385022783279419,
      "epoch": 0.8926716676283901,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004931713530776284,
      "loss": 6.1987,
      "mean_token_accuracy": 0.16712776869535445,
      "num_tokens": 19606172.0,
      "step": 7735
    },
    {
      "entropy": 6.298589515686035,
      "epoch": 0.8932487016733988,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004931612604654083,
      "loss": 6.1536,
      "mean_token_accuracy": 0.17061897069215776,
      "num_tokens": 19618245.0,
      "step": 7740
    },
    {
      "entropy": 6.362525987625122,
      "epoch": 0.8938257357184074,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004931511605154428,
      "loss": 6.0954,
      "mean_token_accuracy": 0.1694071665406227,
      "num_tokens": 19630921.0,
      "step": 7745
    },
    {
      "entropy": 6.318386745452881,
      "epoch": 0.894402769763416,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004931410532280711,
      "loss": 6.1002,
      "mean_token_accuracy": 0.16700265556573868,
      "num_tokens": 19643869.0,
      "step": 7750
    },
    {
      "entropy": 6.338892793655395,
      "epoch": 0.8949798038084247,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004931309386036337,
      "loss": 6.0796,
      "mean_token_accuracy": 0.17166771739721298,
      "num_tokens": 19657819.0,
      "step": 7755
    },
    {
      "entropy": 6.332481241226196,
      "epoch": 0.8955568378534333,
      "grad_norm": 0.875,
      "learning_rate": 0.0004931208166424704,
      "loss": 6.1639,
      "mean_token_accuracy": 0.16711659282445906,
      "num_tokens": 19670847.0,
      "step": 7760
    },
    {
      "entropy": 6.290749549865723,
      "epoch": 0.896133871898442,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931106873449219,
      "loss": 6.0904,
      "mean_token_accuracy": 0.16827981024980546,
      "num_tokens": 19684055.0,
      "step": 7765
    },
    {
      "entropy": 6.255268526077271,
      "epoch": 0.8967109059434507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004931005507113285,
      "loss": 5.9975,
      "mean_token_accuracy": 0.1766640618443489,
      "num_tokens": 19696731.0,
      "step": 7770
    },
    {
      "entropy": 6.371890735626221,
      "epoch": 0.8972879399884593,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004930904067420317,
      "loss": 6.1297,
      "mean_token_accuracy": 0.16785583049058914,
      "num_tokens": 19708787.0,
      "step": 7775
    },
    {
      "entropy": 6.323027563095093,
      "epoch": 0.897864974033468,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004930802554373723,
      "loss": 6.1471,
      "mean_token_accuracy": 0.16464302837848663,
      "num_tokens": 19722570.0,
      "step": 7780
    },
    {
      "entropy": 6.376014757156372,
      "epoch": 0.8984420080784766,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004930700967976918,
      "loss": 6.1277,
      "mean_token_accuracy": 0.17113870084285737,
      "num_tokens": 19734690.0,
      "step": 7785
    },
    {
      "entropy": 6.341323328018189,
      "epoch": 0.8990190421234853,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000493059930823332,
      "loss": 6.1289,
      "mean_token_accuracy": 0.1661938413977623,
      "num_tokens": 19747907.0,
      "step": 7790
    },
    {
      "entropy": 6.278427219390869,
      "epoch": 0.8995960761684939,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004930497575146346,
      "loss": 6.0954,
      "mean_token_accuracy": 0.17348378151655197,
      "num_tokens": 19760869.0,
      "step": 7795
    },
    {
      "entropy": 6.302192831039429,
      "epoch": 0.9001731102135025,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930395768719421,
      "loss": 5.9991,
      "mean_token_accuracy": 0.1784247100353241,
      "num_tokens": 19773097.0,
      "step": 7800
    },
    {
      "entropy": 6.289451456069946,
      "epoch": 0.9007501442585113,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004930293888955966,
      "loss": 6.1225,
      "mean_token_accuracy": 0.16142976507544518,
      "num_tokens": 19786070.0,
      "step": 7805
    },
    {
      "entropy": 6.357674837112427,
      "epoch": 0.9013271783035199,
      "grad_norm": 0.859375,
      "learning_rate": 0.000493019193585941,
      "loss": 6.1119,
      "mean_token_accuracy": 0.1719981074333191,
      "num_tokens": 19799244.0,
      "step": 7810
    },
    {
      "entropy": 6.2633469581604,
      "epoch": 0.9019042123485286,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000493008990943318,
      "loss": 6.1486,
      "mean_token_accuracy": 0.16575743407011032,
      "num_tokens": 19811965.0,
      "step": 7815
    },
    {
      "entropy": 6.37284574508667,
      "epoch": 0.9024812463935372,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004929987809680709,
      "loss": 6.1125,
      "mean_token_accuracy": 0.16965578645467758,
      "num_tokens": 19825531.0,
      "step": 7820
    },
    {
      "entropy": 6.3032989501953125,
      "epoch": 0.9030582804385459,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004929885636605432,
      "loss": 6.0596,
      "mean_token_accuracy": 0.1738504096865654,
      "num_tokens": 19839166.0,
      "step": 7825
    },
    {
      "entropy": 6.281108570098877,
      "epoch": 0.9036353144835545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929783390210784,
      "loss": 6.0935,
      "mean_token_accuracy": 0.16765460520982742,
      "num_tokens": 19852187.0,
      "step": 7830
    },
    {
      "entropy": 6.308529853820801,
      "epoch": 0.9042123485285631,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004929681070500204,
      "loss": 6.1392,
      "mean_token_accuracy": 0.16539253145456315,
      "num_tokens": 19864309.0,
      "step": 7835
    },
    {
      "entropy": 6.293935537338257,
      "epoch": 0.9047893825735719,
      "grad_norm": 0.875,
      "learning_rate": 0.0004929578677477135,
      "loss": 6.1353,
      "mean_token_accuracy": 0.1615119993686676,
      "num_tokens": 19877475.0,
      "step": 7840
    },
    {
      "entropy": 6.360176944732666,
      "epoch": 0.9053664166185805,
      "grad_norm": 0.875,
      "learning_rate": 0.0004929476211145019,
      "loss": 6.0895,
      "mean_token_accuracy": 0.1671235129237175,
      "num_tokens": 19889984.0,
      "step": 7845
    },
    {
      "entropy": 6.29410719871521,
      "epoch": 0.9059434506635892,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929373671507303,
      "loss": 6.0358,
      "mean_token_accuracy": 0.18178427070379258,
      "num_tokens": 19902135.0,
      "step": 7850
    },
    {
      "entropy": 6.2419897556304935,
      "epoch": 0.9065204847085978,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004929271058567436,
      "loss": 6.0894,
      "mean_token_accuracy": 0.173341503739357,
      "num_tokens": 19915214.0,
      "step": 7855
    },
    {
      "entropy": 6.316197299957276,
      "epoch": 0.9070975187536064,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000492916837232887,
      "loss": 6.1189,
      "mean_token_accuracy": 0.16358882784843445,
      "num_tokens": 19929113.0,
      "step": 7860
    },
    {
      "entropy": 6.398955392837524,
      "epoch": 0.9076745527986151,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004929065612795058,
      "loss": 6.1305,
      "mean_token_accuracy": 0.1644856885075569,
      "num_tokens": 19942727.0,
      "step": 7865
    },
    {
      "entropy": 6.257999897003174,
      "epoch": 0.9082515868436237,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004928962779969456,
      "loss": 6.0676,
      "mean_token_accuracy": 0.1740820974111557,
      "num_tokens": 19954863.0,
      "step": 7870
    },
    {
      "entropy": 6.227055358886719,
      "epoch": 0.9088286208886325,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004928859873855524,
      "loss": 6.0157,
      "mean_token_accuracy": 0.17968434393405913,
      "num_tokens": 19967555.0,
      "step": 7875
    },
    {
      "entropy": 6.388265609741211,
      "epoch": 0.9094056549336411,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004928756894456723,
      "loss": 6.1258,
      "mean_token_accuracy": 0.16835675835609437,
      "num_tokens": 19979992.0,
      "step": 7880
    },
    {
      "entropy": 6.325529479980469,
      "epoch": 0.9099826889786498,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004928653841776515,
      "loss": 6.1594,
      "mean_token_accuracy": 0.16871841698884965,
      "num_tokens": 19993231.0,
      "step": 7885
    },
    {
      "entropy": 6.3724034309387205,
      "epoch": 0.9105597230236584,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928550715818368,
      "loss": 6.1318,
      "mean_token_accuracy": 0.16597728282213212,
      "num_tokens": 20005839.0,
      "step": 7890
    },
    {
      "entropy": 6.320882320404053,
      "epoch": 0.911136757068667,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004928447516585749,
      "loss": 6.1033,
      "mean_token_accuracy": 0.1678915873169899,
      "num_tokens": 20017423.0,
      "step": 7895
    },
    {
      "entropy": 6.278266286849975,
      "epoch": 0.9117137911136757,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000492834424408213,
      "loss": 6.1393,
      "mean_token_accuracy": 0.16712916046380996,
      "num_tokens": 20030989.0,
      "step": 7900
    },
    {
      "entropy": 6.356175470352173,
      "epoch": 0.9122908251586843,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004928240898310984,
      "loss": 6.1949,
      "mean_token_accuracy": 0.16000735461711885,
      "num_tokens": 20043763.0,
      "step": 7905
    },
    {
      "entropy": 6.308985090255737,
      "epoch": 0.9128678592036931,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004928137479275789,
      "loss": 6.0511,
      "mean_token_accuracy": 0.17675268799066543,
      "num_tokens": 20057012.0,
      "step": 7910
    },
    {
      "entropy": 6.290041828155518,
      "epoch": 0.9134448932487017,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000492803398698002,
      "loss": 6.0633,
      "mean_token_accuracy": 0.17207225859165193,
      "num_tokens": 20069093.0,
      "step": 7915
    },
    {
      "entropy": 6.248224306106567,
      "epoch": 0.9140219272937103,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004927930421427161,
      "loss": 6.0368,
      "mean_token_accuracy": 0.17193564623594285,
      "num_tokens": 20081392.0,
      "step": 7920
    },
    {
      "entropy": 6.298941898345947,
      "epoch": 0.914598961338719,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927826782620694,
      "loss": 6.0669,
      "mean_token_accuracy": 0.1737733945250511,
      "num_tokens": 20092621.0,
      "step": 7925
    },
    {
      "entropy": 6.283964967727661,
      "epoch": 0.9151759953837276,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004927723070564104,
      "loss": 6.091,
      "mean_token_accuracy": 0.1684947982430458,
      "num_tokens": 20105423.0,
      "step": 7930
    },
    {
      "entropy": 6.348589468002319,
      "epoch": 0.9157530294287363,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004927619285260881,
      "loss": 6.078,
      "mean_token_accuracy": 0.16632856130599977,
      "num_tokens": 20117835.0,
      "step": 7935
    },
    {
      "entropy": 6.2604001522064205,
      "epoch": 0.9163300634737449,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004927515426714515,
      "loss": 5.9604,
      "mean_token_accuracy": 0.18144481480121613,
      "num_tokens": 20130563.0,
      "step": 7940
    },
    {
      "entropy": 6.2235535144805905,
      "epoch": 0.9169070975187537,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00049274114949285,
      "loss": 6.049,
      "mean_token_accuracy": 0.17362369149923323,
      "num_tokens": 20144507.0,
      "step": 7945
    },
    {
      "entropy": 6.298459100723266,
      "epoch": 0.9174841315637623,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000492730748990633,
      "loss": 6.0629,
      "mean_token_accuracy": 0.17503987103700638,
      "num_tokens": 20156341.0,
      "step": 7950
    },
    {
      "entropy": 6.308384609222412,
      "epoch": 0.9180611656087709,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004927203411651504,
      "loss": 6.1037,
      "mean_token_accuracy": 0.1752057820558548,
      "num_tokens": 20169410.0,
      "step": 7955
    },
    {
      "entropy": 6.322392272949219,
      "epoch": 0.9186381996537796,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927099260167523,
      "loss": 6.1004,
      "mean_token_accuracy": 0.1718139886856079,
      "num_tokens": 20181233.0,
      "step": 7960
    },
    {
      "entropy": 6.331369638442993,
      "epoch": 0.9192152336987882,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004926995035457889,
      "loss": 6.1059,
      "mean_token_accuracy": 0.16894071847200393,
      "num_tokens": 20193234.0,
      "step": 7965
    },
    {
      "entropy": 6.249935102462769,
      "epoch": 0.9197922677437969,
      "grad_norm": 0.828125,
      "learning_rate": 0.000492689073752611,
      "loss": 6.0168,
      "mean_token_accuracy": 0.17317250818014146,
      "num_tokens": 20206784.0,
      "step": 7970
    },
    {
      "entropy": 6.263874673843384,
      "epoch": 0.9203693017888055,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926786366375691,
      "loss": 6.015,
      "mean_token_accuracy": 0.18037747889757155,
      "num_tokens": 20218597.0,
      "step": 7975
    },
    {
      "entropy": 6.214323377609253,
      "epoch": 0.9209463358338142,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004926681922010145,
      "loss": 6.0014,
      "mean_token_accuracy": 0.17502579391002654,
      "num_tokens": 20232703.0,
      "step": 7980
    },
    {
      "entropy": 6.322141790390015,
      "epoch": 0.9215233698788229,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926577404432982,
      "loss": 6.083,
      "mean_token_accuracy": 0.16904457658529282,
      "num_tokens": 20245811.0,
      "step": 7985
    },
    {
      "entropy": 6.22695517539978,
      "epoch": 0.9221004039238315,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926472813647721,
      "loss": 5.9668,
      "mean_token_accuracy": 0.17562677413225175,
      "num_tokens": 20258055.0,
      "step": 7990
    },
    {
      "entropy": 6.290423822402954,
      "epoch": 0.9226774379688402,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004926368149657876,
      "loss": 6.0638,
      "mean_token_accuracy": 0.17215737104415893,
      "num_tokens": 20271250.0,
      "step": 7995
    },
    {
      "entropy": 6.26991376876831,
      "epoch": 0.9232544720138488,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004926263412466972,
      "loss": 6.0108,
      "mean_token_accuracy": 0.17998398542404176,
      "num_tokens": 20284013.0,
      "step": 8000
    },
    {
      "entropy": 6.283149528503418,
      "epoch": 0.9238315060588574,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004926158602078527,
      "loss": 6.085,
      "mean_token_accuracy": 0.1768463060259819,
      "num_tokens": 20297413.0,
      "step": 8005
    },
    {
      "entropy": 6.258905601501465,
      "epoch": 0.9244085401038661,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004926053718496069,
      "loss": 6.0607,
      "mean_token_accuracy": 0.17365098595619202,
      "num_tokens": 20310026.0,
      "step": 8010
    },
    {
      "entropy": 6.2548730850219725,
      "epoch": 0.9249855741488748,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004925948761723126,
      "loss": 6.0253,
      "mean_token_accuracy": 0.17915035635232926,
      "num_tokens": 20322118.0,
      "step": 8015
    },
    {
      "entropy": 6.292974901199341,
      "epoch": 0.9255626081938835,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004925843731763226,
      "loss": 6.1118,
      "mean_token_accuracy": 0.16737297773361207,
      "num_tokens": 20335252.0,
      "step": 8020
    },
    {
      "entropy": 6.393998670578003,
      "epoch": 0.9261396422388921,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004925738628619904,
      "loss": 6.1285,
      "mean_token_accuracy": 0.16633883416652678,
      "num_tokens": 20348201.0,
      "step": 8025
    },
    {
      "entropy": 6.302540397644043,
      "epoch": 0.9267166762839008,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004925633452296693,
      "loss": 6.0785,
      "mean_token_accuracy": 0.17153316140174865,
      "num_tokens": 20360250.0,
      "step": 8030
    },
    {
      "entropy": 6.287873649597168,
      "epoch": 0.9272937103289094,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004925528202797131,
      "loss": 6.0467,
      "mean_token_accuracy": 0.17138591706752776,
      "num_tokens": 20372376.0,
      "step": 8035
    },
    {
      "entropy": 6.3305846691131595,
      "epoch": 0.927870744373918,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004925422880124761,
      "loss": 6.0498,
      "mean_token_accuracy": 0.1675078883767128,
      "num_tokens": 20384484.0,
      "step": 8040
    },
    {
      "entropy": 6.297106599807739,
      "epoch": 0.9284477784189267,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004925317484283121,
      "loss": 6.1007,
      "mean_token_accuracy": 0.17022451460361482,
      "num_tokens": 20396519.0,
      "step": 8045
    },
    {
      "entropy": 6.2978801250457765,
      "epoch": 0.9290248124639354,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004925212015275758,
      "loss": 6.0651,
      "mean_token_accuracy": 0.17090250551700592,
      "num_tokens": 20409108.0,
      "step": 8050
    },
    {
      "entropy": 6.301689147949219,
      "epoch": 0.9296018465089441,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000492510647310622,
      "loss": 6.0951,
      "mean_token_accuracy": 0.17218709588050843,
      "num_tokens": 20421929.0,
      "step": 8055
    },
    {
      "entropy": 6.268299961090088,
      "epoch": 0.9301788805539527,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004925000857778055,
      "loss": 6.11,
      "mean_token_accuracy": 0.17204220592975616,
      "num_tokens": 20435453.0,
      "step": 8060
    },
    {
      "entropy": 6.288744974136352,
      "epoch": 0.9307559145989613,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004924895169294818,
      "loss": 6.0847,
      "mean_token_accuracy": 0.16962202191352843,
      "num_tokens": 20447867.0,
      "step": 8065
    },
    {
      "entropy": 6.378092336654663,
      "epoch": 0.93133294864397,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004924789407660062,
      "loss": 6.1279,
      "mean_token_accuracy": 0.17544363737106322,
      "num_tokens": 20461150.0,
      "step": 8070
    },
    {
      "entropy": 6.2715552806854244,
      "epoch": 0.9319099826889786,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924683572877345,
      "loss": 5.9922,
      "mean_token_accuracy": 0.1773497223854065,
      "num_tokens": 20474012.0,
      "step": 8075
    },
    {
      "entropy": 6.303848218917847,
      "epoch": 0.9324870167339873,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924577664950225,
      "loss": 6.0621,
      "mean_token_accuracy": 0.16593163311481476,
      "num_tokens": 20485833.0,
      "step": 8080
    },
    {
      "entropy": 6.335386610031128,
      "epoch": 0.933064050778996,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004924471683882266,
      "loss": 6.1018,
      "mean_token_accuracy": 0.1694898262619972,
      "num_tokens": 20499277.0,
      "step": 8085
    },
    {
      "entropy": 6.28719310760498,
      "epoch": 0.9336410848240047,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924365629677031,
      "loss": 6.0842,
      "mean_token_accuracy": 0.17266414165496827,
      "num_tokens": 20510789.0,
      "step": 8090
    },
    {
      "entropy": 6.26942253112793,
      "epoch": 0.9342181188690133,
      "grad_norm": 0.9375,
      "learning_rate": 0.000492425950233809,
      "loss": 6.0536,
      "mean_token_accuracy": 0.1782476782798767,
      "num_tokens": 20523273.0,
      "step": 8095
    },
    {
      "entropy": 6.358910179138183,
      "epoch": 0.9347951529140219,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000492415330186901,
      "loss": 6.1267,
      "mean_token_accuracy": 0.1688690036535263,
      "num_tokens": 20535466.0,
      "step": 8100
    },
    {
      "entropy": 6.415497732162476,
      "epoch": 0.9353721869590306,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004924047028273364,
      "loss": 6.1822,
      "mean_token_accuracy": 0.1660977765917778,
      "num_tokens": 20548108.0,
      "step": 8105
    },
    {
      "entropy": 6.251081037521362,
      "epoch": 0.9359492210040392,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004923940681554725,
      "loss": 6.0452,
      "mean_token_accuracy": 0.17263105362653733,
      "num_tokens": 20559859.0,
      "step": 8110
    },
    {
      "entropy": 6.285415458679199,
      "epoch": 0.9365262550490479,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004923834261716672,
      "loss": 6.168,
      "mean_token_accuracy": 0.16842860281467437,
      "num_tokens": 20572569.0,
      "step": 8115
    },
    {
      "entropy": 6.300700378417969,
      "epoch": 0.9371032890940566,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004923727768762782,
      "loss": 6.0822,
      "mean_token_accuracy": 0.16640304476022721,
      "num_tokens": 20585014.0,
      "step": 8120
    },
    {
      "entropy": 6.27948899269104,
      "epoch": 0.9376803231390652,
      "grad_norm": 0.84375,
      "learning_rate": 0.000492362120269664,
      "loss": 6.1351,
      "mean_token_accuracy": 0.1714727073907852,
      "num_tokens": 20597216.0,
      "step": 8125
    },
    {
      "entropy": 6.2442076206207275,
      "epoch": 0.9382573571840739,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004923514563521827,
      "loss": 6.0315,
      "mean_token_accuracy": 0.1827470049262047,
      "num_tokens": 20609193.0,
      "step": 8130
    },
    {
      "entropy": 6.295881223678589,
      "epoch": 0.9388343912290825,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004923407851241933,
      "loss": 6.0744,
      "mean_token_accuracy": 0.1688932627439499,
      "num_tokens": 20620777.0,
      "step": 8135
    },
    {
      "entropy": 6.251183223724365,
      "epoch": 0.9394114252740912,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004923301065860545,
      "loss": 6.0663,
      "mean_token_accuracy": 0.17291734665632247,
      "num_tokens": 20633244.0,
      "step": 8140
    },
    {
      "entropy": 6.292584037780761,
      "epoch": 0.9399884593190998,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004923194207381254,
      "loss": 6.0839,
      "mean_token_accuracy": 0.17595846354961395,
      "num_tokens": 20645622.0,
      "step": 8145
    },
    {
      "entropy": 6.329843950271607,
      "epoch": 0.9405654933641084,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004923087275807656,
      "loss": 6.1581,
      "mean_token_accuracy": 0.16922140568494798,
      "num_tokens": 20658793.0,
      "step": 8150
    },
    {
      "entropy": 6.264755916595459,
      "epoch": 0.9411425274091172,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004922980271143347,
      "loss": 5.9793,
      "mean_token_accuracy": 0.18203234672546387,
      "num_tokens": 20673009.0,
      "step": 8155
    },
    {
      "entropy": 6.368770456314087,
      "epoch": 0.9417195614541258,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922873193391927,
      "loss": 6.1535,
      "mean_token_accuracy": 0.17132313102483748,
      "num_tokens": 20685061.0,
      "step": 8160
    },
    {
      "entropy": 6.282876491546631,
      "epoch": 0.9422965954991345,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004922766042556994,
      "loss": 5.9642,
      "mean_token_accuracy": 0.17185274958610536,
      "num_tokens": 20696854.0,
      "step": 8165
    },
    {
      "entropy": 6.231480884552002,
      "epoch": 0.9428736295441431,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004922658818642156,
      "loss": 6.114,
      "mean_token_accuracy": 0.16710520386695862,
      "num_tokens": 20709259.0,
      "step": 8170
    },
    {
      "entropy": 6.261470890045166,
      "epoch": 0.9434506635891518,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004922551521651018,
      "loss": 5.9355,
      "mean_token_accuracy": 0.18289135992527009,
      "num_tokens": 20722382.0,
      "step": 8175
    },
    {
      "entropy": 6.2622942447662355,
      "epoch": 0.9440276976341604,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922444151587189,
      "loss": 5.9892,
      "mean_token_accuracy": 0.18124618977308274,
      "num_tokens": 20734809.0,
      "step": 8180
    },
    {
      "entropy": 6.1864806652069095,
      "epoch": 0.944604731679169,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922336708454279,
      "loss": 5.9677,
      "mean_token_accuracy": 0.17568960040807724,
      "num_tokens": 20747490.0,
      "step": 8185
    },
    {
      "entropy": 6.301131677627564,
      "epoch": 0.9451817657241778,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004922229192255903,
      "loss": 6.0495,
      "mean_token_accuracy": 0.17025782018899918,
      "num_tokens": 20759398.0,
      "step": 8190
    },
    {
      "entropy": 6.299096727371216,
      "epoch": 0.9457587997691864,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004922121602995678,
      "loss": 6.0754,
      "mean_token_accuracy": 0.17925020754337312,
      "num_tokens": 20773057.0,
      "step": 8195
    },
    {
      "entropy": 6.3332267761230465,
      "epoch": 0.9463358338141951,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004922013940677221,
      "loss": 6.0702,
      "mean_token_accuracy": 0.1656668648123741,
      "num_tokens": 20786084.0,
      "step": 8200
    },
    {
      "entropy": 6.283103036880493,
      "epoch": 0.9469128678592037,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004921906205304155,
      "loss": 6.011,
      "mean_token_accuracy": 0.17718872129917146,
      "num_tokens": 20798459.0,
      "step": 8205
    },
    {
      "entropy": 6.129636430740357,
      "epoch": 0.9474899019042123,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004921798396880101,
      "loss": 5.8363,
      "mean_token_accuracy": 0.18672388941049575,
      "num_tokens": 20811242.0,
      "step": 8210
    },
    {
      "entropy": 6.194643783569336,
      "epoch": 0.948066935949221,
      "grad_norm": 0.875,
      "learning_rate": 0.0004921690515408688,
      "loss": 6.0814,
      "mean_token_accuracy": 0.16723225861787797,
      "num_tokens": 20824757.0,
      "step": 8215
    },
    {
      "entropy": 6.365654611587525,
      "epoch": 0.9486439699942296,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004921582560893543,
      "loss": 6.0779,
      "mean_token_accuracy": 0.17129451036453247,
      "num_tokens": 20836794.0,
      "step": 8220
    },
    {
      "entropy": 6.330667543411255,
      "epoch": 0.9492210040392384,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004921474533338297,
      "loss": 6.0159,
      "mean_token_accuracy": 0.169456784427166,
      "num_tokens": 20849240.0,
      "step": 8225
    },
    {
      "entropy": 6.233933162689209,
      "epoch": 0.949798038084247,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004921366432746585,
      "loss": 6.056,
      "mean_token_accuracy": 0.17418997287750243,
      "num_tokens": 20861362.0,
      "step": 8230
    },
    {
      "entropy": 6.289814519882202,
      "epoch": 0.9503750721292556,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492125825912204,
      "loss": 6.0669,
      "mean_token_accuracy": 0.179733544588089,
      "num_tokens": 20874925.0,
      "step": 8235
    },
    {
      "entropy": 6.290691471099853,
      "epoch": 0.9509521061742643,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004921150012468302,
      "loss": 6.0607,
      "mean_token_accuracy": 0.17308035045862197,
      "num_tokens": 20888098.0,
      "step": 8240
    },
    {
      "entropy": 6.262841844558716,
      "epoch": 0.9515291402192729,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004921041692789013,
      "loss": 6.0561,
      "mean_token_accuracy": 0.17421980947256088,
      "num_tokens": 20901125.0,
      "step": 8245
    },
    {
      "entropy": 6.324599552154541,
      "epoch": 0.9521061742642816,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004920933300087813,
      "loss": 6.1211,
      "mean_token_accuracy": 0.16891735643148423,
      "num_tokens": 20914815.0,
      "step": 8250
    },
    {
      "entropy": 6.191049814224243,
      "epoch": 0.9526832083092902,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920824834368353,
      "loss": 5.9801,
      "mean_token_accuracy": 0.18388531357049942,
      "num_tokens": 20928153.0,
      "step": 8255
    },
    {
      "entropy": 6.379620599746704,
      "epoch": 0.953260242354299,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004920716295634275,
      "loss": 6.139,
      "mean_token_accuracy": 0.16048989295959473,
      "num_tokens": 20940601.0,
      "step": 8260
    },
    {
      "entropy": 6.324798536300659,
      "epoch": 0.9538372763993076,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004920607683889235,
      "loss": 6.1358,
      "mean_token_accuracy": 0.1698945254087448,
      "num_tokens": 20953158.0,
      "step": 8265
    },
    {
      "entropy": 6.3390051364898685,
      "epoch": 0.9544143104443162,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004920498999136882,
      "loss": 6.1179,
      "mean_token_accuracy": 0.16918241530656813,
      "num_tokens": 20966572.0,
      "step": 8270
    },
    {
      "entropy": 6.382406949996948,
      "epoch": 0.9549913444893249,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004920390241380874,
      "loss": 6.0955,
      "mean_token_accuracy": 0.17139170467853546,
      "num_tokens": 20978432.0,
      "step": 8275
    },
    {
      "entropy": 6.287047624588013,
      "epoch": 0.9555683785343335,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920281410624868,
      "loss": 6.1385,
      "mean_token_accuracy": 0.1723800152540207,
      "num_tokens": 20991454.0,
      "step": 8280
    },
    {
      "entropy": 6.278736734390259,
      "epoch": 0.9561454125793422,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004920172506872525,
      "loss": 6.0148,
      "mean_token_accuracy": 0.17681244760751724,
      "num_tokens": 21004111.0,
      "step": 8285
    },
    {
      "entropy": 6.261019992828369,
      "epoch": 0.9567224466243508,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004920063530127508,
      "loss": 6.0,
      "mean_token_accuracy": 0.16469207853078843,
      "num_tokens": 21017523.0,
      "step": 8290
    },
    {
      "entropy": 6.340602111816406,
      "epoch": 0.9572994806693594,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919954480393482,
      "loss": 6.0699,
      "mean_token_accuracy": 0.16875892877578735,
      "num_tokens": 21028723.0,
      "step": 8295
    },
    {
      "entropy": 6.284719848632813,
      "epoch": 0.9578765147143682,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004919845357674114,
      "loss": 6.066,
      "mean_token_accuracy": 0.17578147649765014,
      "num_tokens": 21042205.0,
      "step": 8300
    },
    {
      "entropy": 6.282210159301758,
      "epoch": 0.9584535487593768,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004919736161973076,
      "loss": 6.0003,
      "mean_token_accuracy": 0.17048712372779845,
      "num_tokens": 21053798.0,
      "step": 8305
    },
    {
      "entropy": 6.315688991546631,
      "epoch": 0.9590305828043855,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491962689329404,
      "loss": 6.164,
      "mean_token_accuracy": 0.16235949844121933,
      "num_tokens": 21068333.0,
      "step": 8310
    },
    {
      "entropy": 6.3100635528564455,
      "epoch": 0.9596076168493941,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004919517551640681,
      "loss": 6.0228,
      "mean_token_accuracy": 0.18080112636089324,
      "num_tokens": 21079484.0,
      "step": 8315
    },
    {
      "entropy": 6.214249849319458,
      "epoch": 0.9601846508944027,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004919408137016677,
      "loss": 6.0446,
      "mean_token_accuracy": 0.17698175609111785,
      "num_tokens": 21090940.0,
      "step": 8320
    },
    {
      "entropy": 6.36199107170105,
      "epoch": 0.9607616849394114,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004919298649425708,
      "loss": 6.0958,
      "mean_token_accuracy": 0.16862892359495163,
      "num_tokens": 21104351.0,
      "step": 8325
    },
    {
      "entropy": 6.322522687911987,
      "epoch": 0.96133871898442,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004919189088871456,
      "loss": 6.0604,
      "mean_token_accuracy": 0.16790530383586882,
      "num_tokens": 21116272.0,
      "step": 8330
    },
    {
      "entropy": 6.316466999053955,
      "epoch": 0.9619157530294288,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004919079455357608,
      "loss": 6.1344,
      "mean_token_accuracy": 0.16736244559288024,
      "num_tokens": 21129448.0,
      "step": 8335
    },
    {
      "entropy": 6.3490674018859865,
      "epoch": 0.9624927870744374,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004918969748887847,
      "loss": 6.1193,
      "mean_token_accuracy": 0.1631232514977455,
      "num_tokens": 21143668.0,
      "step": 8340
    },
    {
      "entropy": 6.34621114730835,
      "epoch": 0.9630698211194461,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004918859969465868,
      "loss": 6.1126,
      "mean_token_accuracy": 0.16959568858146667,
      "num_tokens": 21156662.0,
      "step": 8345
    },
    {
      "entropy": 6.289714431762695,
      "epoch": 0.9636468551644547,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004918750117095361,
      "loss": 6.068,
      "mean_token_accuracy": 0.1759003534913063,
      "num_tokens": 21170271.0,
      "step": 8350
    },
    {
      "entropy": 6.267118072509765,
      "epoch": 0.9642238892094633,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004918640191780021,
      "loss": 6.1094,
      "mean_token_accuracy": 0.16548217386007308,
      "num_tokens": 21183878.0,
      "step": 8355
    },
    {
      "entropy": 6.34764256477356,
      "epoch": 0.964800923254472,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004918530193523546,
      "loss": 6.0237,
      "mean_token_accuracy": 0.173106586933136,
      "num_tokens": 21196088.0,
      "step": 8360
    },
    {
      "entropy": 6.302213096618653,
      "epoch": 0.9653779572994806,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004918420122329634,
      "loss": 6.1573,
      "mean_token_accuracy": 0.16970721781253814,
      "num_tokens": 21208640.0,
      "step": 8365
    },
    {
      "entropy": 6.271123838424683,
      "epoch": 0.9659549913444894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918309978201989,
      "loss": 5.9581,
      "mean_token_accuracy": 0.1844442144036293,
      "num_tokens": 21221246.0,
      "step": 8370
    },
    {
      "entropy": 6.25211730003357,
      "epoch": 0.966532025389498,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004918199761144315,
      "loss": 6.0108,
      "mean_token_accuracy": 0.1769232526421547,
      "num_tokens": 21233834.0,
      "step": 8375
    },
    {
      "entropy": 6.292379760742188,
      "epoch": 0.9671090594345066,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004918089471160318,
      "loss": 6.0398,
      "mean_token_accuracy": 0.17788860201835632,
      "num_tokens": 21247388.0,
      "step": 8380
    },
    {
      "entropy": 6.304123020172119,
      "epoch": 0.9676860934795153,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917979108253709,
      "loss": 6.0699,
      "mean_token_accuracy": 0.1677810087800026,
      "num_tokens": 21259800.0,
      "step": 8385
    },
    {
      "entropy": 6.304227828979492,
      "epoch": 0.9682631275245239,
      "grad_norm": 0.875,
      "learning_rate": 0.00049178686724282,
      "loss": 6.0953,
      "mean_token_accuracy": 0.170858334004879,
      "num_tokens": 21271776.0,
      "step": 8390
    },
    {
      "entropy": 6.274011754989624,
      "epoch": 0.9688401615695326,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004917758163687506,
      "loss": 5.9803,
      "mean_token_accuracy": 0.17564422339200975,
      "num_tokens": 21285368.0,
      "step": 8395
    },
    {
      "entropy": 6.191785097122192,
      "epoch": 0.9694171956145412,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004917647582035341,
      "loss": 5.9786,
      "mean_token_accuracy": 0.17851267009973526,
      "num_tokens": 21299077.0,
      "step": 8400
    },
    {
      "entropy": 6.300303649902344,
      "epoch": 0.96999422965955,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004917536927475428,
      "loss": 6.0761,
      "mean_token_accuracy": 0.17253753840923308,
      "num_tokens": 21312379.0,
      "step": 8405
    },
    {
      "entropy": 6.277495002746582,
      "epoch": 0.9705712637045586,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004917426200011486,
      "loss": 6.0348,
      "mean_token_accuracy": 0.1745515689253807,
      "num_tokens": 21325530.0,
      "step": 8410
    },
    {
      "entropy": 6.155578470230102,
      "epoch": 0.9711482977495672,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000491731539964724,
      "loss": 5.9194,
      "mean_token_accuracy": 0.18623323291540145,
      "num_tokens": 21339200.0,
      "step": 8415
    },
    {
      "entropy": 6.288159465789795,
      "epoch": 0.9717253317945759,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004917204526386418,
      "loss": 5.9781,
      "mean_token_accuracy": 0.18108310252428056,
      "num_tokens": 21351072.0,
      "step": 8420
    },
    {
      "entropy": 6.213099575042724,
      "epoch": 0.9723023658395845,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004917093580232748,
      "loss": 6.0366,
      "mean_token_accuracy": 0.17771366089582444,
      "num_tokens": 21363691.0,
      "step": 8425
    },
    {
      "entropy": 6.306379938125611,
      "epoch": 0.9728793998845932,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004916982561189962,
      "loss": 6.0446,
      "mean_token_accuracy": 0.16407431066036224,
      "num_tokens": 21376896.0,
      "step": 8430
    },
    {
      "entropy": 6.317812871932984,
      "epoch": 0.9734564339296018,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004916871469261794,
      "loss": 6.0712,
      "mean_token_accuracy": 0.1748368188738823,
      "num_tokens": 21388590.0,
      "step": 8435
    },
    {
      "entropy": 6.295198154449463,
      "epoch": 0.9740334679746105,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000491676030445198,
      "loss": 6.1191,
      "mean_token_accuracy": 0.17043811678886414,
      "num_tokens": 21401912.0,
      "step": 8440
    },
    {
      "entropy": 6.25452184677124,
      "epoch": 0.9746105020196192,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004916649066764259,
      "loss": 6.0693,
      "mean_token_accuracy": 0.17368588596582413,
      "num_tokens": 21415472.0,
      "step": 8445
    },
    {
      "entropy": 6.265368223190308,
      "epoch": 0.9751875360646278,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916537756202375,
      "loss": 6.0291,
      "mean_token_accuracy": 0.17431591004133223,
      "num_tokens": 21426758.0,
      "step": 8450
    },
    {
      "entropy": 6.249980115890503,
      "epoch": 0.9757645701096365,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004916426372770069,
      "loss": 6.0221,
      "mean_token_accuracy": 0.1822528839111328,
      "num_tokens": 21439347.0,
      "step": 8455
    },
    {
      "entropy": 6.344882678985596,
      "epoch": 0.9763416041546451,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004916314916471087,
      "loss": 6.1197,
      "mean_token_accuracy": 0.16698339730501174,
      "num_tokens": 21451359.0,
      "step": 8460
    },
    {
      "entropy": 6.38165602684021,
      "epoch": 0.9769186381996537,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004916203387309179,
      "loss": 6.1209,
      "mean_token_accuracy": 0.16740904450416566,
      "num_tokens": 21464926.0,
      "step": 8465
    },
    {
      "entropy": 6.275937747955322,
      "epoch": 0.9774956722446624,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004916091785288096,
      "loss": 6.0692,
      "mean_token_accuracy": 0.1693897143006325,
      "num_tokens": 21477234.0,
      "step": 8470
    },
    {
      "entropy": 6.180343294143677,
      "epoch": 0.9780727062896711,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004915980110411593,
      "loss": 5.9165,
      "mean_token_accuracy": 0.18313064575195312,
      "num_tokens": 21488751.0,
      "step": 8475
    },
    {
      "entropy": 6.286865949630737,
      "epoch": 0.9786497403346798,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004915868362683425,
      "loss": 6.0493,
      "mean_token_accuracy": 0.17128912955522538,
      "num_tokens": 21500532.0,
      "step": 8480
    },
    {
      "entropy": 6.272981214523315,
      "epoch": 0.9792267743796884,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004915756542107349,
      "loss": 6.0414,
      "mean_token_accuracy": 0.1687467187643051,
      "num_tokens": 21513725.0,
      "step": 8485
    },
    {
      "entropy": 6.193865776062012,
      "epoch": 0.979803808424697,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004915644648687127,
      "loss": 5.9724,
      "mean_token_accuracy": 0.18019705414772033,
      "num_tokens": 21526886.0,
      "step": 8490
    },
    {
      "entropy": 6.174581098556518,
      "epoch": 0.9803808424697057,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004915532682426524,
      "loss": 5.9082,
      "mean_token_accuracy": 0.18099150508642198,
      "num_tokens": 21539278.0,
      "step": 8495
    },
    {
      "entropy": 6.3271393299102785,
      "epoch": 0.9809578765147143,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004915420643329306,
      "loss": 6.1483,
      "mean_token_accuracy": 0.16990142911672593,
      "num_tokens": 21551864.0,
      "step": 8500
    },
    {
      "entropy": 6.335036182403565,
      "epoch": 0.981534910559723,
      "grad_norm": 0.828125,
      "learning_rate": 0.000491530853139924,
      "loss": 6.0726,
      "mean_token_accuracy": 0.1718694746494293,
      "num_tokens": 21564398.0,
      "step": 8505
    },
    {
      "entropy": 6.335596990585327,
      "epoch": 0.9821119446047317,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004915196346640095,
      "loss": 6.0901,
      "mean_token_accuracy": 0.1714094787836075,
      "num_tokens": 21575998.0,
      "step": 8510
    },
    {
      "entropy": 6.278130578994751,
      "epoch": 0.9826889786497404,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915084089055648,
      "loss": 6.054,
      "mean_token_accuracy": 0.1700270876288414,
      "num_tokens": 21589243.0,
      "step": 8515
    },
    {
      "entropy": 6.203781700134277,
      "epoch": 0.983266012694749,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004914971758649673,
      "loss": 5.9992,
      "mean_token_accuracy": 0.1781109556555748,
      "num_tokens": 21601229.0,
      "step": 8520
    },
    {
      "entropy": 6.335890817642212,
      "epoch": 0.9838430467397576,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004914859355425948,
      "loss": 6.0358,
      "mean_token_accuracy": 0.17270396500825883,
      "num_tokens": 21613270.0,
      "step": 8525
    },
    {
      "entropy": 6.192272806167603,
      "epoch": 0.9844200807847663,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004914746879388255,
      "loss": 5.9736,
      "mean_token_accuracy": 0.17878717929124832,
      "num_tokens": 21626463.0,
      "step": 8530
    },
    {
      "entropy": 6.245885515213013,
      "epoch": 0.9849971148297749,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004914634330540373,
      "loss": 6.0914,
      "mean_token_accuracy": 0.17799934148788452,
      "num_tokens": 21638172.0,
      "step": 8535
    },
    {
      "entropy": 6.308247900009155,
      "epoch": 0.9855741488747836,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004914521708886093,
      "loss": 6.0125,
      "mean_token_accuracy": 0.17956866174936295,
      "num_tokens": 21650015.0,
      "step": 8540
    },
    {
      "entropy": 6.33227424621582,
      "epoch": 0.9861511829197923,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004914409014429201,
      "loss": 6.1115,
      "mean_token_accuracy": 0.1672136142849922,
      "num_tokens": 21662502.0,
      "step": 8545
    },
    {
      "entropy": 6.221246337890625,
      "epoch": 0.986728216964801,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004914296247173486,
      "loss": 6.0021,
      "mean_token_accuracy": 0.18216463774442673,
      "num_tokens": 21675691.0,
      "step": 8550
    },
    {
      "entropy": 6.290973472595215,
      "epoch": 0.9873052510098096,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004914183407122741,
      "loss": 6.0722,
      "mean_token_accuracy": 0.16632219403982162,
      "num_tokens": 21687904.0,
      "step": 8555
    },
    {
      "entropy": 6.309543466567993,
      "epoch": 0.9878822850548182,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004914070494280763,
      "loss": 6.1354,
      "mean_token_accuracy": 0.16766521334648132,
      "num_tokens": 21700371.0,
      "step": 8560
    },
    {
      "entropy": 6.235837602615357,
      "epoch": 0.9884593190998269,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004913957508651349,
      "loss": 6.0335,
      "mean_token_accuracy": 0.18258669674396516,
      "num_tokens": 21715235.0,
      "step": 8565
    },
    {
      "entropy": 6.253808879852295,
      "epoch": 0.9890363531448355,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004913844450238299,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17600261121988298,
      "num_tokens": 21727994.0,
      "step": 8570
    },
    {
      "entropy": 6.281888723373413,
      "epoch": 0.9896133871898442,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004913731319045416,
      "loss": 6.0036,
      "mean_token_accuracy": 0.17418931275606156,
      "num_tokens": 21739894.0,
      "step": 8575
    },
    {
      "entropy": 6.280125093460083,
      "epoch": 0.9901904212348529,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004913618115076505,
      "loss": 6.1333,
      "mean_token_accuracy": 0.16924936175346375,
      "num_tokens": 21752957.0,
      "step": 8580
    },
    {
      "entropy": 6.270187520980835,
      "epoch": 0.9907674552798615,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004913504838335372,
      "loss": 6.0117,
      "mean_token_accuracy": 0.17379884719848632,
      "num_tokens": 21765547.0,
      "step": 8585
    },
    {
      "entropy": 6.266125345230103,
      "epoch": 0.9913444893248702,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004913391488825829,
      "loss": 6.104,
      "mean_token_accuracy": 0.17694249451160432,
      "num_tokens": 21778817.0,
      "step": 8590
    },
    {
      "entropy": 6.2261824131011965,
      "epoch": 0.9919215233698788,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004913278066551689,
      "loss": 6.0004,
      "mean_token_accuracy": 0.1725752294063568,
      "num_tokens": 21791002.0,
      "step": 8595
    },
    {
      "entropy": 6.2803973197937015,
      "epoch": 0.9924985574148875,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913164571516765,
      "loss": 6.0834,
      "mean_token_accuracy": 0.17212606966495514,
      "num_tokens": 21803877.0,
      "step": 8600
    },
    {
      "entropy": 6.300022172927856,
      "epoch": 0.9930755914598961,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004913051003724876,
      "loss": 5.9912,
      "mean_token_accuracy": 0.1742602214217186,
      "num_tokens": 21815884.0,
      "step": 8605
    },
    {
      "entropy": 6.240463399887085,
      "epoch": 0.9936526255049047,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912937363179839,
      "loss": 6.0434,
      "mean_token_accuracy": 0.17520978897809983,
      "num_tokens": 21826990.0,
      "step": 8610
    },
    {
      "entropy": 6.244440984725952,
      "epoch": 0.9942296595499135,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491282364988548,
      "loss": 6.0255,
      "mean_token_accuracy": 0.17350778728723526,
      "num_tokens": 21838835.0,
      "step": 8615
    },
    {
      "entropy": 6.383248662948608,
      "epoch": 0.9948066935949221,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912709863845621,
      "loss": 6.1339,
      "mean_token_accuracy": 0.17144082933664323,
      "num_tokens": 21850924.0,
      "step": 8620
    },
    {
      "entropy": 6.318079805374145,
      "epoch": 0.9953837276399308,
      "grad_norm": 0.875,
      "learning_rate": 0.000491259600506409,
      "loss": 6.1002,
      "mean_token_accuracy": 0.17307236343622207,
      "num_tokens": 21862720.0,
      "step": 8625
    },
    {
      "entropy": 6.25905818939209,
      "epoch": 0.9959607616849394,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004912482073544716,
      "loss": 6.0542,
      "mean_token_accuracy": 0.17808614820241928,
      "num_tokens": 21874655.0,
      "step": 8630
    },
    {
      "entropy": 6.23261022567749,
      "epoch": 0.996537795729948,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912368069291333,
      "loss": 6.0173,
      "mean_token_accuracy": 0.1786980226635933,
      "num_tokens": 21886911.0,
      "step": 8635
    },
    {
      "entropy": 6.205492162704468,
      "epoch": 0.9971148297749567,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004912253992307773,
      "loss": 5.9927,
      "mean_token_accuracy": 0.18051304966211318,
      "num_tokens": 21899614.0,
      "step": 8640
    },
    {
      "entropy": 6.31227822303772,
      "epoch": 0.9976918638199653,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004912139842597875,
      "loss": 6.0011,
      "mean_token_accuracy": 0.17408534735441208,
      "num_tokens": 21912756.0,
      "step": 8645
    },
    {
      "entropy": 6.221034145355224,
      "epoch": 0.9982688978649741,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004912025620165477,
      "loss": 6.0133,
      "mean_token_accuracy": 0.17002497762441635,
      "num_tokens": 21924911.0,
      "step": 8650
    },
    {
      "entropy": 6.188537883758545,
      "epoch": 0.9988459319099827,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004911911325014421,
      "loss": 5.9556,
      "mean_token_accuracy": 0.1789535701274872,
      "num_tokens": 21938204.0,
      "step": 8655
    },
    {
      "entropy": 6.342658185958863,
      "epoch": 0.9994229659549914,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004911796957148552,
      "loss": 6.041,
      "mean_token_accuracy": 0.17369647771120073,
      "num_tokens": 21951285.0,
      "step": 8660
    },
    {
      "entropy": 6.299900817871094,
      "epoch": 1.0,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004911682516571715,
      "loss": 6.0679,
      "mean_token_accuracy": 0.17429974675178528,
      "num_tokens": 21963117.0,
      "step": 8665
    },
    {
      "entropy": 6.271995592117309,
      "epoch": 1.0005770340450086,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004911568003287761,
      "loss": 5.9582,
      "mean_token_accuracy": 0.17250487804412842,
      "num_tokens": 21976103.0,
      "step": 8670
    },
    {
      "entropy": 6.1640949726104735,
      "epoch": 1.0011540680900173,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004911453417300541,
      "loss": 5.8559,
      "mean_token_accuracy": 0.19138497710227967,
      "num_tokens": 21987905.0,
      "step": 8675
    },
    {
      "entropy": 6.174138498306275,
      "epoch": 1.001731102135026,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004911338758613909,
      "loss": 5.9414,
      "mean_token_accuracy": 0.1773781344294548,
      "num_tokens": 22000872.0,
      "step": 8680
    },
    {
      "entropy": 6.288146829605102,
      "epoch": 1.0023081361800346,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004911224027231722,
      "loss": 5.9653,
      "mean_token_accuracy": 0.17484614700078965,
      "num_tokens": 22012725.0,
      "step": 8685
    },
    {
      "entropy": 6.250091552734375,
      "epoch": 1.0028851702250432,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004911109223157838,
      "loss": 5.9012,
      "mean_token_accuracy": 0.18360436111688613,
      "num_tokens": 22024802.0,
      "step": 8690
    },
    {
      "entropy": 6.269646549224854,
      "epoch": 1.0034622042700518,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004910994346396118,
      "loss": 5.977,
      "mean_token_accuracy": 0.17402059137821196,
      "num_tokens": 22037481.0,
      "step": 8695
    },
    {
      "entropy": 6.29124493598938,
      "epoch": 1.0040392383150605,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910879396950427,
      "loss": 5.9727,
      "mean_token_accuracy": 0.17953110933303834,
      "num_tokens": 22049387.0,
      "step": 8700
    },
    {
      "entropy": 6.334495115280151,
      "epoch": 1.0046162723600693,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000491076437482463,
      "loss": 5.9789,
      "mean_token_accuracy": 0.1691194236278534,
      "num_tokens": 22061323.0,
      "step": 8705
    },
    {
      "entropy": 6.277500820159912,
      "epoch": 1.005193306405078,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004910649280022599,
      "loss": 5.8992,
      "mean_token_accuracy": 0.18170523196458815,
      "num_tokens": 22073163.0,
      "step": 8710
    },
    {
      "entropy": 6.146124935150146,
      "epoch": 1.0057703404500866,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004910534112548201,
      "loss": 5.9132,
      "mean_token_accuracy": 0.17684150934219361,
      "num_tokens": 22085583.0,
      "step": 8715
    },
    {
      "entropy": 6.330292797088623,
      "epoch": 1.0063473744950953,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004910418872405312,
      "loss": 6.0036,
      "mean_token_accuracy": 0.17630979269742966,
      "num_tokens": 22098241.0,
      "step": 8720
    },
    {
      "entropy": 6.291881656646728,
      "epoch": 1.006924408540104,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004910303559597806,
      "loss": 5.8999,
      "mean_token_accuracy": 0.1765839233994484,
      "num_tokens": 22111626.0,
      "step": 8725
    },
    {
      "entropy": 6.200822734832764,
      "epoch": 1.0075014425851125,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004910188174129564,
      "loss": 5.9235,
      "mean_token_accuracy": 0.17677416950464248,
      "num_tokens": 22124653.0,
      "step": 8730
    },
    {
      "entropy": 6.312221431732178,
      "epoch": 1.0080784766301212,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004910072716004466,
      "loss": 5.9421,
      "mean_token_accuracy": 0.17987631410360336,
      "num_tokens": 22137154.0,
      "step": 8735
    },
    {
      "entropy": 6.337836599349975,
      "epoch": 1.0086555106751298,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004909957185226394,
      "loss": 6.0464,
      "mean_token_accuracy": 0.16782609969377518,
      "num_tokens": 22149374.0,
      "step": 8740
    },
    {
      "entropy": 6.2611909866333,
      "epoch": 1.0092325447201385,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004909841581799236,
      "loss": 6.0153,
      "mean_token_accuracy": 0.17667713165283203,
      "num_tokens": 22161988.0,
      "step": 8745
    },
    {
      "entropy": 6.318277978897095,
      "epoch": 1.009809578765147,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004909725905726879,
      "loss": 6.0643,
      "mean_token_accuracy": 0.17233245372772216,
      "num_tokens": 22173490.0,
      "step": 8750
    },
    {
      "entropy": 6.302417469024658,
      "epoch": 1.0103866128101557,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004909610157013214,
      "loss": 5.9981,
      "mean_token_accuracy": 0.17494946718215942,
      "num_tokens": 22185065.0,
      "step": 8755
    },
    {
      "entropy": 6.222137498855591,
      "epoch": 1.0109636468551644,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004909494335662134,
      "loss": 5.9457,
      "mean_token_accuracy": 0.1837028831243515,
      "num_tokens": 22198325.0,
      "step": 8760
    },
    {
      "entropy": 6.141578102111817,
      "epoch": 1.011540680900173,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004909378441677535,
      "loss": 5.8738,
      "mean_token_accuracy": 0.18548515439033508,
      "num_tokens": 22211827.0,
      "step": 8765
    },
    {
      "entropy": 6.244017887115478,
      "epoch": 1.0121177149451817,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004909262475063314,
      "loss": 5.9302,
      "mean_token_accuracy": 0.176451376080513,
      "num_tokens": 22223411.0,
      "step": 8770
    },
    {
      "entropy": 6.229418849945068,
      "epoch": 1.0126947489901905,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004909146435823373,
      "loss": 5.9406,
      "mean_token_accuracy": 0.17601740062236787,
      "num_tokens": 22237752.0,
      "step": 8775
    },
    {
      "entropy": 6.316747760772705,
      "epoch": 1.0132717830351992,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004909030323961613,
      "loss": 5.9883,
      "mean_token_accuracy": 0.17417865097522736,
      "num_tokens": 22249778.0,
      "step": 8780
    },
    {
      "entropy": 6.272009181976318,
      "epoch": 1.0138488170802078,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004908914139481942,
      "loss": 5.9631,
      "mean_token_accuracy": 0.17759780436754227,
      "num_tokens": 22262514.0,
      "step": 8785
    },
    {
      "entropy": 6.274535417556763,
      "epoch": 1.0144258511252164,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004908797882388265,
      "loss": 5.9212,
      "mean_token_accuracy": 0.17650485783815384,
      "num_tokens": 22273662.0,
      "step": 8790
    },
    {
      "entropy": 6.2734862804412845,
      "epoch": 1.015002885170225,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004908681552684495,
      "loss": 6.0197,
      "mean_token_accuracy": 0.17017416656017303,
      "num_tokens": 22286213.0,
      "step": 8795
    },
    {
      "entropy": 6.206152629852295,
      "epoch": 1.0155799192152337,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004908565150374542,
      "loss": 5.8059,
      "mean_token_accuracy": 0.18659797310829163,
      "num_tokens": 22299191.0,
      "step": 8800
    },
    {
      "entropy": 6.220057153701783,
      "epoch": 1.0161569532602424,
      "grad_norm": 0.875,
      "learning_rate": 0.0004908448675462323,
      "loss": 5.9168,
      "mean_token_accuracy": 0.18124910145998002,
      "num_tokens": 22312047.0,
      "step": 8805
    },
    {
      "entropy": 6.246239233016968,
      "epoch": 1.016733987305251,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004908332127951756,
      "loss": 6.0308,
      "mean_token_accuracy": 0.17177972793579102,
      "num_tokens": 22326069.0,
      "step": 8810
    },
    {
      "entropy": 6.313067102432251,
      "epoch": 1.0173110213502596,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004908215507846757,
      "loss": 5.9858,
      "mean_token_accuracy": 0.17213782370090486,
      "num_tokens": 22338593.0,
      "step": 8815
    },
    {
      "entropy": 6.271481513977051,
      "epoch": 1.0178880553952683,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004908098815151254,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1799215242266655,
      "num_tokens": 22352286.0,
      "step": 8820
    },
    {
      "entropy": 6.271336078643799,
      "epoch": 1.018465089440277,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004907982049869168,
      "loss": 5.9306,
      "mean_token_accuracy": 0.1777525871992111,
      "num_tokens": 22364270.0,
      "step": 8825
    },
    {
      "entropy": 6.226083993911743,
      "epoch": 1.0190421234852856,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004907865212004428,
      "loss": 5.9637,
      "mean_token_accuracy": 0.17348791360855104,
      "num_tokens": 22377799.0,
      "step": 8830
    },
    {
      "entropy": 6.261759424209595,
      "epoch": 1.0196191575302942,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004907748301560963,
      "loss": 5.9771,
      "mean_token_accuracy": 0.17696553617715835,
      "num_tokens": 22389913.0,
      "step": 8835
    },
    {
      "entropy": 6.26409945487976,
      "epoch": 1.0201961915753028,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004907631318542707,
      "loss": 5.9988,
      "mean_token_accuracy": 0.17762383669614792,
      "num_tokens": 22401545.0,
      "step": 8840
    },
    {
      "entropy": 6.210157775878907,
      "epoch": 1.0207732256203117,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004907514262953594,
      "loss": 5.8914,
      "mean_token_accuracy": 0.1783894494175911,
      "num_tokens": 22414084.0,
      "step": 8845
    },
    {
      "entropy": 6.22634482383728,
      "epoch": 1.0213502596653203,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004907397134797559,
      "loss": 5.8787,
      "mean_token_accuracy": 0.18252388536930084,
      "num_tokens": 22427406.0,
      "step": 8850
    },
    {
      "entropy": 6.323730850219727,
      "epoch": 1.021927293710329,
      "grad_norm": 1.125,
      "learning_rate": 0.0004907279934078543,
      "loss": 5.9997,
      "mean_token_accuracy": 0.1736849933862686,
      "num_tokens": 22441118.0,
      "step": 8855
    },
    {
      "entropy": 6.31430492401123,
      "epoch": 1.0225043277553376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907162660800488,
      "loss": 6.0627,
      "mean_token_accuracy": 0.17202019542455674,
      "num_tokens": 22453799.0,
      "step": 8860
    },
    {
      "entropy": 6.3228006839752195,
      "epoch": 1.0230813618003463,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004907045314967338,
      "loss": 6.0158,
      "mean_token_accuracy": 0.1706335499882698,
      "num_tokens": 22466573.0,
      "step": 8865
    },
    {
      "entropy": 6.205880975723266,
      "epoch": 1.023658395845355,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004906927896583041,
      "loss": 5.8774,
      "mean_token_accuracy": 0.18316041827201843,
      "num_tokens": 22478968.0,
      "step": 8870
    },
    {
      "entropy": 6.245298290252686,
      "epoch": 1.0242354298903635,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004906810405651546,
      "loss": 5.885,
      "mean_token_accuracy": 0.18523868173360825,
      "num_tokens": 22490669.0,
      "step": 8875
    },
    {
      "entropy": 6.280638694763184,
      "epoch": 1.0248124639353722,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906692842176803,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1728905886411667,
      "num_tokens": 22502235.0,
      "step": 8880
    },
    {
      "entropy": 6.2999005794525145,
      "epoch": 1.0253894979803808,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004906575206162769,
      "loss": 5.958,
      "mean_token_accuracy": 0.17361630797386168,
      "num_tokens": 22514535.0,
      "step": 8885
    },
    {
      "entropy": 6.253720092773437,
      "epoch": 1.0259665320253895,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004906457497613399,
      "loss": 5.9391,
      "mean_token_accuracy": 0.17980763763189317,
      "num_tokens": 22527773.0,
      "step": 8890
    },
    {
      "entropy": 6.2260314464569095,
      "epoch": 1.026543566070398,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004906339716532651,
      "loss": 5.9518,
      "mean_token_accuracy": 0.1829774111509323,
      "num_tokens": 22541853.0,
      "step": 8895
    },
    {
      "entropy": 6.360072422027588,
      "epoch": 1.0271206001154067,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004906221862924488,
      "loss": 6.0153,
      "mean_token_accuracy": 0.16649125069379805,
      "num_tokens": 22554473.0,
      "step": 8900
    },
    {
      "entropy": 6.310372686386108,
      "epoch": 1.0276976341604154,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004906103936792875,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17342302650213243,
      "num_tokens": 22567107.0,
      "step": 8905
    },
    {
      "entropy": 6.255186080932617,
      "epoch": 1.028274668205424,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004905985938141777,
      "loss": 6.1353,
      "mean_token_accuracy": 0.16475126296281814,
      "num_tokens": 22579806.0,
      "step": 8910
    },
    {
      "entropy": 6.290994596481323,
      "epoch": 1.0288517022504329,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004905867866975163,
      "loss": 5.9484,
      "mean_token_accuracy": 0.17861126959323884,
      "num_tokens": 22593460.0,
      "step": 8915
    },
    {
      "entropy": 6.27809362411499,
      "epoch": 1.0294287362954415,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004905749723297003,
      "loss": 6.0062,
      "mean_token_accuracy": 0.17625349313020705,
      "num_tokens": 22607113.0,
      "step": 8920
    },
    {
      "entropy": 6.285610055923462,
      "epoch": 1.0300057703404502,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004905631507111271,
      "loss": 5.9193,
      "mean_token_accuracy": 0.17798546701669693,
      "num_tokens": 22619749.0,
      "step": 8925
    },
    {
      "entropy": 6.206954050064087,
      "epoch": 1.0305828043854588,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004905513218421946,
      "loss": 5.8888,
      "mean_token_accuracy": 0.18107479810714722,
      "num_tokens": 22632584.0,
      "step": 8930
    },
    {
      "entropy": 6.265821838378907,
      "epoch": 1.0311598384304674,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004905394857233004,
      "loss": 5.9746,
      "mean_token_accuracy": 0.17853020131587982,
      "num_tokens": 22645181.0,
      "step": 8935
    },
    {
      "entropy": 6.315801954269409,
      "epoch": 1.031736872475476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905276423548426,
      "loss": 5.9959,
      "mean_token_accuracy": 0.1747600942850113,
      "num_tokens": 22657666.0,
      "step": 8940
    },
    {
      "entropy": 6.293545866012574,
      "epoch": 1.0323139065204847,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004905157917372197,
      "loss": 6.0498,
      "mean_token_accuracy": 0.16805843263864517,
      "num_tokens": 22670404.0,
      "step": 8945
    },
    {
      "entropy": 6.278645467758179,
      "epoch": 1.0328909405654934,
      "grad_norm": 0.875,
      "learning_rate": 0.0004905039338708302,
      "loss": 6.038,
      "mean_token_accuracy": 0.1695145219564438,
      "num_tokens": 22683558.0,
      "step": 8950
    },
    {
      "entropy": 6.292709922790527,
      "epoch": 1.033467974610502,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904920687560728,
      "loss": 5.8988,
      "mean_token_accuracy": 0.18147481679916383,
      "num_tokens": 22694907.0,
      "step": 8955
    },
    {
      "entropy": 6.294043922424317,
      "epoch": 1.0340450086555106,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004904801963933469,
      "loss": 6.0771,
      "mean_token_accuracy": 0.16681005507707597,
      "num_tokens": 22707447.0,
      "step": 8960
    },
    {
      "entropy": 6.271177196502686,
      "epoch": 1.0346220427005193,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004904683167830515,
      "loss": 5.9436,
      "mean_token_accuracy": 0.1795396849513054,
      "num_tokens": 22719150.0,
      "step": 8965
    },
    {
      "entropy": 6.298140621185302,
      "epoch": 1.035199076745528,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004904564299255862,
      "loss": 6.0155,
      "mean_token_accuracy": 0.1723608747124672,
      "num_tokens": 22732489.0,
      "step": 8970
    },
    {
      "entropy": 6.189728879928589,
      "epoch": 1.0357761107905366,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904445358213511,
      "loss": 5.8812,
      "mean_token_accuracy": 0.1754599243402481,
      "num_tokens": 22745543.0,
      "step": 8975
    },
    {
      "entropy": 6.282424354553223,
      "epoch": 1.0363531448355452,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004904326344707461,
      "loss": 5.9783,
      "mean_token_accuracy": 0.17686663717031478,
      "num_tokens": 22757643.0,
      "step": 8980
    },
    {
      "entropy": 6.321871137619018,
      "epoch": 1.036930178880554,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004904207258741712,
      "loss": 5.971,
      "mean_token_accuracy": 0.17963920831680297,
      "num_tokens": 22770707.0,
      "step": 8985
    },
    {
      "entropy": 6.256516742706299,
      "epoch": 1.0375072129255627,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004904088100320274,
      "loss": 5.9979,
      "mean_token_accuracy": 0.1751512423157692,
      "num_tokens": 22782630.0,
      "step": 8990
    },
    {
      "entropy": 6.2503931522369385,
      "epoch": 1.0380842469705713,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004903968869447151,
      "loss": 6.0444,
      "mean_token_accuracy": 0.17432793825864792,
      "num_tokens": 22795693.0,
      "step": 8995
    },
    {
      "entropy": 6.373712539672852,
      "epoch": 1.03866128101558,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004903849566126356,
      "loss": 6.0568,
      "mean_token_accuracy": 0.1701628342270851,
      "num_tokens": 22807715.0,
      "step": 9000
    },
    {
      "epoch": 1.03866128101558,
      "eval_entropy": 6.111053488421989,
      "eval_loss": 6.069920539855957,
      "eval_mean_token_accuracy": 0.17735485498043227,
      "eval_num_tokens": 22807715.0,
      "eval_runtime": 17.4995,
      "eval_samples_per_second": 1607.814,
      "eval_steps_per_second": 200.977,
      "step": 9000
    },
    {
      "entropy": 6.238695287704468,
      "epoch": 1.0392383150605886,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903730190361902,
      "loss": 5.9607,
      "mean_token_accuracy": 0.17794350385665894,
      "num_tokens": 22820133.0,
      "step": 9005
    },
    {
      "entropy": 6.164699935913086,
      "epoch": 1.0398153491055973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049036107421578,
      "loss": 5.9384,
      "mean_token_accuracy": 0.18351739048957824,
      "num_tokens": 22833928.0,
      "step": 9010
    },
    {
      "entropy": 6.231976461410523,
      "epoch": 1.040392383150606,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903491221518073,
      "loss": 5.8808,
      "mean_token_accuracy": 0.18397735953330993,
      "num_tokens": 22847883.0,
      "step": 9015
    },
    {
      "entropy": 6.303796625137329,
      "epoch": 1.0409694171956145,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004903371628446737,
      "loss": 6.0043,
      "mean_token_accuracy": 0.17286169528961182,
      "num_tokens": 22861540.0,
      "step": 9020
    },
    {
      "entropy": 6.285726165771484,
      "epoch": 1.0415464512406232,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903251962947817,
      "loss": 5.9635,
      "mean_token_accuracy": 0.17470211386680604,
      "num_tokens": 22874778.0,
      "step": 9025
    },
    {
      "entropy": 6.251114177703857,
      "epoch": 1.0421234852856318,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004903132225025335,
      "loss": 6.0005,
      "mean_token_accuracy": 0.1789627730846405,
      "num_tokens": 22886868.0,
      "step": 9030
    },
    {
      "entropy": 6.249149990081787,
      "epoch": 1.0427005193306405,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004903012414683322,
      "loss": 5.9499,
      "mean_token_accuracy": 0.1717926323413849,
      "num_tokens": 22899549.0,
      "step": 9035
    },
    {
      "entropy": 6.312658357620239,
      "epoch": 1.043277553375649,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004902892531925805,
      "loss": 5.938,
      "mean_token_accuracy": 0.1823698177933693,
      "num_tokens": 22912301.0,
      "step": 9040
    },
    {
      "entropy": 6.236099672317505,
      "epoch": 1.0438545874206577,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004902772576756818,
      "loss": 5.9579,
      "mean_token_accuracy": 0.17695192545652388,
      "num_tokens": 22924784.0,
      "step": 9045
    },
    {
      "entropy": 6.318867301940918,
      "epoch": 1.0444316214656664,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004902652549180394,
      "loss": 5.9931,
      "mean_token_accuracy": 0.17314996272325517,
      "num_tokens": 22937095.0,
      "step": 9050
    },
    {
      "entropy": 6.260589694976806,
      "epoch": 1.0450086555106752,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004902532449200571,
      "loss": 5.9055,
      "mean_token_accuracy": 0.18783535063266754,
      "num_tokens": 22951385.0,
      "step": 9055
    },
    {
      "entropy": 6.253799724578857,
      "epoch": 1.0455856895556839,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004902412276821386,
      "loss": 5.9815,
      "mean_token_accuracy": 0.17174628376960754,
      "num_tokens": 22965228.0,
      "step": 9060
    },
    {
      "entropy": 6.255420684814453,
      "epoch": 1.0461627236006925,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004902292032046887,
      "loss": 5.9789,
      "mean_token_accuracy": 0.17923253923654556,
      "num_tokens": 22978376.0,
      "step": 9065
    },
    {
      "entropy": 6.247670841217041,
      "epoch": 1.0467397576457012,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004902171714881112,
      "loss": 5.9685,
      "mean_token_accuracy": 0.177823606133461,
      "num_tokens": 22991572.0,
      "step": 9070
    },
    {
      "entropy": 6.275672149658203,
      "epoch": 1.0473167916907098,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004902051325328112,
      "loss": 5.913,
      "mean_token_accuracy": 0.1792851135134697,
      "num_tokens": 23004001.0,
      "step": 9075
    },
    {
      "entropy": 6.175007724761963,
      "epoch": 1.0478938257357184,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004901930863391934,
      "loss": 5.8908,
      "mean_token_accuracy": 0.18301753997802733,
      "num_tokens": 23016383.0,
      "step": 9080
    },
    {
      "entropy": 6.248683834075928,
      "epoch": 1.048470859780727,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000490181032907663,
      "loss": 6.0054,
      "mean_token_accuracy": 0.1760744646191597,
      "num_tokens": 23027836.0,
      "step": 9085
    },
    {
      "entropy": 6.274583148956299,
      "epoch": 1.0490478938257357,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004901689722386255,
      "loss": 5.9435,
      "mean_token_accuracy": 0.17630271166563033,
      "num_tokens": 23040083.0,
      "step": 9090
    },
    {
      "entropy": 6.245274591445923,
      "epoch": 1.0496249278707444,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901569043324864,
      "loss": 5.9585,
      "mean_token_accuracy": 0.17506776601076127,
      "num_tokens": 23051232.0,
      "step": 9095
    },
    {
      "entropy": 6.30389895439148,
      "epoch": 1.050201961915753,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004901448291896518,
      "loss": 6.0466,
      "mean_token_accuracy": 0.17152084708213805,
      "num_tokens": 23063355.0,
      "step": 9100
    },
    {
      "entropy": 6.1282142162322994,
      "epoch": 1.0507789959607616,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004901327468105277,
      "loss": 5.7307,
      "mean_token_accuracy": 0.19000066667795182,
      "num_tokens": 23076268.0,
      "step": 9105
    },
    {
      "entropy": 6.226790714263916,
      "epoch": 1.0513560300057703,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901206571955205,
      "loss": 5.8925,
      "mean_token_accuracy": 0.17903839647769929,
      "num_tokens": 23088753.0,
      "step": 9110
    },
    {
      "entropy": 6.22412486076355,
      "epoch": 1.051933064050779,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004901085603450369,
      "loss": 5.8371,
      "mean_token_accuracy": 0.1830509915947914,
      "num_tokens": 23100452.0,
      "step": 9115
    },
    {
      "entropy": 6.3040382862091064,
      "epoch": 1.0525100980957875,
      "grad_norm": 0.875,
      "learning_rate": 0.0004900964562594838,
      "loss": 5.9607,
      "mean_token_accuracy": 0.17867771685123443,
      "num_tokens": 23113368.0,
      "step": 9120
    },
    {
      "entropy": 6.1810688972473145,
      "epoch": 1.0530871321407964,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000490084344939268,
      "loss": 5.879,
      "mean_token_accuracy": 0.18695793747901918,
      "num_tokens": 23125940.0,
      "step": 9125
    },
    {
      "entropy": 6.1982675075531,
      "epoch": 1.053664166185805,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900722263847973,
      "loss": 5.9404,
      "mean_token_accuracy": 0.17940602749586104,
      "num_tokens": 23137282.0,
      "step": 9130
    },
    {
      "entropy": 6.339962768554687,
      "epoch": 1.0542412002308137,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004900601005964791,
      "loss": 6.0175,
      "mean_token_accuracy": 0.17628165930509568,
      "num_tokens": 23149569.0,
      "step": 9135
    },
    {
      "entropy": 6.248095893859864,
      "epoch": 1.0548182342758223,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004900479675747212,
      "loss": 5.9997,
      "mean_token_accuracy": 0.17854331880807878,
      "num_tokens": 23163290.0,
      "step": 9140
    },
    {
      "entropy": 6.253164529800415,
      "epoch": 1.055395268320831,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900358273199316,
      "loss": 5.9455,
      "mean_token_accuracy": 0.18596586138010024,
      "num_tokens": 23175092.0,
      "step": 9145
    },
    {
      "entropy": 6.251778507232666,
      "epoch": 1.0559723023658396,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000490023679832519,
      "loss": 5.9085,
      "mean_token_accuracy": 0.1864987775683403,
      "num_tokens": 23189064.0,
      "step": 9150
    },
    {
      "entropy": 6.272374963760376,
      "epoch": 1.0565493364108482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900115251128916,
      "loss": 5.9827,
      "mean_token_accuracy": 0.1733475297689438,
      "num_tokens": 23202619.0,
      "step": 9155
    },
    {
      "entropy": 6.248503303527832,
      "epoch": 1.057126370455857,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004899993631614583,
      "loss": 5.878,
      "mean_token_accuracy": 0.18134477585554123,
      "num_tokens": 23214575.0,
      "step": 9160
    },
    {
      "entropy": 6.313458633422852,
      "epoch": 1.0577034045008655,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004899871939786283,
      "loss": 6.0151,
      "mean_token_accuracy": 0.16623128354549407,
      "num_tokens": 23226526.0,
      "step": 9165
    },
    {
      "entropy": 6.203559637069702,
      "epoch": 1.0582804385458742,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004899750175648107,
      "loss": 5.9677,
      "mean_token_accuracy": 0.1791852667927742,
      "num_tokens": 23239942.0,
      "step": 9170
    },
    {
      "entropy": 6.271019649505615,
      "epoch": 1.0588574725908828,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004899628339204154,
      "loss": 5.9499,
      "mean_token_accuracy": 0.17290004342794418,
      "num_tokens": 23252798.0,
      "step": 9175
    },
    {
      "entropy": 6.23073182106018,
      "epoch": 1.0594345066358914,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004899506430458518,
      "loss": 5.8975,
      "mean_token_accuracy": 0.1814522922039032,
      "num_tokens": 23266195.0,
      "step": 9180
    },
    {
      "entropy": 6.150527381896973,
      "epoch": 1.0600115406809,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004899384449415302,
      "loss": 5.8518,
      "mean_token_accuracy": 0.18595437854528427,
      "num_tokens": 23279057.0,
      "step": 9185
    },
    {
      "entropy": 6.353490591049194,
      "epoch": 1.0605885747259087,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004899262396078606,
      "loss": 6.0096,
      "mean_token_accuracy": 0.1729474201798439,
      "num_tokens": 23292011.0,
      "step": 9190
    },
    {
      "entropy": 6.259584188461304,
      "epoch": 1.0611656087709176,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004899140270452539,
      "loss": 5.8957,
      "mean_token_accuracy": 0.1865168884396553,
      "num_tokens": 23303719.0,
      "step": 9195
    },
    {
      "entropy": 6.225439500808716,
      "epoch": 1.0617426428159262,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004899018072541204,
      "loss": 6.0134,
      "mean_token_accuracy": 0.17329226285219193,
      "num_tokens": 23317291.0,
      "step": 9200
    },
    {
      "entropy": 6.250166511535644,
      "epoch": 1.0623196768609349,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004898895802348715,
      "loss": 5.9503,
      "mean_token_accuracy": 0.17840168476104737,
      "num_tokens": 23329546.0,
      "step": 9205
    },
    {
      "entropy": 6.230170488357544,
      "epoch": 1.0628967109059435,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004898773459879183,
      "loss": 5.889,
      "mean_token_accuracy": 0.18481861054897308,
      "num_tokens": 23342171.0,
      "step": 9210
    },
    {
      "entropy": 6.214110469818115,
      "epoch": 1.0634737449509521,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004898651045136724,
      "loss": 5.9536,
      "mean_token_accuracy": 0.16962333470582963,
      "num_tokens": 23355400.0,
      "step": 9215
    },
    {
      "entropy": 6.293714904785157,
      "epoch": 1.0640507789959608,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004898528558125453,
      "loss": 5.9134,
      "mean_token_accuracy": 0.18154116123914718,
      "num_tokens": 23368011.0,
      "step": 9220
    },
    {
      "entropy": 6.196040391921997,
      "epoch": 1.0646278130409694,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004898405998849492,
      "loss": 5.9827,
      "mean_token_accuracy": 0.18173130601644516,
      "num_tokens": 23380938.0,
      "step": 9225
    },
    {
      "entropy": 6.306158065795898,
      "epoch": 1.065204847085978,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004898283367312962,
      "loss": 6.0392,
      "mean_token_accuracy": 0.1732558473944664,
      "num_tokens": 23392880.0,
      "step": 9230
    },
    {
      "entropy": 6.306704902648926,
      "epoch": 1.0657818811309867,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004898160663519988,
      "loss": 5.954,
      "mean_token_accuracy": 0.1785948559641838,
      "num_tokens": 23407809.0,
      "step": 9235
    },
    {
      "entropy": 6.292022180557251,
      "epoch": 1.0663589151759953,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898037887474697,
      "loss": 5.9722,
      "mean_token_accuracy": 0.17465235590934752,
      "num_tokens": 23419908.0,
      "step": 9240
    },
    {
      "entropy": 6.31259560585022,
      "epoch": 1.066935949221004,
      "grad_norm": 0.875,
      "learning_rate": 0.0004897915039181219,
      "loss": 6.0402,
      "mean_token_accuracy": 0.17240495830774308,
      "num_tokens": 23431743.0,
      "step": 9245
    },
    {
      "entropy": 6.356765460968018,
      "epoch": 1.0675129832660126,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004897792118643685,
      "loss": 5.9364,
      "mean_token_accuracy": 0.1779252678155899,
      "num_tokens": 23444234.0,
      "step": 9250
    },
    {
      "entropy": 6.241831493377686,
      "epoch": 1.0680900173110213,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004897669125866231,
      "loss": 5.9037,
      "mean_token_accuracy": 0.17907557785511016,
      "num_tokens": 23457393.0,
      "step": 9255
    },
    {
      "entropy": 6.186018705368042,
      "epoch": 1.06866705135603,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004897546060852993,
      "loss": 6.0073,
      "mean_token_accuracy": 0.17096005827188493,
      "num_tokens": 23471520.0,
      "step": 9260
    },
    {
      "entropy": 6.283358240127564,
      "epoch": 1.0692440854010385,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004897422923608108,
      "loss": 5.939,
      "mean_token_accuracy": 0.17451726645231247,
      "num_tokens": 23484211.0,
      "step": 9265
    },
    {
      "entropy": 6.303541088104248,
      "epoch": 1.0698211194460474,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004897299714135721,
      "loss": 6.001,
      "mean_token_accuracy": 0.1734578415751457,
      "num_tokens": 23497110.0,
      "step": 9270
    },
    {
      "entropy": 6.2064720630645756,
      "epoch": 1.070398153491056,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004897176432439974,
      "loss": 5.9615,
      "mean_token_accuracy": 0.17626977413892747,
      "num_tokens": 23508803.0,
      "step": 9275
    },
    {
      "entropy": 6.209423160552978,
      "epoch": 1.0709751875360647,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004897053078525016,
      "loss": 5.875,
      "mean_token_accuracy": 0.18555210083723067,
      "num_tokens": 23522209.0,
      "step": 9280
    },
    {
      "entropy": 6.298757934570313,
      "epoch": 1.0715522215810733,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004896929652394993,
      "loss": 5.9379,
      "mean_token_accuracy": 0.17657751739025115,
      "num_tokens": 23534728.0,
      "step": 9285
    },
    {
      "entropy": 6.208915424346924,
      "epoch": 1.072129255626082,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004896806154054057,
      "loss": 6.0239,
      "mean_token_accuracy": 0.1762393593788147,
      "num_tokens": 23548389.0,
      "step": 9290
    },
    {
      "entropy": 6.274517631530761,
      "epoch": 1.0727062896710906,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004896682583506363,
      "loss": 5.9138,
      "mean_token_accuracy": 0.1762578532099724,
      "num_tokens": 23560260.0,
      "step": 9295
    },
    {
      "entropy": 6.2781901359558105,
      "epoch": 1.0732833237160992,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004896558940756067,
      "loss": 5.9373,
      "mean_token_accuracy": 0.17987986654043198,
      "num_tokens": 23572771.0,
      "step": 9300
    },
    {
      "entropy": 6.25750150680542,
      "epoch": 1.0738603577611079,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004896435225807327,
      "loss": 5.9743,
      "mean_token_accuracy": 0.1788585141301155,
      "num_tokens": 23584769.0,
      "step": 9305
    },
    {
      "entropy": 6.243601226806641,
      "epoch": 1.0744373918061165,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004896311438664304,
      "loss": 5.9157,
      "mean_token_accuracy": 0.18181499391794204,
      "num_tokens": 23598912.0,
      "step": 9310
    },
    {
      "entropy": 6.265085172653198,
      "epoch": 1.0750144258511252,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004896187579331163,
      "loss": 5.9643,
      "mean_token_accuracy": 0.17543943077325821,
      "num_tokens": 23611781.0,
      "step": 9315
    },
    {
      "entropy": 6.238484287261963,
      "epoch": 1.0755914598961338,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004896063647812068,
      "loss": 6.004,
      "mean_token_accuracy": 0.17591085135936738,
      "num_tokens": 23624116.0,
      "step": 9320
    },
    {
      "entropy": 6.303693580627441,
      "epoch": 1.0761684939411424,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004895939644111189,
      "loss": 5.9993,
      "mean_token_accuracy": 0.17500557154417037,
      "num_tokens": 23637456.0,
      "step": 9325
    },
    {
      "entropy": 6.248747205734253,
      "epoch": 1.076745527986151,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895815568232694,
      "loss": 5.9724,
      "mean_token_accuracy": 0.17546537071466445,
      "num_tokens": 23649572.0,
      "step": 9330
    },
    {
      "entropy": 6.23223295211792,
      "epoch": 1.07732256203116,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004895691420180759,
      "loss": 5.9348,
      "mean_token_accuracy": 0.18016553670167923,
      "num_tokens": 23662934.0,
      "step": 9335
    },
    {
      "entropy": 6.185261869430542,
      "epoch": 1.0778995960761686,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000489556719995956,
      "loss": 5.8191,
      "mean_token_accuracy": 0.18600068539381026,
      "num_tokens": 23676258.0,
      "step": 9340
    },
    {
      "entropy": 6.313590621948242,
      "epoch": 1.0784766301211772,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004895442907573274,
      "loss": 5.9763,
      "mean_token_accuracy": 0.16796712577342987,
      "num_tokens": 23689361.0,
      "step": 9345
    },
    {
      "entropy": 6.226434850692749,
      "epoch": 1.0790536641661859,
      "grad_norm": 0.828125,
      "learning_rate": 0.000489531854302608,
      "loss": 5.922,
      "mean_token_accuracy": 0.1783987820148468,
      "num_tokens": 23702157.0,
      "step": 9350
    },
    {
      "entropy": 6.250939226150512,
      "epoch": 1.0796306982111945,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004895194106322164,
      "loss": 5.9617,
      "mean_token_accuracy": 0.17884741276502608,
      "num_tokens": 23713846.0,
      "step": 9355
    },
    {
      "entropy": 6.186641645431519,
      "epoch": 1.0802077322562031,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004895069597465709,
      "loss": 5.8775,
      "mean_token_accuracy": 0.18720127791166305,
      "num_tokens": 23727685.0,
      "step": 9360
    },
    {
      "entropy": 6.217104291915893,
      "epoch": 1.0807847663012118,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004894945016460904,
      "loss": 5.9061,
      "mean_token_accuracy": 0.1839689537882805,
      "num_tokens": 23740806.0,
      "step": 9365
    },
    {
      "entropy": 6.239234161376953,
      "epoch": 1.0813618003462204,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004894820363311938,
      "loss": 5.8817,
      "mean_token_accuracy": 0.1822696641087532,
      "num_tokens": 23752968.0,
      "step": 9370
    },
    {
      "entropy": 6.314483737945556,
      "epoch": 1.081938834391229,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004894695638023005,
      "loss": 6.0193,
      "mean_token_accuracy": 0.1787770375609398,
      "num_tokens": 23765141.0,
      "step": 9375
    },
    {
      "entropy": 6.237100076675415,
      "epoch": 1.0825158684362377,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048945708405983,
      "loss": 5.9004,
      "mean_token_accuracy": 0.17711787521839142,
      "num_tokens": 23778294.0,
      "step": 9380
    },
    {
      "entropy": 6.233566188812256,
      "epoch": 1.0830929024812463,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004894445971042019,
      "loss": 5.9295,
      "mean_token_accuracy": 0.18573664426803588,
      "num_tokens": 23790687.0,
      "step": 9385
    },
    {
      "entropy": 6.180399990081787,
      "epoch": 1.083669936526255,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004894321029358364,
      "loss": 5.9022,
      "mean_token_accuracy": 0.17696404159069062,
      "num_tokens": 23803569.0,
      "step": 9390
    },
    {
      "entropy": 6.214170026779175,
      "epoch": 1.0842469705712636,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004894196015551536,
      "loss": 5.9511,
      "mean_token_accuracy": 0.17121729403734207,
      "num_tokens": 23816693.0,
      "step": 9395
    },
    {
      "entropy": 6.28152756690979,
      "epoch": 1.0848240046162723,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000489407092962574,
      "loss": 5.9768,
      "mean_token_accuracy": 0.1818251922726631,
      "num_tokens": 23829139.0,
      "step": 9400
    },
    {
      "entropy": 6.264230871200562,
      "epoch": 1.085401038661281,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004893945771585183,
      "loss": 5.9816,
      "mean_token_accuracy": 0.18013782054185867,
      "num_tokens": 23842489.0,
      "step": 9405
    },
    {
      "entropy": 6.258744764328003,
      "epoch": 1.0859780727062898,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893820541434075,
      "loss": 5.9245,
      "mean_token_accuracy": 0.18056803643703462,
      "num_tokens": 23854555.0,
      "step": 9410
    },
    {
      "entropy": 6.28338360786438,
      "epoch": 1.0865551067512984,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004893695239176629,
      "loss": 5.9667,
      "mean_token_accuracy": 0.17003070265054704,
      "num_tokens": 23868294.0,
      "step": 9415
    },
    {
      "entropy": 6.231619834899902,
      "epoch": 1.087132140796307,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004893569864817057,
      "loss": 5.8322,
      "mean_token_accuracy": 0.18655450493097306,
      "num_tokens": 23880475.0,
      "step": 9420
    },
    {
      "entropy": 6.282496976852417,
      "epoch": 1.0877091748413157,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004893444418359579,
      "loss": 6.0479,
      "mean_token_accuracy": 0.17448573410511017,
      "num_tokens": 23893008.0,
      "step": 9425
    },
    {
      "entropy": 6.271885251998901,
      "epoch": 1.0882862088863243,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000489331889980841,
      "loss": 5.9571,
      "mean_token_accuracy": 0.1754467710852623,
      "num_tokens": 23906216.0,
      "step": 9430
    },
    {
      "entropy": 6.291494274139405,
      "epoch": 1.088863242931333,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004893193309167778,
      "loss": 5.9126,
      "mean_token_accuracy": 0.1802252322435379,
      "num_tokens": 23918978.0,
      "step": 9435
    },
    {
      "entropy": 6.277163743972778,
      "epoch": 1.0894402769763416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004893067646441902,
      "loss": 5.951,
      "mean_token_accuracy": 0.17773516476154327,
      "num_tokens": 23931536.0,
      "step": 9440
    },
    {
      "entropy": 6.255194568634034,
      "epoch": 1.0900173110213502,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000489294191163501,
      "loss": 5.977,
      "mean_token_accuracy": 0.17196260392665863,
      "num_tokens": 23945364.0,
      "step": 9445
    },
    {
      "entropy": 6.301913547515869,
      "epoch": 1.0905943450663589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004892816104751331,
      "loss": 5.9412,
      "mean_token_accuracy": 0.17413746565580368,
      "num_tokens": 23956654.0,
      "step": 9450
    },
    {
      "entropy": 6.3506751537323,
      "epoch": 1.0911713791113675,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892690225795096,
      "loss": 5.9964,
      "mean_token_accuracy": 0.17218484878540039,
      "num_tokens": 23970902.0,
      "step": 9455
    },
    {
      "entropy": 6.3299782276153564,
      "epoch": 1.0917484131563762,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004892564274770542,
      "loss": 6.0246,
      "mean_token_accuracy": 0.16886082738637925,
      "num_tokens": 23984312.0,
      "step": 9460
    },
    {
      "entropy": 6.31231279373169,
      "epoch": 1.0923254472013848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004892438251681901,
      "loss": 5.9304,
      "mean_token_accuracy": 0.17664276510477067,
      "num_tokens": 23996040.0,
      "step": 9465
    },
    {
      "entropy": 6.158179473876953,
      "epoch": 1.0929024812463934,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004892312156533413,
      "loss": 5.8788,
      "mean_token_accuracy": 0.18310317695140838,
      "num_tokens": 24008737.0,
      "step": 9470
    },
    {
      "entropy": 6.207678937911988,
      "epoch": 1.0934795152914023,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004892185989329321,
      "loss": 5.9407,
      "mean_token_accuracy": 0.18016728907823562,
      "num_tokens": 24021735.0,
      "step": 9475
    },
    {
      "entropy": 6.2856221199035645,
      "epoch": 1.094056549336411,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004892059750073868,
      "loss": 5.902,
      "mean_token_accuracy": 0.1802988663315773,
      "num_tokens": 24034391.0,
      "step": 9480
    },
    {
      "entropy": 6.242883634567261,
      "epoch": 1.0946335833814196,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004891933438771299,
      "loss": 5.9472,
      "mean_token_accuracy": 0.17923660278320314,
      "num_tokens": 24047331.0,
      "step": 9485
    },
    {
      "entropy": 6.31348352432251,
      "epoch": 1.0952106174264282,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004891807055425862,
      "loss": 5.9492,
      "mean_token_accuracy": 0.17951114773750304,
      "num_tokens": 24060111.0,
      "step": 9490
    },
    {
      "entropy": 6.214091396331787,
      "epoch": 1.0957876514714369,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004891680600041809,
      "loss": 5.9672,
      "mean_token_accuracy": 0.17410264760255814,
      "num_tokens": 24072275.0,
      "step": 9495
    },
    {
      "entropy": 6.277815198898315,
      "epoch": 1.0963646855164455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004891554072623392,
      "loss": 5.9395,
      "mean_token_accuracy": 0.17899050116539,
      "num_tokens": 24084647.0,
      "step": 9500
    },
    {
      "entropy": 6.235111665725708,
      "epoch": 1.0969417195614541,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004891427473174869,
      "loss": 5.8234,
      "mean_token_accuracy": 0.196658094227314,
      "num_tokens": 24097565.0,
      "step": 9505
    },
    {
      "entropy": 6.260201501846313,
      "epoch": 1.0975187536064628,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004891300801700496,
      "loss": 5.9365,
      "mean_token_accuracy": 0.1819838061928749,
      "num_tokens": 24111603.0,
      "step": 9510
    },
    {
      "entropy": 6.236711692810059,
      "epoch": 1.0980957876514714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004891174058204534,
      "loss": 5.8659,
      "mean_token_accuracy": 0.18519302904605867,
      "num_tokens": 24122763.0,
      "step": 9515
    },
    {
      "entropy": 6.232049989700317,
      "epoch": 1.09867282169648,
      "grad_norm": 0.875,
      "learning_rate": 0.0004891047242691246,
      "loss": 5.8994,
      "mean_token_accuracy": 0.18187192976474761,
      "num_tokens": 24134591.0,
      "step": 9520
    },
    {
      "entropy": 6.302640342712403,
      "epoch": 1.0992498557414887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004890920355164897,
      "loss": 6.005,
      "mean_token_accuracy": 0.17682669907808304,
      "num_tokens": 24147574.0,
      "step": 9525
    },
    {
      "entropy": 6.284323501586914,
      "epoch": 1.0998268897864973,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004890793395629756,
      "loss": 5.96,
      "mean_token_accuracy": 0.17549359649419785,
      "num_tokens": 24159760.0,
      "step": 9530
    },
    {
      "entropy": 6.217068767547607,
      "epoch": 1.100403923831506,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004890666364090093,
      "loss": 5.9141,
      "mean_token_accuracy": 0.17851177901029586,
      "num_tokens": 24173577.0,
      "step": 9535
    },
    {
      "entropy": 6.247491645812988,
      "epoch": 1.1009809578765146,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004890539260550181,
      "loss": 5.9312,
      "mean_token_accuracy": 0.18220448940992356,
      "num_tokens": 24185180.0,
      "step": 9540
    },
    {
      "entropy": 6.230571937561035,
      "epoch": 1.1015579919215233,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004890412085014292,
      "loss": 5.9122,
      "mean_token_accuracy": 0.17490753084421157,
      "num_tokens": 24198967.0,
      "step": 9545
    },
    {
      "entropy": 6.246636056900025,
      "epoch": 1.1021350259665321,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004890284837486706,
      "loss": 5.9863,
      "mean_token_accuracy": 0.17697999775409698,
      "num_tokens": 24210770.0,
      "step": 9550
    },
    {
      "entropy": 6.234810495376587,
      "epoch": 1.1027120600115408,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004890157517971704,
      "loss": 5.9202,
      "mean_token_accuracy": 0.1824613019824028,
      "num_tokens": 24222970.0,
      "step": 9555
    },
    {
      "entropy": 6.289969635009766,
      "epoch": 1.1032890940565494,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004890030126473566,
      "loss": 5.9877,
      "mean_token_accuracy": 0.17253732681274414,
      "num_tokens": 24235591.0,
      "step": 9560
    },
    {
      "entropy": 6.273422431945801,
      "epoch": 1.103866128101558,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004889902662996578,
      "loss": 5.9278,
      "mean_token_accuracy": 0.18337472528219223,
      "num_tokens": 24248844.0,
      "step": 9565
    },
    {
      "entropy": 6.245015287399292,
      "epoch": 1.1044431621465667,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004889775127545027,
      "loss": 5.9184,
      "mean_token_accuracy": 0.18381413966417312,
      "num_tokens": 24260267.0,
      "step": 9570
    },
    {
      "entropy": 6.299554252624512,
      "epoch": 1.1050201961915753,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004889647520123202,
      "loss": 6.0006,
      "mean_token_accuracy": 0.1752782255411148,
      "num_tokens": 24273019.0,
      "step": 9575
    },
    {
      "entropy": 6.218487787246704,
      "epoch": 1.105597230236584,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004889519840735396,
      "loss": 5.893,
      "mean_token_accuracy": 0.18125057965517044,
      "num_tokens": 24284805.0,
      "step": 9580
    },
    {
      "entropy": 6.21037278175354,
      "epoch": 1.1061742642815926,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004889392089385903,
      "loss": 5.973,
      "mean_token_accuracy": 0.16952159404754638,
      "num_tokens": 24297080.0,
      "step": 9585
    },
    {
      "entropy": 6.26860785484314,
      "epoch": 1.1067512983266012,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004889264266079019,
      "loss": 5.9377,
      "mean_token_accuracy": 0.17853870689868928,
      "num_tokens": 24308237.0,
      "step": 9590
    },
    {
      "entropy": 6.288543272018432,
      "epoch": 1.1073283323716099,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004889136370819045,
      "loss": 5.9408,
      "mean_token_accuracy": 0.17501837462186814,
      "num_tokens": 24322435.0,
      "step": 9595
    },
    {
      "entropy": 6.273493242263794,
      "epoch": 1.1079053664166185,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889008403610281,
      "loss": 5.968,
      "mean_token_accuracy": 0.17190479189157487,
      "num_tokens": 24335372.0,
      "step": 9600
    },
    {
      "entropy": 6.213484191894532,
      "epoch": 1.1084824004616272,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004888880364457033,
      "loss": 5.8487,
      "mean_token_accuracy": 0.1867966130375862,
      "num_tokens": 24347956.0,
      "step": 9605
    },
    {
      "entropy": 6.262439870834351,
      "epoch": 1.1090594345066358,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004888752253363604,
      "loss": 5.9227,
      "mean_token_accuracy": 0.17678878605365753,
      "num_tokens": 24360378.0,
      "step": 9610
    },
    {
      "entropy": 6.221158456802368,
      "epoch": 1.1096364685516447,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888624070334308,
      "loss": 5.9427,
      "mean_token_accuracy": 0.1770041763782501,
      "num_tokens": 24372492.0,
      "step": 9615
    },
    {
      "entropy": 6.3089710712432865,
      "epoch": 1.1102135025966533,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004888495815373452,
      "loss": 5.9736,
      "mean_token_accuracy": 0.17231011241674424,
      "num_tokens": 24384634.0,
      "step": 9620
    },
    {
      "entropy": 6.328020906448364,
      "epoch": 1.110790536641662,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004888367488485351,
      "loss": 5.9813,
      "mean_token_accuracy": 0.17650600969791413,
      "num_tokens": 24397117.0,
      "step": 9625
    },
    {
      "entropy": 6.252351140975952,
      "epoch": 1.1113675706866706,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004888239089674323,
      "loss": 5.937,
      "mean_token_accuracy": 0.1831529676914215,
      "num_tokens": 24409071.0,
      "step": 9630
    },
    {
      "entropy": 6.168025636672974,
      "epoch": 1.1119446047316792,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004888110618944686,
      "loss": 5.9083,
      "mean_token_accuracy": 0.17700492590665817,
      "num_tokens": 24422016.0,
      "step": 9635
    },
    {
      "entropy": 6.246804857254029,
      "epoch": 1.1125216387766879,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004887982076300759,
      "loss": 5.9738,
      "mean_token_accuracy": 0.17761478424072266,
      "num_tokens": 24433356.0,
      "step": 9640
    },
    {
      "entropy": 6.17910943031311,
      "epoch": 1.1130986728216965,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0004887853461746867,
      "loss": 5.8712,
      "mean_token_accuracy": 0.18242733478546141,
      "num_tokens": 24449027.0,
      "step": 9645
    },
    {
      "entropy": 6.256478500366211,
      "epoch": 1.1136757068667051,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004887724775287336,
      "loss": 5.9588,
      "mean_token_accuracy": 0.17390252351760865,
      "num_tokens": 24461592.0,
      "step": 9650
    },
    {
      "entropy": 6.232667350769043,
      "epoch": 1.1142527409117138,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004887596016926494,
      "loss": 5.9558,
      "mean_token_accuracy": 0.17697520256042482,
      "num_tokens": 24474003.0,
      "step": 9655
    },
    {
      "entropy": 6.27250862121582,
      "epoch": 1.1148297749567224,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004887467186668673,
      "loss": 5.8966,
      "mean_token_accuracy": 0.18020493239164354,
      "num_tokens": 24487096.0,
      "step": 9660
    },
    {
      "entropy": 6.188425970077515,
      "epoch": 1.115406809001731,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004887338284518204,
      "loss": 5.8248,
      "mean_token_accuracy": 0.19169029593467712,
      "num_tokens": 24499685.0,
      "step": 9665
    },
    {
      "entropy": 6.272452688217163,
      "epoch": 1.1159838430467397,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004887209310479423,
      "loss": 5.9744,
      "mean_token_accuracy": 0.17825918048620223,
      "num_tokens": 24512372.0,
      "step": 9670
    },
    {
      "entropy": 6.117705202102661,
      "epoch": 1.1165608770917483,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004887080264556668,
      "loss": 5.7236,
      "mean_token_accuracy": 0.1982050806283951,
      "num_tokens": 24524694.0,
      "step": 9675
    },
    {
      "entropy": 6.209850311279297,
      "epoch": 1.117137911136757,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004886951146754282,
      "loss": 5.8215,
      "mean_token_accuracy": 0.18689839094877242,
      "num_tokens": 24537779.0,
      "step": 9680
    },
    {
      "entropy": 6.225719881057739,
      "epoch": 1.1177149451817656,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004886821957076603,
      "loss": 5.8692,
      "mean_token_accuracy": 0.18366943895816804,
      "num_tokens": 24551012.0,
      "step": 9685
    },
    {
      "entropy": 6.254965305328369,
      "epoch": 1.1182919792267745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000488669269552798,
      "loss": 5.9029,
      "mean_token_accuracy": 0.17704096138477327,
      "num_tokens": 24562464.0,
      "step": 9690
    },
    {
      "entropy": 6.249507427215576,
      "epoch": 1.1188690132717831,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488656336211276,
      "loss": 5.8927,
      "mean_token_accuracy": 0.18272412568330765,
      "num_tokens": 24574223.0,
      "step": 9695
    },
    {
      "entropy": 6.260955953598023,
      "epoch": 1.1194460473167918,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004886433956835292,
      "loss": 5.9489,
      "mean_token_accuracy": 0.17904128283262252,
      "num_tokens": 24586464.0,
      "step": 9700
    },
    {
      "entropy": 6.195665121078491,
      "epoch": 1.1200230813618004,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004886304479699929,
      "loss": 5.9495,
      "mean_token_accuracy": 0.17086247503757476,
      "num_tokens": 24599071.0,
      "step": 9705
    },
    {
      "entropy": 6.196687364578247,
      "epoch": 1.120600115406809,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004886174930711027,
      "loss": 5.8241,
      "mean_token_accuracy": 0.19134058952331542,
      "num_tokens": 24611302.0,
      "step": 9710
    },
    {
      "entropy": 6.252506589889526,
      "epoch": 1.1211771494518177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004886045309872941,
      "loss": 5.994,
      "mean_token_accuracy": 0.17830884009599685,
      "num_tokens": 24624725.0,
      "step": 9715
    },
    {
      "entropy": 6.284151792526245,
      "epoch": 1.1217541834968263,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004885915617190034,
      "loss": 6.0126,
      "mean_token_accuracy": 0.1770956590771675,
      "num_tokens": 24638797.0,
      "step": 9720
    },
    {
      "entropy": 6.290525484085083,
      "epoch": 1.122331217541835,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004885785852666664,
      "loss": 5.9424,
      "mean_token_accuracy": 0.17862018346786498,
      "num_tokens": 24651462.0,
      "step": 9725
    },
    {
      "entropy": 6.1948224067687985,
      "epoch": 1.1229082515868436,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004885656016307199,
      "loss": 5.9289,
      "mean_token_accuracy": 0.1830155670642853,
      "num_tokens": 24664351.0,
      "step": 9730
    },
    {
      "entropy": 6.262338495254516,
      "epoch": 1.1234852856318522,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004885526108116004,
      "loss": 5.9382,
      "mean_token_accuracy": 0.17929500043392183,
      "num_tokens": 24677332.0,
      "step": 9735
    },
    {
      "entropy": 6.256038618087769,
      "epoch": 1.1240623196768609,
      "grad_norm": 0.796875,
      "learning_rate": 0.000488539612809745,
      "loss": 5.9463,
      "mean_token_accuracy": 0.1767493635416031,
      "num_tokens": 24690408.0,
      "step": 9740
    },
    {
      "entropy": 6.249878311157227,
      "epoch": 1.1246393537218695,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004885266076255907,
      "loss": 6.0597,
      "mean_token_accuracy": 0.17239488065242767,
      "num_tokens": 24703776.0,
      "step": 9745
    },
    {
      "entropy": 6.231577062606812,
      "epoch": 1.1252163877668782,
      "grad_norm": 1.53125,
      "learning_rate": 0.000488513595259575,
      "loss": 5.8191,
      "mean_token_accuracy": 0.18337346464395524,
      "num_tokens": 24716385.0,
      "step": 9750
    },
    {
      "entropy": 6.1917156219482425,
      "epoch": 1.125793421811887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004885005757121357,
      "loss": 5.8534,
      "mean_token_accuracy": 0.18059920519590378,
      "num_tokens": 24729055.0,
      "step": 9755
    },
    {
      "entropy": 6.252173852920532,
      "epoch": 1.1263704558568954,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004884875489837105,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1834695264697075,
      "num_tokens": 24741168.0,
      "step": 9760
    },
    {
      "entropy": 6.255601978302002,
      "epoch": 1.1269474899019043,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004884745150747378,
      "loss": 5.8826,
      "mean_token_accuracy": 0.18673125207424163,
      "num_tokens": 24753682.0,
      "step": 9765
    },
    {
      "entropy": 6.098710680007935,
      "epoch": 1.127524523946913,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004884614739856556,
      "loss": 5.8953,
      "mean_token_accuracy": 0.1864827513694763,
      "num_tokens": 24767528.0,
      "step": 9770
    },
    {
      "entropy": 6.274254846572876,
      "epoch": 1.1281015579919216,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004884484257169028,
      "loss": 5.9888,
      "mean_token_accuracy": 0.17384368181228638,
      "num_tokens": 24780989.0,
      "step": 9775
    },
    {
      "entropy": 6.248011779785156,
      "epoch": 1.1286785920369302,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004884353702689183,
      "loss": 5.9519,
      "mean_token_accuracy": 0.17618138194084168,
      "num_tokens": 24792595.0,
      "step": 9780
    },
    {
      "entropy": 6.205896472930908,
      "epoch": 1.1292556260819389,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004884223076421411,
      "loss": 5.892,
      "mean_token_accuracy": 0.17567367106676102,
      "num_tokens": 24805665.0,
      "step": 9785
    },
    {
      "entropy": 6.25036506652832,
      "epoch": 1.1298326601269475,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004884092378370106,
      "loss": 5.8332,
      "mean_token_accuracy": 0.1854578211903572,
      "num_tokens": 24818276.0,
      "step": 9790
    },
    {
      "entropy": 6.215784978866577,
      "epoch": 1.1304096941719561,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004883961608539664,
      "loss": 5.9196,
      "mean_token_accuracy": 0.17698087692260742,
      "num_tokens": 24831828.0,
      "step": 9795
    },
    {
      "entropy": 6.2280348777771,
      "epoch": 1.1309867282169648,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004883830766934484,
      "loss": 5.8719,
      "mean_token_accuracy": 0.18407948464155197,
      "num_tokens": 24844597.0,
      "step": 9800
    },
    {
      "entropy": 6.228781223297119,
      "epoch": 1.1315637622619734,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004883699853558965,
      "loss": 5.8851,
      "mean_token_accuracy": 0.18682559728622436,
      "num_tokens": 24856136.0,
      "step": 9805
    },
    {
      "entropy": 6.309537792205811,
      "epoch": 1.132140796306982,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004883568868417511,
      "loss": 6.0515,
      "mean_token_accuracy": 0.168924218416214,
      "num_tokens": 24870888.0,
      "step": 9810
    },
    {
      "entropy": 6.268776369094849,
      "epoch": 1.1327178303519907,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004883437811514528,
      "loss": 5.9651,
      "mean_token_accuracy": 0.17877494543790817,
      "num_tokens": 24884239.0,
      "step": 9815
    },
    {
      "entropy": 6.246257781982422,
      "epoch": 1.1332948643969993,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004883306682854424,
      "loss": 5.8892,
      "mean_token_accuracy": 0.1902910977602005,
      "num_tokens": 24896370.0,
      "step": 9820
    },
    {
      "entropy": 6.221540832519532,
      "epoch": 1.133871898442008,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004883175482441611,
      "loss": 5.9755,
      "mean_token_accuracy": 0.1689189612865448,
      "num_tokens": 24908868.0,
      "step": 9825
    },
    {
      "entropy": 6.2697690486907955,
      "epoch": 1.1344489324870168,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004883044210280499,
      "loss": 6.0028,
      "mean_token_accuracy": 0.1793764740228653,
      "num_tokens": 24920630.0,
      "step": 9830
    },
    {
      "entropy": 6.169307231903076,
      "epoch": 1.1350259665320255,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004882912866375505,
      "loss": 5.8111,
      "mean_token_accuracy": 0.18559192270040512,
      "num_tokens": 24932787.0,
      "step": 9835
    },
    {
      "entropy": 6.248462390899658,
      "epoch": 1.1356030005770341,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048827814507310455,
      "loss": 5.872,
      "mean_token_accuracy": 0.18284289091825484,
      "num_tokens": 24944436.0,
      "step": 9840
    },
    {
      "entropy": 6.273605632781982,
      "epoch": 1.1361800346220428,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048826499633515416,
      "loss": 5.8866,
      "mean_token_accuracy": 0.18615373224020004,
      "num_tokens": 24956944.0,
      "step": 9845
    },
    {
      "entropy": 6.292548513412475,
      "epoch": 1.1367570686670514,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048825184042414156,
      "loss": 5.9501,
      "mean_token_accuracy": 0.18019989877939224,
      "num_tokens": 24971540.0,
      "step": 9850
    },
    {
      "entropy": 6.25846586227417,
      "epoch": 1.13733410271206,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004882386773405092,
      "loss": 6.0058,
      "mean_token_accuracy": 0.17531268745660783,
      "num_tokens": 24984056.0,
      "step": 9855
    },
    {
      "entropy": 6.257551765441894,
      "epoch": 1.1379111367570687,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004882255070846999,
      "loss": 5.9961,
      "mean_token_accuracy": 0.1747207224369049,
      "num_tokens": 24995625.0,
      "step": 9860
    },
    {
      "entropy": 6.286048555374146,
      "epoch": 1.1384881708020773,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048821232965715663,
      "loss": 5.8659,
      "mean_token_accuracy": 0.1780352622270584,
      "num_tokens": 25007336.0,
      "step": 9865
    },
    {
      "entropy": 6.233601236343384,
      "epoch": 1.139065204847086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004881991450583225,
      "loss": 5.9502,
      "mean_token_accuracy": 0.17873996645212173,
      "num_tokens": 25019446.0,
      "step": 9870
    },
    {
      "entropy": 6.261944389343261,
      "epoch": 1.1396422388920946,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000488185953288641,
      "loss": 5.8917,
      "mean_token_accuracy": 0.17839810252189636,
      "num_tokens": 25031518.0,
      "step": 9875
    },
    {
      "entropy": 6.189383268356323,
      "epoch": 1.1402192729371032,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004881727543485559,
      "loss": 5.8397,
      "mean_token_accuracy": 0.1882979691028595,
      "num_tokens": 25043365.0,
      "step": 9880
    },
    {
      "entropy": 6.242651414871216,
      "epoch": 1.1407963069821119,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048815954823851107,
      "loss": 5.9929,
      "mean_token_accuracy": 0.17533566653728486,
      "num_tokens": 25055748.0,
      "step": 9885
    },
    {
      "entropy": 6.257275199890136,
      "epoch": 1.1413733410271205,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048814633495895067,
      "loss": 5.8662,
      "mean_token_accuracy": 0.1791042134165764,
      "num_tokens": 25067904.0,
      "step": 9890
    },
    {
      "entropy": 6.218337726593018,
      "epoch": 1.1419503750721294,
      "grad_norm": 0.875,
      "learning_rate": 0.0004881331145103192,
      "loss": 5.9254,
      "mean_token_accuracy": 0.18318034708499908,
      "num_tokens": 25081584.0,
      "step": 9895
    },
    {
      "entropy": 6.2279399871826175,
      "epoch": 1.1425274091171378,
      "grad_norm": 0.875,
      "learning_rate": 0.0004881198868930614,
      "loss": 5.9486,
      "mean_token_accuracy": 0.1760874032974243,
      "num_tokens": 25094555.0,
      "step": 9900
    },
    {
      "entropy": 6.301339721679687,
      "epoch": 1.1431044431621467,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048810665210762195,
      "loss": 5.9941,
      "mean_token_accuracy": 0.17765888273715974,
      "num_tokens": 25108405.0,
      "step": 9905
    },
    {
      "entropy": 6.254843235015869,
      "epoch": 1.1436814772071553,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048809341015444615,
      "loss": 5.965,
      "mean_token_accuracy": 0.17438613772392272,
      "num_tokens": 25121404.0,
      "step": 9910
    },
    {
      "entropy": 6.247920989990234,
      "epoch": 1.144258511252164,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048808016103397934,
      "loss": 5.898,
      "mean_token_accuracy": 0.1843980923295021,
      "num_tokens": 25134773.0,
      "step": 9915
    },
    {
      "entropy": 6.215767669677734,
      "epoch": 1.1448355452971726,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004880669047466671,
      "loss": 5.8805,
      "mean_token_accuracy": 0.18316199630498886,
      "num_tokens": 25147873.0,
      "step": 9920
    },
    {
      "entropy": 6.170156478881836,
      "epoch": 1.1454125793421812,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048805364129295554,
      "loss": 5.8399,
      "mean_token_accuracy": 0.1915966898202896,
      "num_tokens": 25162081.0,
      "step": 9925
    },
    {
      "entropy": 6.311111640930176,
      "epoch": 1.1459896133871899,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048804037067329037,
      "loss": 6.0068,
      "mean_token_accuracy": 0.1734829768538475,
      "num_tokens": 25174203.0,
      "step": 9930
    },
    {
      "entropy": 6.244969701766967,
      "epoch": 1.1465666474321985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004880270928881183,
      "loss": 6.0155,
      "mean_token_accuracy": 0.17717694640159606,
      "num_tokens": 25187238.0,
      "step": 9935
    },
    {
      "entropy": 6.257837438583374,
      "epoch": 1.1471436814772071,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004880138079378857,
      "loss": 5.8506,
      "mean_token_accuracy": 0.18542557209730148,
      "num_tokens": 25199456.0,
      "step": 9940
    },
    {
      "entropy": 6.240293550491333,
      "epoch": 1.1477207155222158,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004880005158230395,
      "loss": 5.9263,
      "mean_token_accuracy": 0.17797180414199829,
      "num_tokens": 25215163.0,
      "step": 9945
    },
    {
      "entropy": 6.21778826713562,
      "epoch": 1.1482977495672244,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004879872165440268,
      "loss": 5.9156,
      "mean_token_accuracy": 0.18256105482578278,
      "num_tokens": 25226946.0,
      "step": 9950
    },
    {
      "entropy": 6.188058137893677,
      "epoch": 1.148874783612233,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004879739101012948,
      "loss": 5.8295,
      "mean_token_accuracy": 0.18217216730117797,
      "num_tokens": 25238691.0,
      "step": 9955
    },
    {
      "entropy": 6.194533586502075,
      "epoch": 1.1494518176572417,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004879605964952911,
      "loss": 5.8397,
      "mean_token_accuracy": 0.1822061464190483,
      "num_tokens": 25251513.0,
      "step": 9960
    },
    {
      "entropy": 6.212546968460083,
      "epoch": 1.1500288517022503,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048794727572646366,
      "loss": 5.7971,
      "mean_token_accuracy": 0.18410493582487106,
      "num_tokens": 25263513.0,
      "step": 9965
    },
    {
      "entropy": 6.172162199020386,
      "epoch": 1.1506058857472592,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004879339477952603,
      "loss": 5.929,
      "mean_token_accuracy": 0.1787589520215988,
      "num_tokens": 25276989.0,
      "step": 9970
    },
    {
      "entropy": 6.263672733306885,
      "epoch": 1.1511829197922678,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048792061270212935,
      "loss": 5.9198,
      "mean_token_accuracy": 0.1863589271903038,
      "num_tokens": 25289474.0,
      "step": 9975
    },
    {
      "entropy": 6.266945791244507,
      "epoch": 1.1517599538372765,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004879072704475193,
      "loss": 5.8971,
      "mean_token_accuracy": 0.18463608622550964,
      "num_tokens": 25302102.0,
      "step": 9980
    },
    {
      "entropy": 6.179267358779907,
      "epoch": 1.1523369878822851,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048789392103187906,
      "loss": 5.8966,
      "mean_token_accuracy": 0.1830424666404724,
      "num_tokens": 25315645.0,
      "step": 9985
    },
    {
      "entropy": 6.3015937328338625,
      "epoch": 1.1529140219272938,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004878805644556575,
      "loss": 5.9686,
      "mean_token_accuracy": 0.17896064668893813,
      "num_tokens": 25328272.0,
      "step": 9990
    },
    {
      "entropy": 6.221786451339722,
      "epoch": 1.1534910559723024,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000487867200719304,
      "loss": 5.9012,
      "mean_token_accuracy": 0.1820850118994713,
      "num_tokens": 25340449.0,
      "step": 9995
    },
    {
      "entropy": 6.201902627944946,
      "epoch": 1.154068090017311,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004878538298232678,
      "loss": 5.8348,
      "mean_token_accuracy": 0.1860969915986061,
      "num_tokens": 25353232.0,
      "step": 10000
    },
    {
      "entropy": 6.22259202003479,
      "epoch": 1.1546451240623197,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004878404517679988,
      "loss": 5.9191,
      "mean_token_accuracy": 0.18249674141407013,
      "num_tokens": 25364575.0,
      "step": 10005
    },
    {
      "entropy": 6.1438665866851805,
      "epoch": 1.1552221581073283,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048782706655394695,
      "loss": 5.8503,
      "mean_token_accuracy": 0.1820308521389961,
      "num_tokens": 25377992.0,
      "step": 10010
    },
    {
      "entropy": 6.300264453887939,
      "epoch": 1.155799192152337,
      "grad_norm": 0.875,
      "learning_rate": 0.0004878136741815624,
      "loss": 5.9753,
      "mean_token_accuracy": 0.1797061175107956,
      "num_tokens": 25391168.0,
      "step": 10015
    },
    {
      "entropy": 6.178576946258545,
      "epoch": 1.1563762261973456,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004878002746512956,
      "loss": 5.84,
      "mean_token_accuracy": 0.18267696648836135,
      "num_tokens": 25405211.0,
      "step": 10020
    },
    {
      "entropy": 6.244351053237915,
      "epoch": 1.1569532602423542,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004877868679635974,
      "loss": 5.9167,
      "mean_token_accuracy": 0.1836722820997238,
      "num_tokens": 25416902.0,
      "step": 10025
    },
    {
      "entropy": 6.295699548721314,
      "epoch": 1.1575302942873629,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004877734541189185,
      "loss": 5.9542,
      "mean_token_accuracy": 0.17551818937063218,
      "num_tokens": 25429071.0,
      "step": 10030
    },
    {
      "entropy": 6.16551456451416,
      "epoch": 1.1581073283323717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048776003311771013,
      "loss": 5.8618,
      "mean_token_accuracy": 0.18317142874002457,
      "num_tokens": 25441153.0,
      "step": 10035
    },
    {
      "entropy": 6.219573926925659,
      "epoch": 1.1586843623773802,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004877466049604238,
      "loss": 5.8538,
      "mean_token_accuracy": 0.1850452020764351,
      "num_tokens": 25454302.0,
      "step": 10040
    },
    {
      "entropy": 6.284542036056519,
      "epoch": 1.159261396422389,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048773316964751106,
      "loss": 5.9278,
      "mean_token_accuracy": 0.17834423929452897,
      "num_tokens": 25466507.0,
      "step": 10045
    },
    {
      "entropy": 6.229846858978272,
      "epoch": 1.1598384304673977,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004877197271794239,
      "loss": 5.9277,
      "mean_token_accuracy": 0.17894631624221802,
      "num_tokens": 25479952.0,
      "step": 10050
    },
    {
      "entropy": 6.265223646163941,
      "epoch": 1.1604154645124063,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004877062775566142,
      "loss": 5.9963,
      "mean_token_accuracy": 0.17179697901010513,
      "num_tokens": 25491952.0,
      "step": 10055
    },
    {
      "entropy": 6.259234809875489,
      "epoch": 1.160992498557415,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048769282077953464,
      "loss": 5.8731,
      "mean_token_accuracy": 0.18257008939981462,
      "num_tokens": 25505053.0,
      "step": 10060
    },
    {
      "entropy": 6.273874807357788,
      "epoch": 1.1615695326024236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048767935684863775,
      "loss": 5.9783,
      "mean_token_accuracy": 0.17779058516025542,
      "num_tokens": 25517367.0,
      "step": 10065
    },
    {
      "entropy": 6.269525957107544,
      "epoch": 1.1621465666474322,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004876658857643762,
      "loss": 5.8988,
      "mean_token_accuracy": 0.17854540348052977,
      "num_tokens": 25528697.0,
      "step": 10070
    },
    {
      "entropy": 6.264088726043701,
      "epoch": 1.1627236006924409,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004876524075272034,
      "loss": 5.9299,
      "mean_token_accuracy": 0.17620706856250762,
      "num_tokens": 25541109.0,
      "step": 10075
    },
    {
      "entropy": 6.157661485671997,
      "epoch": 1.1633006347374495,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048763892213757234,
      "loss": 5.8426,
      "mean_token_accuracy": 0.18647894114255906,
      "num_tokens": 25553870.0,
      "step": 10080
    },
    {
      "entropy": 6.238891172409057,
      "epoch": 1.1638776687824581,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048762542959593683,
      "loss": 5.9602,
      "mean_token_accuracy": 0.17296314835548401,
      "num_tokens": 25566048.0,
      "step": 10085
    },
    {
      "entropy": 6.17363395690918,
      "epoch": 1.1644547028274668,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004876119299027506,
      "loss": 5.8466,
      "mean_token_accuracy": 0.18547327518463136,
      "num_tokens": 25578282.0,
      "step": 10090
    },
    {
      "entropy": 6.176361227035523,
      "epoch": 1.1650317368724754,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048759842305846766,
      "loss": 5.8192,
      "mean_token_accuracy": 0.18447879701852798,
      "num_tokens": 25590359.0,
      "step": 10095
    },
    {
      "entropy": 6.318369626998901,
      "epoch": 1.165608770917484,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004875849090635425,
      "loss": 5.9689,
      "mean_token_accuracy": 0.17571574300527573,
      "num_tokens": 25603256.0,
      "step": 10100
    },
    {
      "entropy": 6.2096953868865965,
      "epoch": 1.1661858049624927,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048757138791842943,
      "loss": 5.9682,
      "mean_token_accuracy": 0.17687658369541168,
      "num_tokens": 25616844.0,
      "step": 10105
    },
    {
      "entropy": 6.213687705993652,
      "epoch": 1.1667628390075016,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004875578596235832,
      "loss": 5.831,
      "mean_token_accuracy": 0.19320357739925384,
      "num_tokens": 25630490.0,
      "step": 10110
    },
    {
      "entropy": 6.261492919921875,
      "epoch": 1.1673398730525102,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004875443241794591,
      "loss": 5.9008,
      "mean_token_accuracy": 0.17846622616052626,
      "num_tokens": 25645082.0,
      "step": 10115
    },
    {
      "entropy": 6.206713581085205,
      "epoch": 1.1679169070975188,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048753078158651227,
      "loss": 5.9613,
      "mean_token_accuracy": 0.17909290343523027,
      "num_tokens": 25658108.0,
      "step": 10120
    },
    {
      "entropy": 6.1895058155059814,
      "epoch": 1.1684939411425275,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487517231845198,
      "loss": 5.9195,
      "mean_token_accuracy": 0.1787977248430252,
      "num_tokens": 25669883.0,
      "step": 10125
    },
    {
      "entropy": 6.254233312606812,
      "epoch": 1.169070975187536,
      "grad_norm": 0.875,
      "learning_rate": 0.0004875036749559724,
      "loss": 5.9368,
      "mean_token_accuracy": 0.1762844055891037,
      "num_tokens": 25683393.0,
      "step": 10130
    },
    {
      "entropy": 6.29863018989563,
      "epoch": 1.1696480092325447,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048749011091929115,
      "loss": 5.9354,
      "mean_token_accuracy": 0.17928873747587204,
      "num_tokens": 25696564.0,
      "step": 10135
    },
    {
      "entropy": 6.187666034698486,
      "epoch": 1.1702250432775534,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004874765397356105,
      "loss": 5.8535,
      "mean_token_accuracy": 0.18042410612106324,
      "num_tokens": 25709350.0,
      "step": 10140
    },
    {
      "entropy": 6.25699462890625,
      "epoch": 1.170802077322562,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004874629614053871,
      "loss": 5.9027,
      "mean_token_accuracy": 0.1798825353384018,
      "num_tokens": 25721559.0,
      "step": 10145
    },
    {
      "entropy": 6.289753055572509,
      "epoch": 1.1713791113675707,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004874493759290775,
      "loss": 6.0286,
      "mean_token_accuracy": 0.17144690901041032,
      "num_tokens": 25734185.0,
      "step": 10150
    },
    {
      "entropy": 6.258307552337646,
      "epoch": 1.1719561454125793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048743578330713854,
      "loss": 5.8719,
      "mean_token_accuracy": 0.18238241821527482,
      "num_tokens": 25746456.0,
      "step": 10155
    },
    {
      "entropy": 6.175575113296508,
      "epoch": 1.172533179457588,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004874221835400277,
      "loss": 5.9022,
      "mean_token_accuracy": 0.1781415045261383,
      "num_tokens": 25759893.0,
      "step": 10160
    },
    {
      "entropy": 6.2830709457397464,
      "epoch": 1.1731102135025966,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004874085766282022,
      "loss": 6.0066,
      "mean_token_accuracy": 0.17472083121538162,
      "num_tokens": 25772574.0,
      "step": 10165
    },
    {
      "entropy": 6.232170677185058,
      "epoch": 1.1736872475476052,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048739496257211966,
      "loss": 5.8917,
      "mean_token_accuracy": 0.1863439902663231,
      "num_tokens": 25785748.0,
      "step": 10170
    },
    {
      "entropy": 6.2371602058410645,
      "epoch": 1.1742642815926139,
      "grad_norm": 0.875,
      "learning_rate": 0.00048738134137223815,
      "loss": 5.8942,
      "mean_token_accuracy": 0.18070970028638839,
      "num_tokens": 25799073.0,
      "step": 10175
    },
    {
      "entropy": 6.274782943725586,
      "epoch": 1.1748413156376225,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048736771302901566,
      "loss": 5.8917,
      "mean_token_accuracy": 0.17771412879228593,
      "num_tokens": 25810896.0,
      "step": 10180
    },
    {
      "entropy": 6.207719469070435,
      "epoch": 1.1754183496826314,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048735407754291063,
      "loss": 5.8831,
      "mean_token_accuracy": 0.18037483990192413,
      "num_tokens": 25824728.0,
      "step": 10185
    },
    {
      "entropy": 6.2243977069854735,
      "epoch": 1.17599538372764,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048734043491438175,
      "loss": 5.9532,
      "mean_token_accuracy": 0.17193512320518495,
      "num_tokens": 25837569.0,
      "step": 10190
    },
    {
      "entropy": 6.313347673416137,
      "epoch": 1.1765724177726486,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048732678514388773,
      "loss": 5.9964,
      "mean_token_accuracy": 0.17422874867916108,
      "num_tokens": 25850232.0,
      "step": 10195
    },
    {
      "entropy": 6.239552402496338,
      "epoch": 1.1771494518176573,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004873131282318878,
      "loss": 5.9453,
      "mean_token_accuracy": 0.18028897494077684,
      "num_tokens": 25862038.0,
      "step": 10200
    },
    {
      "entropy": 6.289503335952759,
      "epoch": 1.177726485862666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048729946417884126,
      "loss": 6.0211,
      "mean_token_accuracy": 0.17092031836509705,
      "num_tokens": 25874337.0,
      "step": 10205
    },
    {
      "entropy": 6.24993200302124,
      "epoch": 1.1783035199076746,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004872857929852076,
      "loss": 5.9286,
      "mean_token_accuracy": 0.18675503879785538,
      "num_tokens": 25888175.0,
      "step": 10210
    },
    {
      "entropy": 6.2598981857299805,
      "epoch": 1.1788805539526832,
      "grad_norm": 0.875,
      "learning_rate": 0.0004872721146514469,
      "loss": 5.8586,
      "mean_token_accuracy": 0.1848505347967148,
      "num_tokens": 25900747.0,
      "step": 10215
    },
    {
      "entropy": 6.236641550064087,
      "epoch": 1.1794575879976918,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000487258429178019,
      "loss": 5.9043,
      "mean_token_accuracy": 0.17885935604572295,
      "num_tokens": 25913010.0,
      "step": 10220
    },
    {
      "entropy": 6.239579439163208,
      "epoch": 1.1800346220427005,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048724473656538427,
      "loss": 5.8398,
      "mean_token_accuracy": 0.18039740473031998,
      "num_tokens": 25925757.0,
      "step": 10225
    },
    {
      "entropy": 6.242413282394409,
      "epoch": 1.1806116560877091,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004872310368140032,
      "loss": 5.9392,
      "mean_token_accuracy": 0.17444781512022017,
      "num_tokens": 25939574.0,
      "step": 10230
    },
    {
      "entropy": 6.235931491851806,
      "epoch": 1.1811886901327178,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004872173299243368,
      "loss": 5.9627,
      "mean_token_accuracy": 0.17699281573295594,
      "num_tokens": 25951142.0,
      "step": 10235
    },
    {
      "entropy": 6.228603029251099,
      "epoch": 1.1817657241777264,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048720361589684575,
      "loss": 5.9367,
      "mean_token_accuracy": 0.1782004415988922,
      "num_tokens": 25962575.0,
      "step": 10240
    },
    {
      "entropy": 6.250329160690308,
      "epoch": 1.182342758222735,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048718989473199147,
      "loss": 5.9423,
      "mean_token_accuracy": 0.18002069294452666,
      "num_tokens": 25977012.0,
      "step": 10245
    },
    {
      "entropy": 6.258625316619873,
      "epoch": 1.182919792267744,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004871761664302356,
      "loss": 5.959,
      "mean_token_accuracy": 0.1816967323422432,
      "num_tokens": 25991520.0,
      "step": 10250
    },
    {
      "entropy": 6.218161344528198,
      "epoch": 1.1834968263127525,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004871624309920397,
      "loss": 5.8231,
      "mean_token_accuracy": 0.18657898157835007,
      "num_tokens": 26004644.0,
      "step": 10255
    },
    {
      "entropy": 6.177314424514771,
      "epoch": 1.1840738603577612,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048714868841786586,
      "loss": 5.8614,
      "mean_token_accuracy": 0.1862117037177086,
      "num_tokens": 26016452.0,
      "step": 10260
    },
    {
      "entropy": 6.24371371269226,
      "epoch": 1.1846508944027698,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048713493870817626,
      "loss": 5.9082,
      "mean_token_accuracy": 0.18058924227952958,
      "num_tokens": 26028675.0,
      "step": 10265
    },
    {
      "entropy": 6.267215347290039,
      "epoch": 1.1852279284477785,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048712118186343336,
      "loss": 5.9649,
      "mean_token_accuracy": 0.1771223098039627,
      "num_tokens": 26041561.0,
      "step": 10270
    },
    {
      "entropy": 6.134266424179077,
      "epoch": 1.185804962492787,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048710741788409985,
      "loss": 5.7823,
      "mean_token_accuracy": 0.18506606221199035,
      "num_tokens": 26052808.0,
      "step": 10275
    },
    {
      "entropy": 6.268643236160278,
      "epoch": 1.1863819965377957,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004870936467706387,
      "loss": 5.9196,
      "mean_token_accuracy": 0.18116012513637542,
      "num_tokens": 26065624.0,
      "step": 10280
    },
    {
      "entropy": 6.233309602737426,
      "epoch": 1.1869590305828044,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004870798685235131,
      "loss": 5.8521,
      "mean_token_accuracy": 0.18350233584642411,
      "num_tokens": 26078075.0,
      "step": 10285
    },
    {
      "entropy": 6.247022247314453,
      "epoch": 1.187536064627813,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048706608314318654,
      "loss": 5.9407,
      "mean_token_accuracy": 0.18099519610404968,
      "num_tokens": 26090454.0,
      "step": 10290
    },
    {
      "entropy": 6.179354619979859,
      "epoch": 1.1881130986728217,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004870522906301225,
      "loss": 5.8473,
      "mean_token_accuracy": 0.1863680064678192,
      "num_tokens": 26103516.0,
      "step": 10295
    },
    {
      "entropy": 6.260325813293457,
      "epoch": 1.1886901327178303,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000487038490984785,
      "loss": 5.9394,
      "mean_token_accuracy": 0.17333969175815583,
      "num_tokens": 26116420.0,
      "step": 10300
    },
    {
      "entropy": 6.236327075958252,
      "epoch": 1.189267166762839,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048702468420763826,
      "loss": 5.9244,
      "mean_token_accuracy": 0.17825981080532075,
      "num_tokens": 26129685.0,
      "step": 10305
    },
    {
      "entropy": 6.279147291183472,
      "epoch": 1.1898442008078476,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048701087029914657,
      "loss": 5.9212,
      "mean_token_accuracy": 0.17717962116003036,
      "num_tokens": 26142909.0,
      "step": 10310
    },
    {
      "entropy": 6.2377691745758055,
      "epoch": 1.1904212348528562,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004869970492597745,
      "loss": 5.9214,
      "mean_token_accuracy": 0.18025095611810685,
      "num_tokens": 26155229.0,
      "step": 10315
    },
    {
      "entropy": 6.253656911849975,
      "epoch": 1.1909982688978649,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004869832210899871,
      "loss": 5.9412,
      "mean_token_accuracy": 0.17832035273313523,
      "num_tokens": 26168309.0,
      "step": 10320
    },
    {
      "entropy": 6.262561273574829,
      "epoch": 1.1915753029428737,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048696938579024927,
      "loss": 5.9077,
      "mean_token_accuracy": 0.18755118995904924,
      "num_tokens": 26180189.0,
      "step": 10325
    },
    {
      "entropy": 6.160376644134521,
      "epoch": 1.1921523369878824,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048695554336102644,
      "loss": 5.8656,
      "mean_token_accuracy": 0.1821357175707817,
      "num_tokens": 26193125.0,
      "step": 10330
    },
    {
      "entropy": 6.262964963912964,
      "epoch": 1.192729371032891,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048694169380278417,
      "loss": 5.9345,
      "mean_token_accuracy": 0.17666238993406297,
      "num_tokens": 26206179.0,
      "step": 10335
    },
    {
      "entropy": 6.300597906112671,
      "epoch": 1.1933064050778996,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004869278371159884,
      "loss": 5.8622,
      "mean_token_accuracy": 0.17887118458747864,
      "num_tokens": 26217887.0,
      "step": 10340
    },
    {
      "entropy": 6.253102731704712,
      "epoch": 1.1938834391229083,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048691397330110503,
      "loss": 5.8874,
      "mean_token_accuracy": 0.18294475078582764,
      "num_tokens": 26231280.0,
      "step": 10345
    },
    {
      "entropy": 6.139592933654785,
      "epoch": 1.194460473167917,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004869001023586005,
      "loss": 5.7454,
      "mean_token_accuracy": 0.192977836728096,
      "num_tokens": 26245565.0,
      "step": 10350
    },
    {
      "entropy": 6.191368961334229,
      "epoch": 1.1950375072129256,
      "grad_norm": 0.875,
      "learning_rate": 0.0004868862242889413,
      "loss": 5.9006,
      "mean_token_accuracy": 0.177755643427372,
      "num_tokens": 26257313.0,
      "step": 10355
    },
    {
      "entropy": 6.185658502578735,
      "epoch": 1.1956145412579342,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000486872339092594,
      "loss": 5.8749,
      "mean_token_accuracy": 0.18288901597261428,
      "num_tokens": 26270566.0,
      "step": 10360
    },
    {
      "entropy": 6.258494663238525,
      "epoch": 1.1961915753029428,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000486858446770026,
      "loss": 5.9906,
      "mean_token_accuracy": 0.17170356810092927,
      "num_tokens": 26283490.0,
      "step": 10365
    },
    {
      "entropy": 6.300401592254639,
      "epoch": 1.1967686093479515,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004868445473217043,
      "loss": 5.916,
      "mean_token_accuracy": 0.17813777476549147,
      "num_tokens": 26297702.0,
      "step": 10370
    },
    {
      "entropy": 6.175187826156616,
      "epoch": 1.1973456433929601,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004868306407480965,
      "loss": 5.8565,
      "mean_token_accuracy": 0.1803478941321373,
      "num_tokens": 26309543.0,
      "step": 10375
    },
    {
      "entropy": 6.257505321502686,
      "epoch": 1.1979226774379688,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048681672704967036,
      "loss": 6.0132,
      "mean_token_accuracy": 0.17134004235267639,
      "num_tokens": 26322259.0,
      "step": 10380
    },
    {
      "entropy": 6.235605478286743,
      "epoch": 1.1984997114829774,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004868028062268938,
      "loss": 5.8864,
      "mean_token_accuracy": 0.17874849140644072,
      "num_tokens": 26334985.0,
      "step": 10385
    },
    {
      "entropy": 6.136746025085449,
      "epoch": 1.1990767455279863,
      "grad_norm": 0.875,
      "learning_rate": 0.00048678887828023504,
      "loss": 5.7137,
      "mean_token_accuracy": 0.18983878195285797,
      "num_tokens": 26348441.0,
      "step": 10390
    },
    {
      "entropy": 6.245867490768433,
      "epoch": 1.1996537795729947,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004867749432101626,
      "loss": 5.9285,
      "mean_token_accuracy": 0.17594893425703048,
      "num_tokens": 26360848.0,
      "step": 10395
    },
    {
      "entropy": 6.157463836669922,
      "epoch": 1.2002308136180035,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004867610010171451,
      "loss": 5.8415,
      "mean_token_accuracy": 0.19420559704303741,
      "num_tokens": 26373704.0,
      "step": 10400
    },
    {
      "entropy": 6.178312158584594,
      "epoch": 1.2008078476630122,
      "grad_norm": 0.875,
      "learning_rate": 0.00048674705170165147,
      "loss": 5.8732,
      "mean_token_accuracy": 0.18894819617271424,
      "num_tokens": 26387035.0,
      "step": 10405
    },
    {
      "entropy": 6.201609086990357,
      "epoch": 1.2013848817080208,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000486733095264151,
      "loss": 5.8124,
      "mean_token_accuracy": 0.18969660848379136,
      "num_tokens": 26399064.0,
      "step": 10410
    },
    {
      "entropy": 6.232175254821778,
      "epoch": 1.2019619157530295,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048671913170511306,
      "loss": 5.9225,
      "mean_token_accuracy": 0.18029454350471497,
      "num_tokens": 26411930.0,
      "step": 10415
    },
    {
      "entropy": 6.165471363067627,
      "epoch": 1.202538949798038,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004867051610250073,
      "loss": 5.8603,
      "mean_token_accuracy": 0.17913966923952102,
      "num_tokens": 26425855.0,
      "step": 10420
    },
    {
      "entropy": 6.286369514465332,
      "epoch": 1.2031159838430467,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004866911832243035,
      "loss": 5.9369,
      "mean_token_accuracy": 0.17732799798250198,
      "num_tokens": 26437649.0,
      "step": 10425
    },
    {
      "entropy": 6.187501287460327,
      "epoch": 1.2036930178880554,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048667719830347203,
      "loss": 5.8614,
      "mean_token_accuracy": 0.17876151204109192,
      "num_tokens": 26449503.0,
      "step": 10430
    },
    {
      "entropy": 6.213302421569824,
      "epoch": 1.204270051933064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048666320626298307,
      "loss": 5.8923,
      "mean_token_accuracy": 0.1856320843100548,
      "num_tokens": 26462199.0,
      "step": 10435
    },
    {
      "entropy": 6.270549392700195,
      "epoch": 1.2048470859780727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048664920710330735,
      "loss": 5.9156,
      "mean_token_accuracy": 0.18263922184705733,
      "num_tokens": 26475295.0,
      "step": 10440
    },
    {
      "entropy": 6.21199107170105,
      "epoch": 1.2054241200230813,
      "grad_norm": 0.875,
      "learning_rate": 0.00048663520082491564,
      "loss": 5.917,
      "mean_token_accuracy": 0.1792993128299713,
      "num_tokens": 26488846.0,
      "step": 10445
    },
    {
      "entropy": 6.256782817840576,
      "epoch": 1.20600115406809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004866211874282791,
      "loss": 5.918,
      "mean_token_accuracy": 0.18471352458000184,
      "num_tokens": 26501645.0,
      "step": 10450
    },
    {
      "entropy": 6.200133991241455,
      "epoch": 1.2065781881130986,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000486607166913869,
      "loss": 5.8885,
      "mean_token_accuracy": 0.17823684066534043,
      "num_tokens": 26513494.0,
      "step": 10455
    },
    {
      "entropy": 6.251746702194214,
      "epoch": 1.2071552221581072,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048659313928215705,
      "loss": 5.8786,
      "mean_token_accuracy": 0.18445879220962524,
      "num_tokens": 26525931.0,
      "step": 10460
    },
    {
      "entropy": 6.277969169616699,
      "epoch": 1.207732256203116,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004865791045336149,
      "loss": 5.9407,
      "mean_token_accuracy": 0.17653046399354935,
      "num_tokens": 26539445.0,
      "step": 10465
    },
    {
      "entropy": 6.267771673202515,
      "epoch": 1.2083092902481247,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004865650626687146,
      "loss": 5.9413,
      "mean_token_accuracy": 0.17517824470996857,
      "num_tokens": 26551758.0,
      "step": 10470
    },
    {
      "entropy": 6.234518194198609,
      "epoch": 1.2088863242931334,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048655101368792866,
      "loss": 5.8709,
      "mean_token_accuracy": 0.18177429288625718,
      "num_tokens": 26564756.0,
      "step": 10475
    },
    {
      "entropy": 6.158968114852906,
      "epoch": 1.209463358338142,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004865369575917293,
      "loss": 5.8606,
      "mean_token_accuracy": 0.19036365449428558,
      "num_tokens": 26579152.0,
      "step": 10480
    },
    {
      "entropy": 6.187554407119751,
      "epoch": 1.2100403923831506,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048652289438058953,
      "loss": 5.8707,
      "mean_token_accuracy": 0.18196028620004653,
      "num_tokens": 26590749.0,
      "step": 10485
    },
    {
      "entropy": 6.306079912185669,
      "epoch": 1.2106174264281593,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048650882405498226,
      "loss": 5.9852,
      "mean_token_accuracy": 0.1760221853852272,
      "num_tokens": 26603859.0,
      "step": 10490
    },
    {
      "entropy": 6.109821081161499,
      "epoch": 1.211194460473168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004864947466153808,
      "loss": 5.7942,
      "mean_token_accuracy": 0.1986470937728882,
      "num_tokens": 26616002.0,
      "step": 10495
    },
    {
      "entropy": 6.190054035186767,
      "epoch": 1.2117714945181766,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004864806620622585,
      "loss": 5.904,
      "mean_token_accuracy": 0.18009792268276215,
      "num_tokens": 26629092.0,
      "step": 10500
    },
    {
      "entropy": 6.2522890090942385,
      "epoch": 1.2123485285631852,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004864665703960892,
      "loss": 5.8987,
      "mean_token_accuracy": 0.17601345330476761,
      "num_tokens": 26642830.0,
      "step": 10505
    },
    {
      "entropy": 6.2701061248779295,
      "epoch": 1.2129255626081938,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004864524716173469,
      "loss": 6.0476,
      "mean_token_accuracy": 0.1719884142279625,
      "num_tokens": 26656421.0,
      "step": 10510
    },
    {
      "entropy": 6.158242321014404,
      "epoch": 1.2135025966532025,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004864383657265058,
      "loss": 5.8265,
      "mean_token_accuracy": 0.18615743070840834,
      "num_tokens": 26669267.0,
      "step": 10515
    },
    {
      "entropy": 6.2623544216156,
      "epoch": 1.2140796306982111,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00048642425272404016,
      "loss": 5.8812,
      "mean_token_accuracy": 0.18154094368219376,
      "num_tokens": 26681074.0,
      "step": 10520
    },
    {
      "entropy": 6.120004653930664,
      "epoch": 1.2146566647432198,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004864101326104248,
      "loss": 5.8522,
      "mean_token_accuracy": 0.18834602385759353,
      "num_tokens": 26694772.0,
      "step": 10525
    },
    {
      "entropy": 6.258338356018067,
      "epoch": 1.2152336987882286,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004863960053861348,
      "loss": 5.9345,
      "mean_token_accuracy": 0.1835750326514244,
      "num_tokens": 26706173.0,
      "step": 10530
    },
    {
      "entropy": 6.271576881408691,
      "epoch": 1.215810732833237,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048638187105164507,
      "loss": 5.9731,
      "mean_token_accuracy": 0.17596358358860015,
      "num_tokens": 26718273.0,
      "step": 10535
    },
    {
      "entropy": 6.277022218704223,
      "epoch": 1.216387766878246,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004863677296074311,
      "loss": 5.9622,
      "mean_token_accuracy": 0.18062706291675568,
      "num_tokens": 26730781.0,
      "step": 10540
    },
    {
      "entropy": 6.175031661987305,
      "epoch": 1.2169648009232545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004863535810539686,
      "loss": 5.9047,
      "mean_token_accuracy": 0.18018955141305923,
      "num_tokens": 26743661.0,
      "step": 10545
    },
    {
      "entropy": 6.153426218032837,
      "epoch": 1.2175418349682632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048633942539173325,
      "loss": 5.8325,
      "mean_token_accuracy": 0.18449801206588745,
      "num_tokens": 26755307.0,
      "step": 10550
    },
    {
      "entropy": 6.225366735458374,
      "epoch": 1.2181188690132718,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048632526262120144,
      "loss": 5.884,
      "mean_token_accuracy": 0.17940303832292556,
      "num_tokens": 26769679.0,
      "step": 10555
    },
    {
      "entropy": 6.2013906955719,
      "epoch": 1.2186959030582805,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004863110927428493,
      "loss": 5.7885,
      "mean_token_accuracy": 0.187868070602417,
      "num_tokens": 26782141.0,
      "step": 10560
    },
    {
      "entropy": 6.279289627075196,
      "epoch": 1.219272937103289,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004862969157571536,
      "loss": 5.9343,
      "mean_token_accuracy": 0.18196739703416825,
      "num_tokens": 26794721.0,
      "step": 10565
    },
    {
      "entropy": 6.209691190719605,
      "epoch": 1.2198499711482977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048628273166459105,
      "loss": 5.8373,
      "mean_token_accuracy": 0.18883996307849885,
      "num_tokens": 26807844.0,
      "step": 10570
    },
    {
      "entropy": 6.2086797714233395,
      "epoch": 1.2204270051933064,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048626854046563876,
      "loss": 5.9307,
      "mean_token_accuracy": 0.1777234122157097,
      "num_tokens": 26820150.0,
      "step": 10575
    },
    {
      "entropy": 6.270304155349732,
      "epoch": 1.221004039238315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048625434216077404,
      "loss": 5.9057,
      "mean_token_accuracy": 0.18113289326429366,
      "num_tokens": 26833949.0,
      "step": 10580
    },
    {
      "entropy": 6.170538091659546,
      "epoch": 1.2215810732833237,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00048624013675047453,
      "loss": 5.8129,
      "mean_token_accuracy": 0.1849739983677864,
      "num_tokens": 26847637.0,
      "step": 10585
    },
    {
      "entropy": 6.219944906234741,
      "epoch": 1.2221581073283323,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048622592423521783,
      "loss": 5.8991,
      "mean_token_accuracy": 0.1873820409178734,
      "num_tokens": 26859173.0,
      "step": 10590
    },
    {
      "entropy": 6.217339086532593,
      "epoch": 1.222735141373341,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004862117046154822,
      "loss": 5.9168,
      "mean_token_accuracy": 0.181583671271801,
      "num_tokens": 26870883.0,
      "step": 10595
    },
    {
      "entropy": 6.301852369308472,
      "epoch": 1.2233121754183496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048619747789174577,
      "loss": 5.9429,
      "mean_token_accuracy": 0.17812787294387816,
      "num_tokens": 26882813.0,
      "step": 10600
    },
    {
      "entropy": 6.152983570098877,
      "epoch": 1.2238892094633584,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004861832440644871,
      "loss": 5.8425,
      "mean_token_accuracy": 0.18726977556943894,
      "num_tokens": 26893928.0,
      "step": 10605
    },
    {
      "entropy": 6.112976455688477,
      "epoch": 1.224466243508367,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048616900313418486,
      "loss": 5.847,
      "mean_token_accuracy": 0.19203756004571915,
      "num_tokens": 26907431.0,
      "step": 10610
    },
    {
      "entropy": 6.32542233467102,
      "epoch": 1.2250432775533757,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048615475510131815,
      "loss": 5.9232,
      "mean_token_accuracy": 0.17866321057081222,
      "num_tokens": 26920278.0,
      "step": 10615
    },
    {
      "entropy": 6.319180536270141,
      "epoch": 1.2256203115983844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048614049996636614,
      "loss": 6.0067,
      "mean_token_accuracy": 0.174865660071373,
      "num_tokens": 26933348.0,
      "step": 10620
    },
    {
      "entropy": 6.245016145706177,
      "epoch": 1.226197345643393,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048612623772980825,
      "loss": 5.9567,
      "mean_token_accuracy": 0.17695067673921586,
      "num_tokens": 26946017.0,
      "step": 10625
    },
    {
      "entropy": 6.131003570556641,
      "epoch": 1.2267743796884016,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048611196839212426,
      "loss": 5.7407,
      "mean_token_accuracy": 0.1904704287648201,
      "num_tokens": 26960230.0,
      "step": 10630
    },
    {
      "entropy": 6.277625894546508,
      "epoch": 1.2273514137334103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000486097691953794,
      "loss": 5.9388,
      "mean_token_accuracy": 0.17718097567558289,
      "num_tokens": 26971811.0,
      "step": 10635
    },
    {
      "entropy": 6.273577499389648,
      "epoch": 1.227928447778419,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048608340841529784,
      "loss": 5.9199,
      "mean_token_accuracy": 0.18174241185188295,
      "num_tokens": 26983098.0,
      "step": 10640
    },
    {
      "entropy": 6.234274244308471,
      "epoch": 1.2285054818234276,
      "grad_norm": 0.921875,
      "learning_rate": 0.000486069117777116,
      "loss": 5.9388,
      "mean_token_accuracy": 0.16956097930669783,
      "num_tokens": 26996410.0,
      "step": 10645
    },
    {
      "entropy": 6.306857490539551,
      "epoch": 1.2290825158684362,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004860548200397293,
      "loss": 5.9422,
      "mean_token_accuracy": 0.17670237571001052,
      "num_tokens": 27009329.0,
      "step": 10650
    },
    {
      "entropy": 6.227819299697876,
      "epoch": 1.2296595499134448,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048604051520361846,
      "loss": 5.865,
      "mean_token_accuracy": 0.18356099277734755,
      "num_tokens": 27022261.0,
      "step": 10655
    },
    {
      "entropy": 6.267890167236328,
      "epoch": 1.2302365839584535,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048602620326926484,
      "loss": 5.9625,
      "mean_token_accuracy": 0.1835799053311348,
      "num_tokens": 27035006.0,
      "step": 10660
    },
    {
      "entropy": 6.194986963272095,
      "epoch": 1.2308136180034621,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048601188423714965,
      "loss": 5.8461,
      "mean_token_accuracy": 0.18581847697496415,
      "num_tokens": 27048481.0,
      "step": 10665
    },
    {
      "entropy": 6.285878229141235,
      "epoch": 1.231390652048471,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004859975581077545,
      "loss": 5.9902,
      "mean_token_accuracy": 0.1755063235759735,
      "num_tokens": 27061083.0,
      "step": 10670
    },
    {
      "entropy": 6.287797164916992,
      "epoch": 1.2319676860934794,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004859832248815614,
      "loss": 5.9218,
      "mean_token_accuracy": 0.179109850525856,
      "num_tokens": 27073547.0,
      "step": 10675
    },
    {
      "entropy": 6.207475233078003,
      "epoch": 1.2325447201384883,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004859688845590522,
      "loss": 5.903,
      "mean_token_accuracy": 0.18531017899513244,
      "num_tokens": 27085739.0,
      "step": 10680
    },
    {
      "entropy": 6.19006838798523,
      "epoch": 1.233121754183497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048595453714070944,
      "loss": 5.8995,
      "mean_token_accuracy": 0.183668914437294,
      "num_tokens": 27097331.0,
      "step": 10685
    },
    {
      "entropy": 6.343549060821533,
      "epoch": 1.2336987882285055,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004859401826270156,
      "loss": 5.9311,
      "mean_token_accuracy": 0.18099030256271362,
      "num_tokens": 27109181.0,
      "step": 10690
    },
    {
      "entropy": 6.210108757019043,
      "epoch": 1.2342758222735142,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004859258210184536,
      "loss": 5.8374,
      "mean_token_accuracy": 0.18753527104854584,
      "num_tokens": 27120978.0,
      "step": 10695
    },
    {
      "entropy": 6.291655874252319,
      "epoch": 1.2348528563185228,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048591145231550623,
      "loss": 5.8453,
      "mean_token_accuracy": 0.19009490311145782,
      "num_tokens": 27135243.0,
      "step": 10700
    },
    {
      "entropy": 6.1215228080749515,
      "epoch": 1.2354298903635315,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048589707651865697,
      "loss": 5.8195,
      "mean_token_accuracy": 0.18755590319633483,
      "num_tokens": 27149366.0,
      "step": 10705
    },
    {
      "entropy": 6.183349609375,
      "epoch": 1.23600692440854,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004858826936283893,
      "loss": 5.9302,
      "mean_token_accuracy": 0.17416223585605622,
      "num_tokens": 27162552.0,
      "step": 10710
    },
    {
      "entropy": 6.235607481002807,
      "epoch": 1.2365839584535487,
      "grad_norm": 0.859375,
      "learning_rate": 0.000485868303645187,
      "loss": 5.9041,
      "mean_token_accuracy": 0.17430349588394164,
      "num_tokens": 27175312.0,
      "step": 10715
    },
    {
      "entropy": 6.257982540130615,
      "epoch": 1.2371609924985574,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048585390656953397,
      "loss": 5.8728,
      "mean_token_accuracy": 0.1754424527287483,
      "num_tokens": 27187730.0,
      "step": 10720
    },
    {
      "entropy": 6.191701173782349,
      "epoch": 1.237738026543566,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004858395024019147,
      "loss": 5.8606,
      "mean_token_accuracy": 0.18815997540950774,
      "num_tokens": 27200191.0,
      "step": 10725
    },
    {
      "entropy": 6.227496719360351,
      "epoch": 1.2383150605885747,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004858250911428133,
      "loss": 5.9041,
      "mean_token_accuracy": 0.18517004400491716,
      "num_tokens": 27211905.0,
      "step": 10730
    },
    {
      "entropy": 6.256813335418701,
      "epoch": 1.2388920946335833,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004858106727927148,
      "loss": 5.8841,
      "mean_token_accuracy": 0.1832223802804947,
      "num_tokens": 27224153.0,
      "step": 10735
    },
    {
      "entropy": 6.2418036460876465,
      "epoch": 1.239469128678592,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000485796247352104,
      "loss": 5.9085,
      "mean_token_accuracy": 0.17539570927619935,
      "num_tokens": 27237167.0,
      "step": 10740
    },
    {
      "entropy": 6.226529169082641,
      "epoch": 1.2400461627236008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004857818148214662,
      "loss": 5.8998,
      "mean_token_accuracy": 0.18224241435527802,
      "num_tokens": 27247956.0,
      "step": 10745
    },
    {
      "entropy": 6.225143718719482,
      "epoch": 1.2406231967686094,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004857673752012866,
      "loss": 5.9231,
      "mean_token_accuracy": 0.17687112092971802,
      "num_tokens": 27260357.0,
      "step": 10750
    },
    {
      "entropy": 6.233739233016967,
      "epoch": 1.241200230813618,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048575292849205114,
      "loss": 5.8377,
      "mean_token_accuracy": 0.1787156268954277,
      "num_tokens": 27273953.0,
      "step": 10755
    },
    {
      "entropy": 6.18577618598938,
      "epoch": 1.2417772648586267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004857384746942456,
      "loss": 5.9035,
      "mean_token_accuracy": 0.18345262706279755,
      "num_tokens": 27285550.0,
      "step": 10760
    },
    {
      "entropy": 6.270837783813477,
      "epoch": 1.2423542989036354,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004857240138083562,
      "loss": 5.9638,
      "mean_token_accuracy": 0.17507691383361818,
      "num_tokens": 27297817.0,
      "step": 10765
    },
    {
      "entropy": 6.254739141464233,
      "epoch": 1.242931332948644,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004857095458348692,
      "loss": 5.9844,
      "mean_token_accuracy": 0.17733536213636397,
      "num_tokens": 27310398.0,
      "step": 10770
    },
    {
      "entropy": 6.279214096069336,
      "epoch": 1.2435083669936526,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048569507077427134,
      "loss": 5.9343,
      "mean_token_accuracy": 0.1735254243016243,
      "num_tokens": 27322462.0,
      "step": 10775
    },
    {
      "entropy": 6.388853359222412,
      "epoch": 1.2440854010386613,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004856805886270494,
      "loss": 5.9646,
      "mean_token_accuracy": 0.1786048084497452,
      "num_tokens": 27335301.0,
      "step": 10780
    },
    {
      "entropy": 6.171620941162109,
      "epoch": 1.24466243508367,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004856660993936905,
      "loss": 5.8295,
      "mean_token_accuracy": 0.18786240220069886,
      "num_tokens": 27347425.0,
      "step": 10785
    },
    {
      "entropy": 6.226814031600952,
      "epoch": 1.2452394691286786,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000485651603074682,
      "loss": 5.9241,
      "mean_token_accuracy": 0.18220330625772477,
      "num_tokens": 27361149.0,
      "step": 10790
    },
    {
      "entropy": 6.230698013305664,
      "epoch": 1.2458165031736872,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004856370996705115,
      "loss": 5.8823,
      "mean_token_accuracy": 0.18632961809635162,
      "num_tokens": 27373055.0,
      "step": 10795
    },
    {
      "entropy": 6.213808155059814,
      "epoch": 1.2463935372186958,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004856225891816667,
      "loss": 5.9587,
      "mean_token_accuracy": 0.18557358533143997,
      "num_tokens": 27385279.0,
      "step": 10800
    },
    {
      "entropy": 6.2369311332702635,
      "epoch": 1.2469705712637045,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004856080716086358,
      "loss": 5.9224,
      "mean_token_accuracy": 0.18104355782270432,
      "num_tokens": 27399035.0,
      "step": 10805
    },
    {
      "entropy": 6.264719486236572,
      "epoch": 1.2475476053087133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000485593546951907,
      "loss": 5.9186,
      "mean_token_accuracy": 0.18508729040622712,
      "num_tokens": 27410473.0,
      "step": 10810
    },
    {
      "entropy": 6.295024299621582,
      "epoch": 1.2481246393537218,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004855790152119688,
      "loss": 5.9116,
      "mean_token_accuracy": 0.18057319670915603,
      "num_tokens": 27422194.0,
      "step": 10815
    },
    {
      "entropy": 6.195057725906372,
      "epoch": 1.2487016733987306,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004855644763893102,
      "loss": 5.8991,
      "mean_token_accuracy": 0.1836089238524437,
      "num_tokens": 27434800.0,
      "step": 10820
    },
    {
      "entropy": 6.145266151428222,
      "epoch": 1.2492787074437393,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048554993048441987,
      "loss": 5.7857,
      "mean_token_accuracy": 0.19063047766685487,
      "num_tokens": 27447002.0,
      "step": 10825
    },
    {
      "entropy": 6.3162415504455565,
      "epoch": 1.249855741488748,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004855353774977873,
      "loss": 5.9918,
      "mean_token_accuracy": 0.17585002183914183,
      "num_tokens": 27461461.0,
      "step": 10830
    },
    {
      "entropy": 6.218456411361695,
      "epoch": 1.2504327755337565,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048552081742990184,
      "loss": 5.9805,
      "mean_token_accuracy": 0.17204463481903076,
      "num_tokens": 27475062.0,
      "step": 10835
    },
    {
      "entropy": 6.222698593139649,
      "epoch": 1.2510098095787652,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048550625028125327,
      "loss": 5.8455,
      "mean_token_accuracy": 0.18781229108572006,
      "num_tokens": 27487812.0,
      "step": 10840
    },
    {
      "entropy": 6.207841777801514,
      "epoch": 1.2515868436237738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004854916760523315,
      "loss": 5.8881,
      "mean_token_accuracy": 0.18616320043802262,
      "num_tokens": 27501351.0,
      "step": 10845
    },
    {
      "entropy": 6.234964418411255,
      "epoch": 1.2521638776687825,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048547709474362684,
      "loss": 5.8595,
      "mean_token_accuracy": 0.18577065020799638,
      "num_tokens": 27514651.0,
      "step": 10850
    },
    {
      "entropy": 6.318981122970581,
      "epoch": 1.252740911713791,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004854625063556296,
      "loss": 5.984,
      "mean_token_accuracy": 0.1744752883911133,
      "num_tokens": 27527037.0,
      "step": 10855
    },
    {
      "entropy": 6.253472375869751,
      "epoch": 1.2533179457587997,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004854479108888305,
      "loss": 5.9672,
      "mean_token_accuracy": 0.1773965671658516,
      "num_tokens": 27540800.0,
      "step": 10860
    },
    {
      "entropy": 6.235241842269898,
      "epoch": 1.2538949798038084,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048543330834372047,
      "loss": 5.878,
      "mean_token_accuracy": 0.1835606426000595,
      "num_tokens": 27553873.0,
      "step": 10865
    },
    {
      "entropy": 6.168224048614502,
      "epoch": 1.254472013848817,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048541869872079064,
      "loss": 5.8891,
      "mean_token_accuracy": 0.18331090360879898,
      "num_tokens": 27566336.0,
      "step": 10870
    },
    {
      "entropy": 6.299077558517456,
      "epoch": 1.2550490478938257,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004854040820205324,
      "loss": 5.9985,
      "mean_token_accuracy": 0.17636170685291291,
      "num_tokens": 27578677.0,
      "step": 10875
    },
    {
      "entropy": 6.32140531539917,
      "epoch": 1.2556260819388343,
      "grad_norm": 0.875,
      "learning_rate": 0.0004853894582434373,
      "loss": 5.9481,
      "mean_token_accuracy": 0.18338159322738648,
      "num_tokens": 27590734.0,
      "step": 10880
    },
    {
      "entropy": 6.228301954269409,
      "epoch": 1.2562031159838432,
      "grad_norm": 0.875,
      "learning_rate": 0.0004853748273899974,
      "loss": 5.8949,
      "mean_token_accuracy": 0.17739989310503007,
      "num_tokens": 27603193.0,
      "step": 10885
    },
    {
      "entropy": 6.195730686187744,
      "epoch": 1.2567801500288516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004853601894607046,
      "loss": 5.8547,
      "mean_token_accuracy": 0.18741325587034224,
      "num_tokens": 27615248.0,
      "step": 10890
    },
    {
      "entropy": 6.26179986000061,
      "epoch": 1.2573571840738604,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004853455444560514,
      "loss": 5.9447,
      "mean_token_accuracy": 0.18093649297952652,
      "num_tokens": 27627384.0,
      "step": 10895
    },
    {
      "entropy": 6.245268106460571,
      "epoch": 1.257934218118869,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004853308923765302,
      "loss": 5.9537,
      "mean_token_accuracy": 0.176744045317173,
      "num_tokens": 27641014.0,
      "step": 10900
    },
    {
      "entropy": 6.233709287643433,
      "epoch": 1.2585112521638777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000485316233222634,
      "loss": 5.8089,
      "mean_token_accuracy": 0.19169621169567108,
      "num_tokens": 27652325.0,
      "step": 10905
    },
    {
      "entropy": 6.234284591674805,
      "epoch": 1.2590882862088864,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004853015669948557,
      "loss": 5.8979,
      "mean_token_accuracy": 0.18115110993385314,
      "num_tokens": 27664365.0,
      "step": 10910
    },
    {
      "entropy": 6.211205530166626,
      "epoch": 1.259665320253895,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048528689369368863,
      "loss": 5.8122,
      "mean_token_accuracy": 0.18548589795827866,
      "num_tokens": 27677657.0,
      "step": 10915
    },
    {
      "entropy": 6.164569282531739,
      "epoch": 1.2602423542989036,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004852722133196264,
      "loss": 5.7457,
      "mean_token_accuracy": 0.19496161937713624,
      "num_tokens": 27689598.0,
      "step": 10920
    },
    {
      "entropy": 6.142652988433838,
      "epoch": 1.2608193883439123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004852575258731627,
      "loss": 5.7812,
      "mean_token_accuracy": 0.19326534271240234,
      "num_tokens": 27702573.0,
      "step": 10925
    },
    {
      "entropy": 6.221087074279785,
      "epoch": 1.261396422388921,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004852428313547916,
      "loss": 5.9266,
      "mean_token_accuracy": 0.18208504021167754,
      "num_tokens": 27714506.0,
      "step": 10930
    },
    {
      "entropy": 6.0959484577178955,
      "epoch": 1.2619734564339296,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048522812976500726,
      "loss": 5.6263,
      "mean_token_accuracy": 0.2131839096546173,
      "num_tokens": 27728308.0,
      "step": 10935
    },
    {
      "entropy": 6.20897798538208,
      "epoch": 1.2625504904789382,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048521342110430417,
      "loss": 5.9055,
      "mean_token_accuracy": 0.1780867025256157,
      "num_tokens": 27741521.0,
      "step": 10940
    },
    {
      "entropy": 6.283983612060547,
      "epoch": 1.2631275245239468,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00048519870537317713,
      "loss": 5.95,
      "mean_token_accuracy": 0.17134395986795425,
      "num_tokens": 27754033.0,
      "step": 10945
    },
    {
      "entropy": 6.2703940868377686,
      "epoch": 1.2637045585689557,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048518398257212103,
      "loss": 5.9433,
      "mean_token_accuracy": 0.1823036938905716,
      "num_tokens": 27766557.0,
      "step": 10950
    },
    {
      "entropy": 6.212624263763428,
      "epoch": 1.2642815926139641,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004851692527016311,
      "loss": 5.8816,
      "mean_token_accuracy": 0.18911558240652085,
      "num_tokens": 27778018.0,
      "step": 10955
    },
    {
      "entropy": 6.216017198562622,
      "epoch": 1.264858626658973,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004851545157622027,
      "loss": 5.8461,
      "mean_token_accuracy": 0.1850295215845108,
      "num_tokens": 27789514.0,
      "step": 10960
    },
    {
      "entropy": 6.237737607955933,
      "epoch": 1.2654356607039816,
      "grad_norm": 0.875,
      "learning_rate": 0.0004851397717543316,
      "loss": 5.9017,
      "mean_token_accuracy": 0.18139505088329316,
      "num_tokens": 27802759.0,
      "step": 10965
    },
    {
      "entropy": 6.165145397186279,
      "epoch": 1.2660126947489903,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004851250206785137,
      "loss": 5.837,
      "mean_token_accuracy": 0.18212868571281432,
      "num_tokens": 27817064.0,
      "step": 10970
    },
    {
      "entropy": 6.2120749950408936,
      "epoch": 1.266589728793999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048511026253524503,
      "loss": 5.8825,
      "mean_token_accuracy": 0.1782068982720375,
      "num_tokens": 27830627.0,
      "step": 10975
    },
    {
      "entropy": 6.173754358291626,
      "epoch": 1.2671667628390075,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004850954973250221,
      "loss": 5.7905,
      "mean_token_accuracy": 0.1953563630580902,
      "num_tokens": 27842745.0,
      "step": 10980
    },
    {
      "entropy": 6.241191816329956,
      "epoch": 1.2677437968840162,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048508072504834144,
      "loss": 5.9272,
      "mean_token_accuracy": 0.1826498791575432,
      "num_tokens": 27855720.0,
      "step": 10985
    },
    {
      "entropy": 6.162121534347534,
      "epoch": 1.2683208309290248,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048506594570569997,
      "loss": 5.813,
      "mean_token_accuracy": 0.19384430795907975,
      "num_tokens": 27869019.0,
      "step": 10990
    },
    {
      "entropy": 6.239211940765381,
      "epoch": 1.2688978649740335,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004850511592975947,
      "loss": 5.9603,
      "mean_token_accuracy": 0.17428394109010698,
      "num_tokens": 27883563.0,
      "step": 10995
    },
    {
      "entropy": 6.205305671691894,
      "epoch": 1.269474899019042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004850363658245231,
      "loss": 5.8958,
      "mean_token_accuracy": 0.1830536961555481,
      "num_tokens": 27895980.0,
      "step": 11000
    },
    {
      "entropy": 6.1944968700408936,
      "epoch": 1.2700519330640507,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004850215652869826,
      "loss": 5.9488,
      "mean_token_accuracy": 0.1719176724553108,
      "num_tokens": 27907535.0,
      "step": 11005
    },
    {
      "entropy": 6.35361614227295,
      "epoch": 1.2706289671090594,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004850067576854711,
      "loss": 5.8952,
      "mean_token_accuracy": 0.18115041255950928,
      "num_tokens": 27920665.0,
      "step": 11010
    },
    {
      "entropy": 6.160195446014404,
      "epoch": 1.271206001154068,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004849919430204867,
      "loss": 5.7931,
      "mean_token_accuracy": 0.19305038899183274,
      "num_tokens": 27932178.0,
      "step": 11015
    },
    {
      "entropy": 6.192092657089233,
      "epoch": 1.2717830351990767,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004849771212925275,
      "loss": 5.8532,
      "mean_token_accuracy": 0.18312882333993913,
      "num_tokens": 27945319.0,
      "step": 11020
    },
    {
      "entropy": 6.1728551387786865,
      "epoch": 1.2723600692440855,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048496229250209223,
      "loss": 5.8611,
      "mean_token_accuracy": 0.18547921180725097,
      "num_tokens": 27957765.0,
      "step": 11025
    },
    {
      "entropy": 6.2221180438995365,
      "epoch": 1.272937103289094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004849474566496795,
      "loss": 5.8693,
      "mean_token_accuracy": 0.19211741536855698,
      "num_tokens": 27971306.0,
      "step": 11030
    },
    {
      "entropy": 6.179285287857056,
      "epoch": 1.2735141373341028,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004849326137357883,
      "loss": 5.8139,
      "mean_token_accuracy": 0.18820831030607224,
      "num_tokens": 27982998.0,
      "step": 11035
    },
    {
      "entropy": 6.201269102096558,
      "epoch": 1.2740911713791114,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000484917763760918,
      "loss": 5.9292,
      "mean_token_accuracy": 0.18131004124879838,
      "num_tokens": 27995792.0,
      "step": 11040
    },
    {
      "entropy": 6.18898344039917,
      "epoch": 1.27466820542412,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00048490290672556784,
      "loss": 5.7945,
      "mean_token_accuracy": 0.18836279660463334,
      "num_tokens": 28008882.0,
      "step": 11045
    },
    {
      "entropy": 6.215117025375366,
      "epoch": 1.2752452394691287,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004848880426302378,
      "loss": 5.9267,
      "mean_token_accuracy": 0.18029106110334397,
      "num_tokens": 28022768.0,
      "step": 11050
    },
    {
      "entropy": 6.212264347076416,
      "epoch": 1.2758222735141374,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004848731714754277,
      "loss": 5.8546,
      "mean_token_accuracy": 0.1832072451710701,
      "num_tokens": 28035196.0,
      "step": 11055
    },
    {
      "entropy": 6.346293973922729,
      "epoch": 1.276399307559146,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004848582932616377,
      "loss": 6.0246,
      "mean_token_accuracy": 0.16925484985113143,
      "num_tokens": 28048089.0,
      "step": 11060
    },
    {
      "entropy": 6.264890193939209,
      "epoch": 1.2769763416041546,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004848434079893682,
      "loss": 5.9431,
      "mean_token_accuracy": 0.1741225779056549,
      "num_tokens": 28060140.0,
      "step": 11065
    },
    {
      "entropy": 6.279630327224732,
      "epoch": 1.2775533756491633,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004848285156591201,
      "loss": 5.904,
      "mean_token_accuracy": 0.17972956001758575,
      "num_tokens": 28072448.0,
      "step": 11070
    },
    {
      "entropy": 6.299330806732177,
      "epoch": 1.278130409694172,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048481361627139384,
      "loss": 5.9808,
      "mean_token_accuracy": 0.17580227553844452,
      "num_tokens": 28084696.0,
      "step": 11075
    },
    {
      "entropy": 6.263509035110474,
      "epoch": 1.2787074437391805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048479870982669096,
      "loss": 5.9356,
      "mean_token_accuracy": 0.18625864684581755,
      "num_tokens": 28097834.0,
      "step": 11080
    },
    {
      "entropy": 6.195324182510376,
      "epoch": 1.2792844777841892,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048478379632551266,
      "loss": 5.819,
      "mean_token_accuracy": 0.19088034480810165,
      "num_tokens": 28110066.0,
      "step": 11085
    },
    {
      "entropy": 6.239222526550293,
      "epoch": 1.279861511829198,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004847688757683606,
      "loss": 5.9292,
      "mean_token_accuracy": 0.18112271279096603,
      "num_tokens": 28122617.0,
      "step": 11090
    },
    {
      "entropy": 6.177471590042114,
      "epoch": 1.2804385458742065,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004847539481557366,
      "loss": 5.8293,
      "mean_token_accuracy": 0.18178199976682663,
      "num_tokens": 28134987.0,
      "step": 11095
    },
    {
      "entropy": 6.306794786453247,
      "epoch": 1.2810155799192153,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048473901348814277,
      "loss": 5.8937,
      "mean_token_accuracy": 0.17665995657444,
      "num_tokens": 28147478.0,
      "step": 11100
    },
    {
      "entropy": 6.251819658279419,
      "epoch": 1.281592613964224,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004847240717660814,
      "loss": 5.9244,
      "mean_token_accuracy": 0.18358440846204757,
      "num_tokens": 28159010.0,
      "step": 11105
    },
    {
      "entropy": 6.200797462463379,
      "epoch": 1.2821696480092326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048470912299005493,
      "loss": 5.9463,
      "mean_token_accuracy": 0.17590773105621338,
      "num_tokens": 28170579.0,
      "step": 11110
    },
    {
      "entropy": 6.257146501541138,
      "epoch": 1.2827466820542412,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048469416716056635,
      "loss": 5.8812,
      "mean_token_accuracy": 0.18537105917930602,
      "num_tokens": 28182569.0,
      "step": 11115
    },
    {
      "entropy": 6.304178762435913,
      "epoch": 1.28332371609925,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004846792042781187,
      "loss": 5.9716,
      "mean_token_accuracy": 0.17825240939855574,
      "num_tokens": 28195736.0,
      "step": 11120
    },
    {
      "entropy": 6.228730297088623,
      "epoch": 1.2839007501442585,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048466423434321513,
      "loss": 5.9442,
      "mean_token_accuracy": 0.18216853886842727,
      "num_tokens": 28210944.0,
      "step": 11125
    },
    {
      "entropy": 6.244191837310791,
      "epoch": 1.2844777841892672,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048464925735635907,
      "loss": 5.8904,
      "mean_token_accuracy": 0.18087451457977294,
      "num_tokens": 28223674.0,
      "step": 11130
    },
    {
      "entropy": 6.281203126907348,
      "epoch": 1.2850548182342758,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048463427331805445,
      "loss": 5.9496,
      "mean_token_accuracy": 0.1721128135919571,
      "num_tokens": 28236643.0,
      "step": 11135
    },
    {
      "entropy": 6.245355081558228,
      "epoch": 1.2856318522792844,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004846192822288052,
      "loss": 5.9319,
      "mean_token_accuracy": 0.17700979709625245,
      "num_tokens": 28249678.0,
      "step": 11140
    },
    {
      "entropy": 6.196260690689087,
      "epoch": 1.286208886324293,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004846042840891155,
      "loss": 5.8687,
      "mean_token_accuracy": 0.18645038157701493,
      "num_tokens": 28262571.0,
      "step": 11145
    },
    {
      "entropy": 6.205413866043091,
      "epoch": 1.2867859203693017,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004845892788994899,
      "loss": 5.8878,
      "mean_token_accuracy": 0.1781436026096344,
      "num_tokens": 28275450.0,
      "step": 11150
    },
    {
      "entropy": 6.265288972854615,
      "epoch": 1.2873629544143104,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004845742666604329,
      "loss": 5.8481,
      "mean_token_accuracy": 0.18640822172164917,
      "num_tokens": 28288224.0,
      "step": 11155
    },
    {
      "entropy": 6.080491209030152,
      "epoch": 1.287939988459319,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048455924737244953,
      "loss": 5.6821,
      "mean_token_accuracy": 0.20160522162914277,
      "num_tokens": 28300515.0,
      "step": 11160
    },
    {
      "entropy": 6.223293685913086,
      "epoch": 1.2885170225043279,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048454422103604505,
      "loss": 5.9138,
      "mean_token_accuracy": 0.18207870572805404,
      "num_tokens": 28313729.0,
      "step": 11165
    },
    {
      "entropy": 6.229456186294556,
      "epoch": 1.2890940565493363,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004845291876517247,
      "loss": 5.8415,
      "mean_token_accuracy": 0.1863294705748558,
      "num_tokens": 28325872.0,
      "step": 11170
    },
    {
      "entropy": 6.28121018409729,
      "epoch": 1.2896710905943451,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048451414721999424,
      "loss": 5.8926,
      "mean_token_accuracy": 0.18683222979307174,
      "num_tokens": 28337820.0,
      "step": 11175
    },
    {
      "entropy": 6.236246109008789,
      "epoch": 1.2902481246393538,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048449909974135954,
      "loss": 5.9102,
      "mean_token_accuracy": 0.18545778095722198,
      "num_tokens": 28350608.0,
      "step": 11180
    },
    {
      "entropy": 6.1874621391296385,
      "epoch": 1.2908251586843624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004844840452163267,
      "loss": 5.8364,
      "mean_token_accuracy": 0.18310509771108627,
      "num_tokens": 28362229.0,
      "step": 11185
    },
    {
      "entropy": 6.182185697555542,
      "epoch": 1.291402192729371,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000484468983645402,
      "loss": 5.8288,
      "mean_token_accuracy": 0.18478621542453766,
      "num_tokens": 28374767.0,
      "step": 11190
    },
    {
      "entropy": 6.160684967041016,
      "epoch": 1.2919792267743797,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048445391502909213,
      "loss": 5.8068,
      "mean_token_accuracy": 0.1793937399983406,
      "num_tokens": 28387401.0,
      "step": 11195
    },
    {
      "entropy": 6.148256254196167,
      "epoch": 1.2925562608193883,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048443883936790386,
      "loss": 5.8714,
      "mean_token_accuracy": 0.1865007311105728,
      "num_tokens": 28400123.0,
      "step": 11200
    },
    {
      "entropy": 6.244810009002686,
      "epoch": 1.293133294864397,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048442375666234427,
      "loss": 5.9275,
      "mean_token_accuracy": 0.1865316689014435,
      "num_tokens": 28411607.0,
      "step": 11205
    },
    {
      "entropy": 6.298503971099853,
      "epoch": 1.2937103289094056,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004844086669129206,
      "loss": 5.9041,
      "mean_token_accuracy": 0.17625187188386918,
      "num_tokens": 28424299.0,
      "step": 11210
    },
    {
      "entropy": 6.29367241859436,
      "epoch": 1.2942873629544143,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048439357012014045,
      "loss": 5.9829,
      "mean_token_accuracy": 0.18334121853113175,
      "num_tokens": 28438646.0,
      "step": 11215
    },
    {
      "entropy": 6.2231042861938475,
      "epoch": 1.294864396999423,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004843784662845116,
      "loss": 5.9187,
      "mean_token_accuracy": 0.1788049578666687,
      "num_tokens": 28452870.0,
      "step": 11220
    },
    {
      "entropy": 6.278390598297119,
      "epoch": 1.2954414310444315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000484363355406542,
      "loss": 5.9678,
      "mean_token_accuracy": 0.17838167697191237,
      "num_tokens": 28465391.0,
      "step": 11225
    },
    {
      "entropy": 6.184773588180542,
      "epoch": 1.2960184650894404,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048434823748674,
      "loss": 5.8141,
      "mean_token_accuracy": 0.18809021711349488,
      "num_tokens": 28477458.0,
      "step": 11230
    },
    {
      "entropy": 6.2303235054016115,
      "epoch": 1.2965954991344488,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048433311252561403,
      "loss": 5.8803,
      "mean_token_accuracy": 0.1797617569565773,
      "num_tokens": 28489092.0,
      "step": 11235
    },
    {
      "entropy": 6.267126274108887,
      "epoch": 1.2971725331794577,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004843179805236728,
      "loss": 5.8906,
      "mean_token_accuracy": 0.19050804376602173,
      "num_tokens": 28501990.0,
      "step": 11240
    },
    {
      "entropy": 6.255610990524292,
      "epoch": 1.297749567224466,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048430284148142516,
      "loss": 5.8569,
      "mean_token_accuracy": 0.18393197506666184,
      "num_tokens": 28515984.0,
      "step": 11245
    },
    {
      "entropy": 6.16234974861145,
      "epoch": 1.298326601269475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004842876953993805,
      "loss": 5.8196,
      "mean_token_accuracy": 0.1858302280306816,
      "num_tokens": 28528093.0,
      "step": 11250
    },
    {
      "entropy": 6.211054754257202,
      "epoch": 1.2989036353144836,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004842725422780482,
      "loss": 5.8786,
      "mean_token_accuracy": 0.18156624287366868,
      "num_tokens": 28541177.0,
      "step": 11255
    },
    {
      "entropy": 6.203247213363648,
      "epoch": 1.2994806693594922,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004842573821179378,
      "loss": 5.8263,
      "mean_token_accuracy": 0.18733591884374617,
      "num_tokens": 28553033.0,
      "step": 11260
    },
    {
      "entropy": 6.057796621322632,
      "epoch": 1.3000577034045009,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004842422149195593,
      "loss": 5.7329,
      "mean_token_accuracy": 0.19602534770965577,
      "num_tokens": 28565787.0,
      "step": 11265
    },
    {
      "entropy": 6.210728502273559,
      "epoch": 1.3006347374495095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048422704068342294,
      "loss": 5.8791,
      "mean_token_accuracy": 0.1837248533964157,
      "num_tokens": 28578823.0,
      "step": 11270
    },
    {
      "entropy": 6.236587333679199,
      "epoch": 1.3012117714945182,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004842118594100389,
      "loss": 5.8589,
      "mean_token_accuracy": 0.18490492999553682,
      "num_tokens": 28590873.0,
      "step": 11275
    },
    {
      "entropy": 6.275784587860107,
      "epoch": 1.3017888055395268,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048419667109991793,
      "loss": 5.8677,
      "mean_token_accuracy": 0.1839105010032654,
      "num_tokens": 28603202.0,
      "step": 11280
    },
    {
      "entropy": 6.2624917984008786,
      "epoch": 1.3023658395845354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048418147575357085,
      "loss": 5.9364,
      "mean_token_accuracy": 0.1802520826458931,
      "num_tokens": 28616803.0,
      "step": 11285
    },
    {
      "entropy": 6.235371446609497,
      "epoch": 1.302942873629544,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004841662733715087,
      "loss": 5.9265,
      "mean_token_accuracy": 0.18370906561613082,
      "num_tokens": 28629439.0,
      "step": 11290
    },
    {
      "entropy": 6.221305084228516,
      "epoch": 1.3035199076745527,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048415106395424294,
      "loss": 5.8409,
      "mean_token_accuracy": 0.19191209226846695,
      "num_tokens": 28642002.0,
      "step": 11295
    },
    {
      "entropy": 6.233723211288452,
      "epoch": 1.3040969417195614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048413584750228494,
      "loss": 5.9276,
      "mean_token_accuracy": 0.17685849070549012,
      "num_tokens": 28653906.0,
      "step": 11300
    },
    {
      "entropy": 6.263261842727661,
      "epoch": 1.3046739757645702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048412062401614653,
      "loss": 5.8631,
      "mean_token_accuracy": 0.1780393034219742,
      "num_tokens": 28666085.0,
      "step": 11305
    },
    {
      "entropy": 6.288744354248047,
      "epoch": 1.3052510098095786,
      "grad_norm": 0.875,
      "learning_rate": 0.00048410539349634,
      "loss": 5.9002,
      "mean_token_accuracy": 0.18006695061922073,
      "num_tokens": 28679298.0,
      "step": 11310
    },
    {
      "entropy": 6.228331756591797,
      "epoch": 1.3058280438545875,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048409015594337725,
      "loss": 5.832,
      "mean_token_accuracy": 0.18585215210914613,
      "num_tokens": 28691146.0,
      "step": 11315
    },
    {
      "entropy": 6.139502239227295,
      "epoch": 1.3064050778995961,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000484074911357771,
      "loss": 5.805,
      "mean_token_accuracy": 0.193204827606678,
      "num_tokens": 28705259.0,
      "step": 11320
    },
    {
      "entropy": 6.277085208892823,
      "epoch": 1.3069821119446048,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048405965974003404,
      "loss": 5.8664,
      "mean_token_accuracy": 0.18343890011310576,
      "num_tokens": 28718004.0,
      "step": 11325
    },
    {
      "entropy": 6.315512657165527,
      "epoch": 1.3075591459896134,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048404440109067927,
      "loss": 5.94,
      "mean_token_accuracy": 0.172566556930542,
      "num_tokens": 28730442.0,
      "step": 11330
    },
    {
      "entropy": 6.176974058151245,
      "epoch": 1.308136180034622,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004840291354102198,
      "loss": 5.8626,
      "mean_token_accuracy": 0.18585751056671143,
      "num_tokens": 28743078.0,
      "step": 11335
    },
    {
      "entropy": 6.193962287902832,
      "epoch": 1.3087132140796307,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004840138626991693,
      "loss": 5.8112,
      "mean_token_accuracy": 0.19411917328834533,
      "num_tokens": 28756975.0,
      "step": 11340
    },
    {
      "entropy": 6.17353081703186,
      "epoch": 1.3092902481246393,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004839985829580413,
      "loss": 5.8615,
      "mean_token_accuracy": 0.19032180905342103,
      "num_tokens": 28770830.0,
      "step": 11345
    },
    {
      "entropy": 6.164017105102539,
      "epoch": 1.309867282169648,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048398329618734977,
      "loss": 5.8813,
      "mean_token_accuracy": 0.18314029425382614,
      "num_tokens": 28782894.0,
      "step": 11350
    },
    {
      "entropy": 6.234577655792236,
      "epoch": 1.3104443162146566,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004839680023876089,
      "loss": 5.9004,
      "mean_token_accuracy": 0.18352519422769548,
      "num_tokens": 28796412.0,
      "step": 11355
    },
    {
      "entropy": 6.258211088180542,
      "epoch": 1.3110213502596653,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004839527015593331,
      "loss": 5.8874,
      "mean_token_accuracy": 0.18190276026725768,
      "num_tokens": 28809004.0,
      "step": 11360
    },
    {
      "entropy": 6.177172517776489,
      "epoch": 1.311598384304674,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048393739370303684,
      "loss": 5.8703,
      "mean_token_accuracy": 0.18527233898639678,
      "num_tokens": 28821104.0,
      "step": 11365
    },
    {
      "entropy": 6.212075662612915,
      "epoch": 1.3121754183496828,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004839220788192353,
      "loss": 5.8582,
      "mean_token_accuracy": 0.1836971163749695,
      "num_tokens": 28833413.0,
      "step": 11370
    },
    {
      "entropy": 6.2552742004394535,
      "epoch": 1.3127524523946912,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048390675690844335,
      "loss": 5.9195,
      "mean_token_accuracy": 0.18566916584968568,
      "num_tokens": 28846233.0,
      "step": 11375
    },
    {
      "entropy": 6.22949743270874,
      "epoch": 1.3133294864397,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004838914279711764,
      "loss": 5.9049,
      "mean_token_accuracy": 0.18077892810106277,
      "num_tokens": 28859296.0,
      "step": 11380
    },
    {
      "entropy": 6.24212155342102,
      "epoch": 1.3139065204847085,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048387609200795003,
      "loss": 5.8705,
      "mean_token_accuracy": 0.18020653128623962,
      "num_tokens": 28872871.0,
      "step": 11385
    },
    {
      "entropy": 6.227920436859131,
      "epoch": 1.3144835545297173,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048386074901928,
      "loss": 5.8728,
      "mean_token_accuracy": 0.18762342929840087,
      "num_tokens": 28885842.0,
      "step": 11390
    },
    {
      "entropy": 6.204128217697144,
      "epoch": 1.315060588574726,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004838453990056825,
      "loss": 5.8685,
      "mean_token_accuracy": 0.18358820378780366,
      "num_tokens": 28898030.0,
      "step": 11395
    },
    {
      "entropy": 6.236626720428466,
      "epoch": 1.3156376226197346,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048383004196767373,
      "loss": 5.9551,
      "mean_token_accuracy": 0.1742781236767769,
      "num_tokens": 28910562.0,
      "step": 11400
    },
    {
      "entropy": 6.389348745346069,
      "epoch": 1.3162146566647432,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004838146779057702,
      "loss": 5.9462,
      "mean_token_accuracy": 0.17498781085014342,
      "num_tokens": 28923169.0,
      "step": 11405
    },
    {
      "entropy": 6.27456111907959,
      "epoch": 1.3167916907097519,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004837993068204887,
      "loss": 5.9501,
      "mean_token_accuracy": 0.17980277687311172,
      "num_tokens": 28935840.0,
      "step": 11410
    },
    {
      "entropy": 6.128229713439941,
      "epoch": 1.3173687247547605,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048378392871234634,
      "loss": 5.8023,
      "mean_token_accuracy": 0.18477270752191544,
      "num_tokens": 28948533.0,
      "step": 11415
    },
    {
      "entropy": 6.1896552562713625,
      "epoch": 1.3179457587997692,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004837685435818601,
      "loss": 5.8504,
      "mean_token_accuracy": 0.18157275319099425,
      "num_tokens": 28961060.0,
      "step": 11420
    },
    {
      "entropy": 6.239120578765869,
      "epoch": 1.3185227928447778,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004837531514295477,
      "loss": 5.8455,
      "mean_token_accuracy": 0.19041848331689834,
      "num_tokens": 28973570.0,
      "step": 11425
    },
    {
      "entropy": 6.142883682250977,
      "epoch": 1.3190998268897864,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004837377522559267,
      "loss": 5.8503,
      "mean_token_accuracy": 0.18818650394678116,
      "num_tokens": 28987519.0,
      "step": 11430
    },
    {
      "entropy": 6.226796340942383,
      "epoch": 1.319676860934795,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048372234606151507,
      "loss": 5.8595,
      "mean_token_accuracy": 0.18620571196079255,
      "num_tokens": 29000790.0,
      "step": 11435
    },
    {
      "entropy": 6.264711380004883,
      "epoch": 1.3202538949798037,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048370693284683106,
      "loss": 5.8852,
      "mean_token_accuracy": 0.170473350584507,
      "num_tokens": 29013263.0,
      "step": 11440
    },
    {
      "entropy": 6.207987976074219,
      "epoch": 1.3208309290248126,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048369151261239303,
      "loss": 5.8496,
      "mean_token_accuracy": 0.19145987778902054,
      "num_tokens": 29024958.0,
      "step": 11445
    },
    {
      "entropy": 6.199948310852051,
      "epoch": 1.321407963069821,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004836760853587196,
      "loss": 5.8533,
      "mean_token_accuracy": 0.18090698421001433,
      "num_tokens": 29037752.0,
      "step": 11450
    },
    {
      "entropy": 6.213617992401123,
      "epoch": 1.3219849971148299,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048366065108632967,
      "loss": 5.7651,
      "mean_token_accuracy": 0.19660505801439285,
      "num_tokens": 29050872.0,
      "step": 11455
    },
    {
      "entropy": 6.2573456287384035,
      "epoch": 1.3225620311598385,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048364520979574246,
      "loss": 5.8451,
      "mean_token_accuracy": 0.1824808672070503,
      "num_tokens": 29062472.0,
      "step": 11460
    },
    {
      "entropy": 6.205182886123657,
      "epoch": 1.3231390652048471,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048362976148747715,
      "loss": 5.905,
      "mean_token_accuracy": 0.18824295401573182,
      "num_tokens": 29075392.0,
      "step": 11465
    },
    {
      "entropy": 6.191983413696289,
      "epoch": 1.3237160992498558,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004836143061620536,
      "loss": 5.8446,
      "mean_token_accuracy": 0.1802932247519493,
      "num_tokens": 29088221.0,
      "step": 11470
    },
    {
      "entropy": 6.265887832641601,
      "epoch": 1.3242931332948644,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004835988438199914,
      "loss": 5.8899,
      "mean_token_accuracy": 0.18557904660701752,
      "num_tokens": 29101050.0,
      "step": 11475
    },
    {
      "entropy": 6.256651878356934,
      "epoch": 1.324870167339873,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004835833744618107,
      "loss": 5.9048,
      "mean_token_accuracy": 0.17918722629547118,
      "num_tokens": 29113928.0,
      "step": 11480
    },
    {
      "entropy": 6.290854024887085,
      "epoch": 1.3254472013848817,
      "grad_norm": 0.875,
      "learning_rate": 0.0004835678980880318,
      "loss": 5.9715,
      "mean_token_accuracy": 0.17373317778110503,
      "num_tokens": 29127196.0,
      "step": 11485
    },
    {
      "entropy": 6.2080904006958,
      "epoch": 1.3260242354298903,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004835524146991753,
      "loss": 5.86,
      "mean_token_accuracy": 0.18338795155286788,
      "num_tokens": 29140498.0,
      "step": 11490
    },
    {
      "entropy": 6.180888605117798,
      "epoch": 1.326601269474899,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048353692429576185,
      "loss": 5.7341,
      "mean_token_accuracy": 0.19421834796667098,
      "num_tokens": 29153910.0,
      "step": 11495
    },
    {
      "entropy": 6.2341631889343265,
      "epoch": 1.3271783035199076,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004835214268783126,
      "loss": 5.9087,
      "mean_token_accuracy": 0.18213239014148713,
      "num_tokens": 29166289.0,
      "step": 11500
    },
    {
      "entropy": 6.2545263290405275,
      "epoch": 1.3277553375649163,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048350592244734866,
      "loss": 5.8928,
      "mean_token_accuracy": 0.18607201874256135,
      "num_tokens": 29178321.0,
      "step": 11505
    },
    {
      "entropy": 6.250811529159546,
      "epoch": 1.3283323716099251,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004834904110033917,
      "loss": 5.8652,
      "mean_token_accuracy": 0.18639709055423737,
      "num_tokens": 29190396.0,
      "step": 11510
    },
    {
      "entropy": 6.206500768661499,
      "epoch": 1.3289094056549335,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048347489254696327,
      "loss": 5.8893,
      "mean_token_accuracy": 0.187374547123909,
      "num_tokens": 29202698.0,
      "step": 11515
    },
    {
      "entropy": 6.169634389877319,
      "epoch": 1.3294864396999424,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004834593670785854,
      "loss": 5.8066,
      "mean_token_accuracy": 0.19477708637714386,
      "num_tokens": 29214508.0,
      "step": 11520
    },
    {
      "entropy": 6.273674535751343,
      "epoch": 1.3300634737449508,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048344383459878024,
      "loss": 5.8929,
      "mean_token_accuracy": 0.18686662763357162,
      "num_tokens": 29225683.0,
      "step": 11525
    },
    {
      "entropy": 6.165818500518799,
      "epoch": 1.3306405077899597,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048342829510807024,
      "loss": 5.8157,
      "mean_token_accuracy": 0.19610528498888016,
      "num_tokens": 29239497.0,
      "step": 11530
    },
    {
      "entropy": 6.218149137496948,
      "epoch": 1.3312175418349683,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000483412748606978,
      "loss": 5.8765,
      "mean_token_accuracy": 0.1863399013876915,
      "num_tokens": 29252230.0,
      "step": 11535
    },
    {
      "entropy": 6.192487716674805,
      "epoch": 1.331794575879977,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004833971950960266,
      "loss": 5.8456,
      "mean_token_accuracy": 0.18255683928728103,
      "num_tokens": 29265349.0,
      "step": 11540
    },
    {
      "entropy": 6.209053659439087,
      "epoch": 1.3323716099249856,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004833816345757391,
      "loss": 5.8677,
      "mean_token_accuracy": 0.18868259489536285,
      "num_tokens": 29278291.0,
      "step": 11545
    },
    {
      "entropy": 6.255635023117065,
      "epoch": 1.3329486439699942,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004833660670466387,
      "loss": 5.8967,
      "mean_token_accuracy": 0.17723693400621415,
      "num_tokens": 29290580.0,
      "step": 11550
    },
    {
      "entropy": 6.300553560256958,
      "epoch": 1.3335256780150029,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004833504925092492,
      "loss": 6.0274,
      "mean_token_accuracy": 0.17564536333084108,
      "num_tokens": 29305738.0,
      "step": 11555
    },
    {
      "entropy": 6.180785226821899,
      "epoch": 1.3341027120600115,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004833349109640944,
      "loss": 5.8241,
      "mean_token_accuracy": 0.18618278205394745,
      "num_tokens": 29316533.0,
      "step": 11560
    },
    {
      "entropy": 6.167458629608154,
      "epoch": 1.3346797461050202,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004833193224116983,
      "loss": 5.8414,
      "mean_token_accuracy": 0.19107311218976974,
      "num_tokens": 29329723.0,
      "step": 11565
    },
    {
      "entropy": 6.246208715438843,
      "epoch": 1.3352567801500288,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048330372685258526,
      "loss": 5.9083,
      "mean_token_accuracy": 0.18134523630142213,
      "num_tokens": 29342830.0,
      "step": 11570
    },
    {
      "entropy": 6.143269824981689,
      "epoch": 1.3358338141950374,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004832881242872799,
      "loss": 5.767,
      "mean_token_accuracy": 0.192962084710598,
      "num_tokens": 29354096.0,
      "step": 11575
    },
    {
      "entropy": 6.192146921157837,
      "epoch": 1.336410848240046,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004832725147163069,
      "loss": 5.8609,
      "mean_token_accuracy": 0.18677292317152022,
      "num_tokens": 29367173.0,
      "step": 11580
    },
    {
      "entropy": 6.18740553855896,
      "epoch": 1.336987882285055,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048325689814019134,
      "loss": 5.8639,
      "mean_token_accuracy": 0.1850790038704872,
      "num_tokens": 29378958.0,
      "step": 11585
    },
    {
      "entropy": 6.20060453414917,
      "epoch": 1.3375649163300634,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004832412745594585,
      "loss": 5.8433,
      "mean_token_accuracy": 0.19058489948511123,
      "num_tokens": 29391360.0,
      "step": 11590
    },
    {
      "entropy": 6.226307058334351,
      "epoch": 1.3381419503750722,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048322564397463376,
      "loss": 5.9037,
      "mean_token_accuracy": 0.18153176605701446,
      "num_tokens": 29404040.0,
      "step": 11595
    },
    {
      "entropy": 6.270101261138916,
      "epoch": 1.3387189844200809,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048321000638624296,
      "loss": 5.9037,
      "mean_token_accuracy": 0.18023875504732131,
      "num_tokens": 29417960.0,
      "step": 11600
    },
    {
      "entropy": 6.239561080932617,
      "epoch": 1.3392960184650895,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000483194361794812,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1824859157204628,
      "num_tokens": 29431453.0,
      "step": 11605
    },
    {
      "entropy": 6.184908533096314,
      "epoch": 1.3398730525100981,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000483178710200867,
      "loss": 5.9163,
      "mean_token_accuracy": 0.1790502518415451,
      "num_tokens": 29443716.0,
      "step": 11610
    },
    {
      "entropy": 6.185495328903198,
      "epoch": 1.3404500865551068,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004831630516049346,
      "loss": 5.79,
      "mean_token_accuracy": 0.1911078229546547,
      "num_tokens": 29455588.0,
      "step": 11615
    },
    {
      "entropy": 6.138625288009644,
      "epoch": 1.3410271206001154,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004831473860075414,
      "loss": 5.7153,
      "mean_token_accuracy": 0.19553344696760178,
      "num_tokens": 29469472.0,
      "step": 11620
    },
    {
      "entropy": 6.218545722961426,
      "epoch": 1.341604154645124,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048313171340921417,
      "loss": 5.819,
      "mean_token_accuracy": 0.1919667214155197,
      "num_tokens": 29482432.0,
      "step": 11625
    },
    {
      "entropy": 6.220783853530884,
      "epoch": 1.3421811886901327,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048311603381048025,
      "loss": 5.8181,
      "mean_token_accuracy": 0.19385574907064437,
      "num_tokens": 29495425.0,
      "step": 11630
    },
    {
      "entropy": 6.2529582500457765,
      "epoch": 1.3427582227351413,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004831003472118668,
      "loss": 5.8692,
      "mean_token_accuracy": 0.19155482500791549,
      "num_tokens": 29507336.0,
      "step": 11635
    },
    {
      "entropy": 6.156981706619263,
      "epoch": 1.34333525678015,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004830846536139016,
      "loss": 5.7989,
      "mean_token_accuracy": 0.19177321940660477,
      "num_tokens": 29519857.0,
      "step": 11640
    },
    {
      "entropy": 6.2249547958374025,
      "epoch": 1.3439122908251586,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004830689530171124,
      "loss": 5.8469,
      "mean_token_accuracy": 0.18324577808380127,
      "num_tokens": 29531842.0,
      "step": 11645
    },
    {
      "entropy": 6.259550285339356,
      "epoch": 1.3444893248701673,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004830532454220273,
      "loss": 5.8985,
      "mean_token_accuracy": 0.18539201468229294,
      "num_tokens": 29543721.0,
      "step": 11650
    },
    {
      "entropy": 6.258128833770752,
      "epoch": 1.345066358915176,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048303753082917474,
      "loss": 5.8919,
      "mean_token_accuracy": 0.18458596765995025,
      "num_tokens": 29555713.0,
      "step": 11655
    },
    {
      "entropy": 6.19729266166687,
      "epoch": 1.3456433929601848,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048302180923908306,
      "loss": 5.8954,
      "mean_token_accuracy": 0.18397297859191894,
      "num_tokens": 29568983.0,
      "step": 11660
    },
    {
      "entropy": 6.282765340805054,
      "epoch": 1.3462204270051932,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048300608065228126,
      "loss": 5.9034,
      "mean_token_accuracy": 0.18377083241939546,
      "num_tokens": 29580778.0,
      "step": 11665
    },
    {
      "entropy": 6.276874732971192,
      "epoch": 1.346797461050202,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048299034506929815,
      "loss": 5.8994,
      "mean_token_accuracy": 0.1784396469593048,
      "num_tokens": 29594395.0,
      "step": 11670
    },
    {
      "entropy": 6.1028321266174315,
      "epoch": 1.3473744950952107,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048297460249066315,
      "loss": 5.7698,
      "mean_token_accuracy": 0.19325246214866637,
      "num_tokens": 29607588.0,
      "step": 11675
    },
    {
      "entropy": 6.24799370765686,
      "epoch": 1.3479515291402193,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004829588529169057,
      "loss": 5.8453,
      "mean_token_accuracy": 0.18600047379732132,
      "num_tokens": 29618822.0,
      "step": 11680
    },
    {
      "entropy": 6.122512245178223,
      "epoch": 1.348528563185228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004829430963485555,
      "loss": 5.8154,
      "mean_token_accuracy": 0.18972329944372177,
      "num_tokens": 29631759.0,
      "step": 11685
    },
    {
      "entropy": 6.250639152526856,
      "epoch": 1.3491055972302366,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004829273327861426,
      "loss": 5.8657,
      "mean_token_accuracy": 0.1783287152647972,
      "num_tokens": 29644381.0,
      "step": 11690
    },
    {
      "entropy": 6.230941534042358,
      "epoch": 1.3496826312752452,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004829115622301971,
      "loss": 5.8787,
      "mean_token_accuracy": 0.1829552635550499,
      "num_tokens": 29657370.0,
      "step": 11695
    },
    {
      "entropy": 6.261565685272217,
      "epoch": 1.3502596653202539,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048289578468124956,
      "loss": 5.9253,
      "mean_token_accuracy": 0.1811118647456169,
      "num_tokens": 29669438.0,
      "step": 11700
    },
    {
      "entropy": 6.252146434783936,
      "epoch": 1.3508366993652625,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048288000013983046,
      "loss": 5.9058,
      "mean_token_accuracy": 0.18210960030555726,
      "num_tokens": 29681277.0,
      "step": 11705
    },
    {
      "entropy": 6.161410093307495,
      "epoch": 1.3514137334102712,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048286420860647086,
      "loss": 5.7474,
      "mean_token_accuracy": 0.18817867189645768,
      "num_tokens": 29693523.0,
      "step": 11710
    },
    {
      "entropy": 6.3111645698547365,
      "epoch": 1.3519907674552798,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048284841008170185,
      "loss": 5.9414,
      "mean_token_accuracy": 0.17817626297473907,
      "num_tokens": 29707589.0,
      "step": 11715
    },
    {
      "entropy": 6.261362600326538,
      "epoch": 1.3525678015002884,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048283260456605487,
      "loss": 5.9501,
      "mean_token_accuracy": 0.17884300798177719,
      "num_tokens": 29719267.0,
      "step": 11720
    },
    {
      "entropy": 6.152270889282226,
      "epoch": 1.3531448355452973,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004828167920600614,
      "loss": 5.7494,
      "mean_token_accuracy": 0.1960026890039444,
      "num_tokens": 29731575.0,
      "step": 11725
    },
    {
      "entropy": 6.235330820083618,
      "epoch": 1.3537218695903057,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004828009725642534,
      "loss": 5.9221,
      "mean_token_accuracy": 0.17483729869127274,
      "num_tokens": 29744762.0,
      "step": 11730
    },
    {
      "entropy": 6.267324018478393,
      "epoch": 1.3542989036353146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004827851460791628,
      "loss": 5.9516,
      "mean_token_accuracy": 0.18018896877765656,
      "num_tokens": 29758300.0,
      "step": 11735
    },
    {
      "entropy": 6.239387941360474,
      "epoch": 1.3548759376803232,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048276931260532213,
      "loss": 5.8519,
      "mean_token_accuracy": 0.18261149078607558,
      "num_tokens": 29771489.0,
      "step": 11740
    },
    {
      "entropy": 6.258696508407593,
      "epoch": 1.3554529717253319,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004827534721432639,
      "loss": 5.9362,
      "mean_token_accuracy": 0.18618210703134536,
      "num_tokens": 29783063.0,
      "step": 11745
    },
    {
      "entropy": 6.273253011703491,
      "epoch": 1.3560300057703405,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004827376246935207,
      "loss": 5.9141,
      "mean_token_accuracy": 0.1834575667977333,
      "num_tokens": 29797136.0,
      "step": 11750
    },
    {
      "entropy": 6.223002815246582,
      "epoch": 1.3566070398153491,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004827217702566257,
      "loss": 5.8635,
      "mean_token_accuracy": 0.18649065494537354,
      "num_tokens": 29809580.0,
      "step": 11755
    },
    {
      "entropy": 6.1100945472717285,
      "epoch": 1.3571840738603578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048270590883311217,
      "loss": 5.7647,
      "mean_token_accuracy": 0.19088437408208847,
      "num_tokens": 29822400.0,
      "step": 11760
    },
    {
      "entropy": 6.2468303680419925,
      "epoch": 1.3577611079053664,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048269004042351363,
      "loss": 5.8651,
      "mean_token_accuracy": 0.18244728446006775,
      "num_tokens": 29834392.0,
      "step": 11765
    },
    {
      "entropy": 6.2506287574768065,
      "epoch": 1.358338141950375,
      "grad_norm": 1.0,
      "learning_rate": 0.0004826741650283637,
      "loss": 5.8709,
      "mean_token_accuracy": 0.18384577184915543,
      "num_tokens": 29846626.0,
      "step": 11770
    },
    {
      "entropy": 6.203633117675781,
      "epoch": 1.3589151759953837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004826582826481963,
      "loss": 5.826,
      "mean_token_accuracy": 0.1876814991235733,
      "num_tokens": 29858988.0,
      "step": 11775
    },
    {
      "entropy": 6.154868125915527,
      "epoch": 1.3594922100403923,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004826423932835458,
      "loss": 5.8045,
      "mean_token_accuracy": 0.1907202571630478,
      "num_tokens": 29871875.0,
      "step": 11780
    },
    {
      "entropy": 6.228389644622803,
      "epoch": 1.360069244085401,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048262649693494653,
      "loss": 5.8503,
      "mean_token_accuracy": 0.18580542951822282,
      "num_tokens": 29884757.0,
      "step": 11785
    },
    {
      "entropy": 6.195999526977539,
      "epoch": 1.3606462781304096,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004826105936029332,
      "loss": 5.8392,
      "mean_token_accuracy": 0.18981667160987853,
      "num_tokens": 29897309.0,
      "step": 11790
    },
    {
      "entropy": 6.284321260452271,
      "epoch": 1.3612233121754183,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825946832880406,
      "loss": 5.9247,
      "mean_token_accuracy": 0.17598619312047958,
      "num_tokens": 29910005.0,
      "step": 11795
    },
    {
      "entropy": 6.22040205001831,
      "epoch": 1.3618003462204271,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048257876599080404,
      "loss": 5.8878,
      "mean_token_accuracy": 0.18229834586381913,
      "num_tokens": 29921912.0,
      "step": 11800
    },
    {
      "entropy": 6.251012372970581,
      "epoch": 1.3623773802654355,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048256284171175874,
      "loss": 5.8468,
      "mean_token_accuracy": 0.18459426164627074,
      "num_tokens": 29934399.0,
      "step": 11805
    },
    {
      "entropy": 6.228803539276123,
      "epoch": 1.3629544143104444,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048254691045144035,
      "loss": 5.8198,
      "mean_token_accuracy": 0.18516346663236619,
      "num_tokens": 29946660.0,
      "step": 11810
    },
    {
      "entropy": 6.212866544723511,
      "epoch": 1.363531448355453,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004825309722103848,
      "loss": 5.845,
      "mean_token_accuracy": 0.1838705688714981,
      "num_tokens": 29958382.0,
      "step": 11815
    },
    {
      "entropy": 6.2703948497772215,
      "epoch": 1.3641084824004617,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000482515026989128,
      "loss": 5.8321,
      "mean_token_accuracy": 0.18664811700582504,
      "num_tokens": 29971241.0,
      "step": 11820
    },
    {
      "entropy": 6.303622579574585,
      "epoch": 1.3646855164454703,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048249907478820634,
      "loss": 5.9203,
      "mean_token_accuracy": 0.18046647310256958,
      "num_tokens": 29984145.0,
      "step": 11825
    },
    {
      "entropy": 6.17892746925354,
      "epoch": 1.365262550490479,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048248311560815637,
      "loss": 5.8125,
      "mean_token_accuracy": 0.18881973773241043,
      "num_tokens": 29997092.0,
      "step": 11830
    },
    {
      "entropy": 6.1425611019134525,
      "epoch": 1.3658395845354876,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004824671494495149,
      "loss": 5.8448,
      "mean_token_accuracy": 0.19086938947439194,
      "num_tokens": 30009369.0,
      "step": 11835
    },
    {
      "entropy": 6.222787857055664,
      "epoch": 1.3664166185804962,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004824511763128189,
      "loss": 5.8248,
      "mean_token_accuracy": 0.18760768324136734,
      "num_tokens": 30021423.0,
      "step": 11840
    },
    {
      "entropy": 6.241295337677002,
      "epoch": 1.3669936526255049,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048243519619860567,
      "loss": 5.8881,
      "mean_token_accuracy": 0.18700434565544127,
      "num_tokens": 30034426.0,
      "step": 11845
    },
    {
      "entropy": 6.197099113464356,
      "epoch": 1.3675706866705135,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004824192091074126,
      "loss": 5.859,
      "mean_token_accuracy": 0.18185582160949706,
      "num_tokens": 30046687.0,
      "step": 11850
    },
    {
      "entropy": 6.2657853126525875,
      "epoch": 1.3681477207155222,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004824032150397775,
      "loss": 5.9146,
      "mean_token_accuracy": 0.17807950228452682,
      "num_tokens": 30058815.0,
      "step": 11855
    },
    {
      "entropy": 6.3250209331512455,
      "epoch": 1.3687247547605308,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048238721399623824,
      "loss": 5.923,
      "mean_token_accuracy": 0.17989212423563003,
      "num_tokens": 30070715.0,
      "step": 11860
    },
    {
      "entropy": 6.244921112060547,
      "epoch": 1.3693017888055397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048237120597733316,
      "loss": 5.7811,
      "mean_token_accuracy": 0.19141092002391816,
      "num_tokens": 30083636.0,
      "step": 11865
    },
    {
      "entropy": 6.23410005569458,
      "epoch": 1.369878822850548,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004823551909836005,
      "loss": 5.9107,
      "mean_token_accuracy": 0.18153376430273055,
      "num_tokens": 30094956.0,
      "step": 11870
    },
    {
      "entropy": 6.22312741279602,
      "epoch": 1.370455856895557,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048233916901557896,
      "loss": 5.8207,
      "mean_token_accuracy": 0.1860801324248314,
      "num_tokens": 30107947.0,
      "step": 11875
    },
    {
      "entropy": 6.237414026260376,
      "epoch": 1.3710328909405656,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048232314007380753,
      "loss": 5.9293,
      "mean_token_accuracy": 0.1775343969464302,
      "num_tokens": 30120557.0,
      "step": 11880
    },
    {
      "entropy": 6.284224987030029,
      "epoch": 1.3716099249855742,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048230710415882524,
      "loss": 5.93,
      "mean_token_accuracy": 0.1874416932463646,
      "num_tokens": 30131734.0,
      "step": 11885
    },
    {
      "entropy": 6.251478290557861,
      "epoch": 1.3721869590305829,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048229106127117144,
      "loss": 5.8175,
      "mean_token_accuracy": 0.18836882412433625,
      "num_tokens": 30144025.0,
      "step": 11890
    },
    {
      "entropy": 6.2390485286712645,
      "epoch": 1.3727639930755915,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004822750114113858,
      "loss": 5.8678,
      "mean_token_accuracy": 0.18512837290763856,
      "num_tokens": 30156246.0,
      "step": 11895
    },
    {
      "entropy": 6.203308248519898,
      "epoch": 1.3733410271206001,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004822589545800081,
      "loss": 5.8418,
      "mean_token_accuracy": 0.18874067962169647,
      "num_tokens": 30169349.0,
      "step": 11900
    },
    {
      "entropy": 6.240908622741699,
      "epoch": 1.3739180611656088,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004822428907775784,
      "loss": 5.8662,
      "mean_token_accuracy": 0.18322746753692626,
      "num_tokens": 30181821.0,
      "step": 11905
    },
    {
      "entropy": 6.248536014556885,
      "epoch": 1.3744950952106174,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048222682000463704,
      "loss": 5.8428,
      "mean_token_accuracy": 0.18755651861429215,
      "num_tokens": 30194278.0,
      "step": 11910
    },
    {
      "entropy": 6.209737348556518,
      "epoch": 1.375072129255626,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004822107422617245,
      "loss": 5.8785,
      "mean_token_accuracy": 0.185113425552845,
      "num_tokens": 30206706.0,
      "step": 11915
    },
    {
      "entropy": 6.194297361373901,
      "epoch": 1.3756491633006347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048219465754938156,
      "loss": 5.8521,
      "mean_token_accuracy": 0.18779730647802353,
      "num_tokens": 30220356.0,
      "step": 11920
    },
    {
      "entropy": 6.26024112701416,
      "epoch": 1.3762261973456433,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048217856586814926,
      "loss": 5.8858,
      "mean_token_accuracy": 0.18770142048597335,
      "num_tokens": 30232821.0,
      "step": 11925
    },
    {
      "entropy": 6.088385343551636,
      "epoch": 1.376803231390652,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048216246721856875,
      "loss": 5.7992,
      "mean_token_accuracy": 0.18888443261384963,
      "num_tokens": 30246094.0,
      "step": 11930
    },
    {
      "entropy": 6.250760412216186,
      "epoch": 1.3773802654356606,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048214636160118164,
      "loss": 5.8677,
      "mean_token_accuracy": 0.189895099401474,
      "num_tokens": 30259313.0,
      "step": 11935
    },
    {
      "entropy": 6.3054070472717285,
      "epoch": 1.3779572994806695,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004821302490165295,
      "loss": 5.9186,
      "mean_token_accuracy": 0.18048689216375352,
      "num_tokens": 30274045.0,
      "step": 11940
    },
    {
      "entropy": 6.209653759002686,
      "epoch": 1.378534333525678,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004821141294651544,
      "loss": 5.8523,
      "mean_token_accuracy": 0.18880870640277864,
      "num_tokens": 30286250.0,
      "step": 11945
    },
    {
      "entropy": 6.226180458068848,
      "epoch": 1.3791113675706868,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048209800294759836,
      "loss": 5.8959,
      "mean_token_accuracy": 0.18101679980754853,
      "num_tokens": 30299417.0,
      "step": 11950
    },
    {
      "entropy": 6.236491250991821,
      "epoch": 1.3796884016156954,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004820818694644039,
      "loss": 5.8486,
      "mean_token_accuracy": 0.18206991553306578,
      "num_tokens": 30311465.0,
      "step": 11955
    },
    {
      "entropy": 6.1253071308135985,
      "epoch": 1.380265435660704,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048206572901611364,
      "loss": 5.6994,
      "mean_token_accuracy": 0.196821990609169,
      "num_tokens": 30325941.0,
      "step": 11960
    },
    {
      "entropy": 6.267721271514892,
      "epoch": 1.3808424697057127,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048204958160327034,
      "loss": 6.0067,
      "mean_token_accuracy": 0.17587004899978637,
      "num_tokens": 30339238.0,
      "step": 11965
    },
    {
      "entropy": 6.245282888412476,
      "epoch": 1.3814195037507213,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048203342722641726,
      "loss": 5.8001,
      "mean_token_accuracy": 0.1954345002770424,
      "num_tokens": 30351082.0,
      "step": 11970
    },
    {
      "entropy": 6.21645884513855,
      "epoch": 1.38199653779573,
      "grad_norm": 0.875,
      "learning_rate": 0.00048201726588609776,
      "loss": 5.8453,
      "mean_token_accuracy": 0.18525812923908233,
      "num_tokens": 30364472.0,
      "step": 11975
    },
    {
      "entropy": 6.177454423904419,
      "epoch": 1.3825735718407386,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048200109758285534,
      "loss": 5.8008,
      "mean_token_accuracy": 0.1869779869914055,
      "num_tokens": 30377113.0,
      "step": 11980
    },
    {
      "entropy": 6.275010967254639,
      "epoch": 1.3831506058857472,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004819849223172337,
      "loss": 5.897,
      "mean_token_accuracy": 0.17949435412883757,
      "num_tokens": 30389313.0,
      "step": 11985
    },
    {
      "entropy": 6.26848292350769,
      "epoch": 1.3837276399307559,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048196874008977716,
      "loss": 5.9753,
      "mean_token_accuracy": 0.17302415668964385,
      "num_tokens": 30402227.0,
      "step": 11990
    },
    {
      "entropy": 6.261510372161865,
      "epoch": 1.3843046739757645,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004819525509010298,
      "loss": 5.8681,
      "mean_token_accuracy": 0.18413420021533966,
      "num_tokens": 30414719.0,
      "step": 11995
    },
    {
      "entropy": 6.225330781936646,
      "epoch": 1.3848817080207732,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004819363547515361,
      "loss": 5.9032,
      "mean_token_accuracy": 0.18110304623842238,
      "num_tokens": 30427989.0,
      "step": 12000
    },
    {
      "epoch": 1.3848817080207732,
      "eval_entropy": 6.099995839585626,
      "eval_loss": 5.940317630767822,
      "eval_mean_token_accuracy": 0.1878560801328737,
      "eval_num_tokens": 30427989.0,
      "eval_runtime": 17.6697,
      "eval_samples_per_second": 1592.326,
      "eval_steps_per_second": 199.041,
      "step": 12000
    },
    {
      "entropy": 6.25681357383728,
      "epoch": 1.385458742065782,
      "grad_norm": 0.828125,
      "learning_rate": 0.000481920151641841,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1821727052330971,
      "num_tokens": 30441097.0,
      "step": 12005
    },
    {
      "entropy": 6.104511880874634,
      "epoch": 1.3860357761107904,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048190394157248935,
      "loss": 5.7095,
      "mean_token_accuracy": 0.1949584513902664,
      "num_tokens": 30453535.0,
      "step": 12010
    },
    {
      "entropy": 6.197775077819824,
      "epoch": 1.3866128101557993,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004818877245440264,
      "loss": 5.9067,
      "mean_token_accuracy": 0.18245501220226287,
      "num_tokens": 30466350.0,
      "step": 12015
    },
    {
      "entropy": 6.310269975662232,
      "epoch": 1.3871898442008077,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048187150055699763,
      "loss": 5.9295,
      "mean_token_accuracy": 0.17709443867206573,
      "num_tokens": 30479183.0,
      "step": 12020
    },
    {
      "entropy": 6.264182233810425,
      "epoch": 1.3877668782458166,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004818552696119487,
      "loss": 5.8241,
      "mean_token_accuracy": 0.19044758081436158,
      "num_tokens": 30491903.0,
      "step": 12025
    },
    {
      "entropy": 6.201251029968262,
      "epoch": 1.3883439122908252,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818390317094255,
      "loss": 5.8259,
      "mean_token_accuracy": 0.18758193552494049,
      "num_tokens": 30504424.0,
      "step": 12030
    },
    {
      "entropy": 6.223074054718017,
      "epoch": 1.3889209463358339,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004818227868499742,
      "loss": 5.8386,
      "mean_token_accuracy": 0.18516270369291304,
      "num_tokens": 30517339.0,
      "step": 12035
    },
    {
      "entropy": 6.232985639572144,
      "epoch": 1.3894979803808425,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004818065350341412,
      "loss": 5.8273,
      "mean_token_accuracy": 0.18675389289855956,
      "num_tokens": 30529218.0,
      "step": 12040
    },
    {
      "entropy": 6.254549884796143,
      "epoch": 1.3900750144258511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048179027626247325,
      "loss": 5.7787,
      "mean_token_accuracy": 0.19106095433235168,
      "num_tokens": 30540601.0,
      "step": 12045
    },
    {
      "entropy": 6.089814472198486,
      "epoch": 1.3906520484708598,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004817740105355169,
      "loss": 5.7445,
      "mean_token_accuracy": 0.194708052277565,
      "num_tokens": 30552834.0,
      "step": 12050
    },
    {
      "entropy": 6.195273160934448,
      "epoch": 1.3912290825158684,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048175773785381947,
      "loss": 5.873,
      "mean_token_accuracy": 0.19092209190130233,
      "num_tokens": 30563799.0,
      "step": 12055
    },
    {
      "entropy": 6.235287094116211,
      "epoch": 1.391806116560877,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048174145821792833,
      "loss": 5.8305,
      "mean_token_accuracy": 0.19392269551754,
      "num_tokens": 30576637.0,
      "step": 12060
    },
    {
      "entropy": 6.2308636665344235,
      "epoch": 1.3923831506058857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004817251716283908,
      "loss": 5.9381,
      "mean_token_accuracy": 0.18031953871250153,
      "num_tokens": 30589720.0,
      "step": 12065
    },
    {
      "entropy": 6.242139720916748,
      "epoch": 1.3929601846508943,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004817088780857548,
      "loss": 5.939,
      "mean_token_accuracy": 0.18436544090509416,
      "num_tokens": 30602895.0,
      "step": 12070
    },
    {
      "entropy": 6.153406286239624,
      "epoch": 1.393537218695903,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048169257759056845,
      "loss": 5.8205,
      "mean_token_accuracy": 0.19048167318105697,
      "num_tokens": 30615411.0,
      "step": 12075
    },
    {
      "entropy": 6.233449935913086,
      "epoch": 1.3941142527409118,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048167627014337994,
      "loss": 5.8519,
      "mean_token_accuracy": 0.18667246848344804,
      "num_tokens": 30627682.0,
      "step": 12080
    },
    {
      "entropy": 6.27116641998291,
      "epoch": 1.3946912867859202,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048165995574473764,
      "loss": 5.9909,
      "mean_token_accuracy": 0.17633870542049407,
      "num_tokens": 30641723.0,
      "step": 12085
    },
    {
      "entropy": 6.261033821105957,
      "epoch": 1.395268320830929,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048164363439519043,
      "loss": 5.9265,
      "mean_token_accuracy": 0.17952186167240142,
      "num_tokens": 30654926.0,
      "step": 12090
    },
    {
      "entropy": 6.25032229423523,
      "epoch": 1.3958453548759377,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004816273060952873,
      "loss": 5.8898,
      "mean_token_accuracy": 0.18118651062250138,
      "num_tokens": 30667641.0,
      "step": 12095
    },
    {
      "entropy": 6.265254402160645,
      "epoch": 1.3964223889209464,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048161097084557726,
      "loss": 5.8878,
      "mean_token_accuracy": 0.18833872377872468,
      "num_tokens": 30680841.0,
      "step": 12100
    },
    {
      "entropy": 6.258693695068359,
      "epoch": 1.396999422965955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048159462864660993,
      "loss": 5.8351,
      "mean_token_accuracy": 0.19087911397218704,
      "num_tokens": 30695829.0,
      "step": 12105
    },
    {
      "entropy": 6.265958213806153,
      "epoch": 1.3975764570109637,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004815782794989348,
      "loss": 5.8061,
      "mean_token_accuracy": 0.18091001808643342,
      "num_tokens": 30707853.0,
      "step": 12110
    },
    {
      "entropy": 6.1866248607635494,
      "epoch": 1.3981534910559723,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004815619234031019,
      "loss": 5.7881,
      "mean_token_accuracy": 0.18840817213058472,
      "num_tokens": 30720568.0,
      "step": 12115
    },
    {
      "entropy": 6.119961404800415,
      "epoch": 1.398730525100981,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004815455603596613,
      "loss": 5.8416,
      "mean_token_accuracy": 0.18913987129926682,
      "num_tokens": 30734282.0,
      "step": 12120
    },
    {
      "entropy": 6.227114629745484,
      "epoch": 1.3993075591459896,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004815291903691634,
      "loss": 5.826,
      "mean_token_accuracy": 0.18959399610757827,
      "num_tokens": 30746186.0,
      "step": 12125
    },
    {
      "entropy": 6.212065601348877,
      "epoch": 1.3998845931909982,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048151281343215873,
      "loss": 5.8853,
      "mean_token_accuracy": 0.18449723869562148,
      "num_tokens": 30757697.0,
      "step": 12130
    },
    {
      "entropy": 6.22585277557373,
      "epoch": 1.4004616272360069,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004814964295491982,
      "loss": 5.844,
      "mean_token_accuracy": 0.1880333364009857,
      "num_tokens": 30770068.0,
      "step": 12135
    },
    {
      "entropy": 6.225053071975708,
      "epoch": 1.4010386612810155,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048148003872083286,
      "loss": 5.7963,
      "mean_token_accuracy": 0.18897532820701599,
      "num_tokens": 30782135.0,
      "step": 12140
    },
    {
      "entropy": 6.235483264923095,
      "epoch": 1.4016156953260244,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048146364094761384,
      "loss": 5.9007,
      "mean_token_accuracy": 0.18461197018623351,
      "num_tokens": 30793448.0,
      "step": 12145
    },
    {
      "entropy": 6.222435474395752,
      "epoch": 1.4021927293710328,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048144723623009297,
      "loss": 5.8254,
      "mean_token_accuracy": 0.18379292041063308,
      "num_tokens": 30806280.0,
      "step": 12150
    },
    {
      "entropy": 6.171455049514771,
      "epoch": 1.4027697634160416,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004814308245688218,
      "loss": 5.8305,
      "mean_token_accuracy": 0.18788353949785233,
      "num_tokens": 30820394.0,
      "step": 12155
    },
    {
      "entropy": 6.238278007507324,
      "epoch": 1.40334679746105,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048141440596435235,
      "loss": 5.8465,
      "mean_token_accuracy": 0.18920082449913025,
      "num_tokens": 30832343.0,
      "step": 12160
    },
    {
      "entropy": 6.246117830276489,
      "epoch": 1.403923831506059,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004813979804172369,
      "loss": 5.871,
      "mean_token_accuracy": 0.18447341322898864,
      "num_tokens": 30844833.0,
      "step": 12165
    },
    {
      "entropy": 6.164185667037964,
      "epoch": 1.4045008655510676,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048138154792802795,
      "loss": 5.7773,
      "mean_token_accuracy": 0.19125905632972717,
      "num_tokens": 30856899.0,
      "step": 12170
    },
    {
      "entropy": 6.207029819488525,
      "epoch": 1.4050778995960762,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004813651084972781,
      "loss": 5.8498,
      "mean_token_accuracy": 0.19137704819440843,
      "num_tokens": 30869172.0,
      "step": 12175
    },
    {
      "entropy": 6.247862768173218,
      "epoch": 1.4056549336410848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048134866212554036,
      "loss": 5.8635,
      "mean_token_accuracy": 0.18843065351247787,
      "num_tokens": 30880751.0,
      "step": 12180
    },
    {
      "entropy": 6.208236360549927,
      "epoch": 1.4062319676860935,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004813322088133679,
      "loss": 5.8379,
      "mean_token_accuracy": 0.1796042203903198,
      "num_tokens": 30893583.0,
      "step": 12185
    },
    {
      "entropy": 6.266119956970215,
      "epoch": 1.4068090017311021,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048131574856131407,
      "loss": 5.9219,
      "mean_token_accuracy": 0.182425419986248,
      "num_tokens": 30906150.0,
      "step": 12190
    },
    {
      "entropy": 6.222499752044678,
      "epoch": 1.4073860357761108,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004812992813699324,
      "loss": 5.8585,
      "mean_token_accuracy": 0.1822350800037384,
      "num_tokens": 30918505.0,
      "step": 12195
    },
    {
      "entropy": 6.202458715438842,
      "epoch": 1.4079630698211194,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000481282807239777,
      "loss": 5.8783,
      "mean_token_accuracy": 0.18459342718124389,
      "num_tokens": 30932519.0,
      "step": 12200
    },
    {
      "entropy": 6.214414358139038,
      "epoch": 1.408540103866128,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004812663261714019,
      "loss": 5.8932,
      "mean_token_accuracy": 0.18110774755477904,
      "num_tokens": 30944482.0,
      "step": 12205
    },
    {
      "entropy": 6.23360424041748,
      "epoch": 1.4091171379111367,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004812498381653613,
      "loss": 5.808,
      "mean_token_accuracy": 0.19237677454948426,
      "num_tokens": 30956726.0,
      "step": 12210
    },
    {
      "entropy": 6.183458662033081,
      "epoch": 1.4096941719561453,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004812333432222098,
      "loss": 5.8185,
      "mean_token_accuracy": 0.1868341565132141,
      "num_tokens": 30968557.0,
      "step": 12215
    },
    {
      "entropy": 6.240329027175903,
      "epoch": 1.4102712060011542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004812168413425023,
      "loss": 5.8649,
      "mean_token_accuracy": 0.19138860553503037,
      "num_tokens": 30981442.0,
      "step": 12220
    },
    {
      "entropy": 6.180831003189087,
      "epoch": 1.4108482400461626,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048120033252679374,
      "loss": 5.7928,
      "mean_token_accuracy": 0.18167011439800262,
      "num_tokens": 30994126.0,
      "step": 12225
    },
    {
      "entropy": 6.263823986053467,
      "epoch": 1.4114252740911715,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048118381677563946,
      "loss": 5.8492,
      "mean_token_accuracy": 0.19315034300088882,
      "num_tokens": 31006688.0,
      "step": 12230
    },
    {
      "entropy": 6.295634746551514,
      "epoch": 1.41200230813618,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004811672940895949,
      "loss": 5.9607,
      "mean_token_accuracy": 0.174312986433506,
      "num_tokens": 31019922.0,
      "step": 12235
    },
    {
      "entropy": 6.296223926544189,
      "epoch": 1.4125793421811887,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004811507644692158,
      "loss": 5.9473,
      "mean_token_accuracy": 0.17719331830739976,
      "num_tokens": 31032068.0,
      "step": 12240
    },
    {
      "entropy": 6.2304267406463625,
      "epoch": 1.4131563762261974,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004811342279150581,
      "loss": 5.8022,
      "mean_token_accuracy": 0.18696216493844986,
      "num_tokens": 31045547.0,
      "step": 12245
    },
    {
      "entropy": 6.223422193527222,
      "epoch": 1.413733410271206,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004811176844276781,
      "loss": 5.8754,
      "mean_token_accuracy": 0.182257342338562,
      "num_tokens": 31059219.0,
      "step": 12250
    },
    {
      "entropy": 6.240674066543579,
      "epoch": 1.4143104443162147,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004811011340076322,
      "loss": 5.9068,
      "mean_token_accuracy": 0.18581583946943284,
      "num_tokens": 31072472.0,
      "step": 12255
    },
    {
      "entropy": 6.283777141571045,
      "epoch": 1.4148874783612233,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048108457665547695,
      "loss": 5.9337,
      "mean_token_accuracy": 0.1770991414785385,
      "num_tokens": 31085394.0,
      "step": 12260
    },
    {
      "entropy": 6.291441917419434,
      "epoch": 1.415464512406232,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004810680123717694,
      "loss": 5.9142,
      "mean_token_accuracy": 0.1822252869606018,
      "num_tokens": 31098651.0,
      "step": 12265
    },
    {
      "entropy": 6.2274573802947994,
      "epoch": 1.4160415464512406,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004810514411570666,
      "loss": 5.891,
      "mean_token_accuracy": 0.18512693494558335,
      "num_tokens": 31111297.0,
      "step": 12270
    },
    {
      "entropy": 6.174374294281006,
      "epoch": 1.4166185804962492,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004810348630119259,
      "loss": 5.8648,
      "mean_token_accuracy": 0.1855068176984787,
      "num_tokens": 31124247.0,
      "step": 12275
    },
    {
      "entropy": 6.223607206344605,
      "epoch": 1.4171956145412579,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048101827793690493,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18701430261135102,
      "num_tokens": 31136106.0,
      "step": 12280
    },
    {
      "entropy": 6.18486213684082,
      "epoch": 1.4177726485862667,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004810016859325615,
      "loss": 5.7435,
      "mean_token_accuracy": 0.19915330708026885,
      "num_tokens": 31148561.0,
      "step": 12285
    },
    {
      "entropy": 6.344252252578736,
      "epoch": 1.4183496826312751,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004809850869994537,
      "loss": 5.87,
      "mean_token_accuracy": 0.1795135736465454,
      "num_tokens": 31162289.0,
      "step": 12290
    },
    {
      "entropy": 6.246892976760864,
      "epoch": 1.418926716676284,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004809684811381398,
      "loss": 5.8392,
      "mean_token_accuracy": 0.1845525473356247,
      "num_tokens": 31174013.0,
      "step": 12295
    },
    {
      "entropy": 6.114938592910766,
      "epoch": 1.4195037507212924,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004809518683491785,
      "loss": 5.7611,
      "mean_token_accuracy": 0.19683301001787185,
      "num_tokens": 31186299.0,
      "step": 12300
    },
    {
      "entropy": 6.251533699035645,
      "epoch": 1.4200807847663013,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048093524863312823,
      "loss": 5.8846,
      "mean_token_accuracy": 0.18240513652563095,
      "num_tokens": 31198458.0,
      "step": 12305
    },
    {
      "entropy": 6.2563111782073975,
      "epoch": 1.42065781881131,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004809186219905482,
      "loss": 5.9158,
      "mean_token_accuracy": 0.18031973838806153,
      "num_tokens": 31210712.0,
      "step": 12310
    },
    {
      "entropy": 6.192594814300537,
      "epoch": 1.4212348528563186,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004809019884219976,
      "loss": 5.8243,
      "mean_token_accuracy": 0.18846655786037445,
      "num_tokens": 31223460.0,
      "step": 12315
    },
    {
      "entropy": 6.208048105239868,
      "epoch": 1.4218118869013272,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048088534792803595,
      "loss": 5.8498,
      "mean_token_accuracy": 0.18974225372076034,
      "num_tokens": 31235298.0,
      "step": 12320
    },
    {
      "entropy": 6.267314434051514,
      "epoch": 1.4223889209463358,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048086870050922286,
      "loss": 5.9004,
      "mean_token_accuracy": 0.17768537402153015,
      "num_tokens": 31247164.0,
      "step": 12325
    },
    {
      "entropy": 6.277236700057983,
      "epoch": 1.4229659549913445,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048085204616611833,
      "loss": 5.8657,
      "mean_token_accuracy": 0.19184055477380751,
      "num_tokens": 31259963.0,
      "step": 12330
    },
    {
      "entropy": 6.250907707214355,
      "epoch": 1.4235429890363531,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048083538489928246,
      "loss": 5.8793,
      "mean_token_accuracy": 0.18113293647766113,
      "num_tokens": 31272646.0,
      "step": 12335
    },
    {
      "entropy": 6.180868768692017,
      "epoch": 1.4241200230813618,
      "grad_norm": 0.875,
      "learning_rate": 0.0004808187167092757,
      "loss": 5.7965,
      "mean_token_accuracy": 0.19398818910121918,
      "num_tokens": 31286474.0,
      "step": 12340
    },
    {
      "entropy": 6.158138370513916,
      "epoch": 1.4246970571263704,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004808020415966586,
      "loss": 5.7427,
      "mean_token_accuracy": 0.193064646422863,
      "num_tokens": 31300086.0,
      "step": 12345
    },
    {
      "entropy": 6.183599472045898,
      "epoch": 1.425274091171379,
      "grad_norm": 0.921875,
      "learning_rate": 0.000480785359561992,
      "loss": 5.7547,
      "mean_token_accuracy": 0.19418373107910156,
      "num_tokens": 31312562.0,
      "step": 12350
    },
    {
      "entropy": 6.231119012832641,
      "epoch": 1.4258511252163877,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048076867060583704,
      "loss": 5.9008,
      "mean_token_accuracy": 0.18114117980003358,
      "num_tokens": 31325200.0,
      "step": 12355
    },
    {
      "entropy": 6.226961946487426,
      "epoch": 1.4264281592613965,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004807519747287551,
      "loss": 5.8358,
      "mean_token_accuracy": 0.18880437314510345,
      "num_tokens": 31337361.0,
      "step": 12360
    },
    {
      "entropy": 6.221896171569824,
      "epoch": 1.427005193306405,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004807352719313078,
      "loss": 5.8151,
      "mean_token_accuracy": 0.1878543585538864,
      "num_tokens": 31351411.0,
      "step": 12365
    },
    {
      "entropy": 6.1694423198699955,
      "epoch": 1.4275822273514138,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004807185622140567,
      "loss": 5.8229,
      "mean_token_accuracy": 0.18979541957378387,
      "num_tokens": 31364616.0,
      "step": 12370
    },
    {
      "entropy": 6.18802695274353,
      "epoch": 1.4281592613964225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048070184557756396,
      "loss": 5.8069,
      "mean_token_accuracy": 0.19088694751262664,
      "num_tokens": 31376515.0,
      "step": 12375
    },
    {
      "entropy": 6.253544616699219,
      "epoch": 1.428736295441431,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048068512202239177,
      "loss": 5.8586,
      "mean_token_accuracy": 0.18943951427936553,
      "num_tokens": 31388408.0,
      "step": 12380
    },
    {
      "entropy": 6.186633634567261,
      "epoch": 1.4293133294864397,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004806683915491028,
      "loss": 5.8307,
      "mean_token_accuracy": 0.18826691806316376,
      "num_tokens": 31401001.0,
      "step": 12385
    },
    {
      "entropy": 6.222378826141357,
      "epoch": 1.4298903635314484,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004806516541582596,
      "loss": 5.7706,
      "mean_token_accuracy": 0.18555983901023865,
      "num_tokens": 31413259.0,
      "step": 12390
    },
    {
      "entropy": 6.314942312240601,
      "epoch": 1.430467397576457,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048063490985042506,
      "loss": 5.9784,
      "mean_token_accuracy": 0.17387469857931137,
      "num_tokens": 31425997.0,
      "step": 12395
    },
    {
      "entropy": 6.21344633102417,
      "epoch": 1.4310444316214657,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004806181586261626,
      "loss": 5.8897,
      "mean_token_accuracy": 0.17805962413549423,
      "num_tokens": 31439967.0,
      "step": 12400
    },
    {
      "entropy": 6.259420156478882,
      "epoch": 1.4316214656664743,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048060140048603544,
      "loss": 5.8386,
      "mean_token_accuracy": 0.18475070148706435,
      "num_tokens": 31453122.0,
      "step": 12405
    },
    {
      "entropy": 6.274561071395874,
      "epoch": 1.432198499711483,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004805846354306073,
      "loss": 5.8736,
      "mean_token_accuracy": 0.18380601108074188,
      "num_tokens": 31465043.0,
      "step": 12410
    },
    {
      "entropy": 6.176213502883911,
      "epoch": 1.4327755337564916,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048056786346044214,
      "loss": 5.7886,
      "mean_token_accuracy": 0.19241779744625093,
      "num_tokens": 31477003.0,
      "step": 12415
    },
    {
      "entropy": 6.188734531402588,
      "epoch": 1.4333525678015002,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048055108457610395,
      "loss": 5.8743,
      "mean_token_accuracy": 0.1861951768398285,
      "num_tokens": 31491612.0,
      "step": 12420
    },
    {
      "entropy": 6.244323778152466,
      "epoch": 1.4339296018465089,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004805342987781571,
      "loss": 5.8058,
      "mean_token_accuracy": 0.1916755333542824,
      "num_tokens": 31504488.0,
      "step": 12425
    },
    {
      "entropy": 6.255792427062988,
      "epoch": 1.4345066358915175,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004805175060671663,
      "loss": 5.9023,
      "mean_token_accuracy": 0.18422753661870955,
      "num_tokens": 31517770.0,
      "step": 12430
    },
    {
      "entropy": 6.158941078186035,
      "epoch": 1.4350836699365264,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004805007064436962,
      "loss": 5.6754,
      "mean_token_accuracy": 0.19200352877378463,
      "num_tokens": 31530262.0,
      "step": 12435
    },
    {
      "entropy": 6.198278379440308,
      "epoch": 1.4356607039815348,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004804838999083119,
      "loss": 5.8457,
      "mean_token_accuracy": 0.17849596440792084,
      "num_tokens": 31543988.0,
      "step": 12440
    },
    {
      "entropy": 6.189879465103149,
      "epoch": 1.4362377380265436,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004804670864615787,
      "loss": 5.7892,
      "mean_token_accuracy": 0.18699323385953903,
      "num_tokens": 31555690.0,
      "step": 12445
    },
    {
      "entropy": 6.182527351379394,
      "epoch": 1.4368147720715523,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048045026610406223,
      "loss": 5.8466,
      "mean_token_accuracy": 0.18502853512763978,
      "num_tokens": 31568542.0,
      "step": 12450
    },
    {
      "entropy": 6.2215944766998295,
      "epoch": 1.437391806116561,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000480433438836328,
      "loss": 5.902,
      "mean_token_accuracy": 0.18309358209371568,
      "num_tokens": 31580538.0,
      "step": 12455
    },
    {
      "entropy": 6.2771049499511715,
      "epoch": 1.4379688401615696,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048041660465894206,
      "loss": 5.7729,
      "mean_token_accuracy": 0.19393041133880615,
      "num_tokens": 31593414.0,
      "step": 12460
    },
    {
      "entropy": 6.179455280303955,
      "epoch": 1.4385458742065782,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004803997635724707,
      "loss": 5.8214,
      "mean_token_accuracy": 0.18613655418157576,
      "num_tokens": 31605931.0,
      "step": 12465
    },
    {
      "entropy": 6.286129140853882,
      "epoch": 1.4391229082515868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004803829155774804,
      "loss": 5.8051,
      "mean_token_accuracy": 0.18124421834945678,
      "num_tokens": 31617949.0,
      "step": 12470
    },
    {
      "entropy": 6.259036111831665,
      "epoch": 1.4396999422965955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004803660606745377,
      "loss": 5.8616,
      "mean_token_accuracy": 0.184785795211792,
      "num_tokens": 31630429.0,
      "step": 12475
    },
    {
      "entropy": 6.1880269050598145,
      "epoch": 1.4402769763416041,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048034919886420953,
      "loss": 5.8446,
      "mean_token_accuracy": 0.18876426219940184,
      "num_tokens": 31641430.0,
      "step": 12480
    },
    {
      "entropy": 6.288239336013794,
      "epoch": 1.4408540103866128,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048033233014706304,
      "loss": 5.9516,
      "mean_token_accuracy": 0.17848838716745377,
      "num_tokens": 31653465.0,
      "step": 12485
    },
    {
      "entropy": 6.277260160446167,
      "epoch": 1.4414310444316214,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048031545452366565,
      "loss": 5.8442,
      "mean_token_accuracy": 0.18718578219413756,
      "num_tokens": 31665550.0,
      "step": 12490
    },
    {
      "entropy": 6.1531486988067625,
      "epoch": 1.44200807847663,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048029857199458493,
      "loss": 5.8269,
      "mean_token_accuracy": 0.1947907954454422,
      "num_tokens": 31677991.0,
      "step": 12495
    },
    {
      "entropy": 6.281120872497558,
      "epoch": 1.442585112521639,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048028168256038873,
      "loss": 5.8773,
      "mean_token_accuracy": 0.1840187504887581,
      "num_tokens": 31690148.0,
      "step": 12500
    },
    {
      "entropy": 6.29625096321106,
      "epoch": 1.4431621465666473,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048026478622164513,
      "loss": 5.9026,
      "mean_token_accuracy": 0.18740195482969285,
      "num_tokens": 31703632.0,
      "step": 12505
    },
    {
      "entropy": 6.215112113952637,
      "epoch": 1.4437391806116562,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048024788297892234,
      "loss": 5.868,
      "mean_token_accuracy": 0.18766642957925797,
      "num_tokens": 31714722.0,
      "step": 12510
    },
    {
      "entropy": 6.27773609161377,
      "epoch": 1.4443162146566648,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000480230972832789,
      "loss": 5.8095,
      "mean_token_accuracy": 0.18923597037792206,
      "num_tokens": 31726554.0,
      "step": 12515
    },
    {
      "entropy": 6.301509237289428,
      "epoch": 1.4448932487016735,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048021405578381384,
      "loss": 5.9398,
      "mean_token_accuracy": 0.18375100195407867,
      "num_tokens": 31738604.0,
      "step": 12520
    },
    {
      "entropy": 6.126003217697144,
      "epoch": 1.445470282746682,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004801971318325659,
      "loss": 5.7896,
      "mean_token_accuracy": 0.19540145993232727,
      "num_tokens": 31750084.0,
      "step": 12525
    },
    {
      "entropy": 6.2102593898773195,
      "epoch": 1.4460473167916907,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004801802009796142,
      "loss": 5.842,
      "mean_token_accuracy": 0.1842881217598915,
      "num_tokens": 31764068.0,
      "step": 12530
    },
    {
      "entropy": 6.2234776496887205,
      "epoch": 1.4466243508366994,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004801632632255285,
      "loss": 5.7923,
      "mean_token_accuracy": 0.18666674345731735,
      "num_tokens": 31777425.0,
      "step": 12535
    },
    {
      "entropy": 6.188868808746338,
      "epoch": 1.447201384881708,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004801463185708783,
      "loss": 5.8304,
      "mean_token_accuracy": 0.18678562343120575,
      "num_tokens": 31789767.0,
      "step": 12540
    },
    {
      "entropy": 6.182145690917968,
      "epoch": 1.4477784189267167,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048012936701623363,
      "loss": 5.7839,
      "mean_token_accuracy": 0.19578426480293273,
      "num_tokens": 31801399.0,
      "step": 12545
    },
    {
      "entropy": 6.246342611312866,
      "epoch": 1.4483554529717253,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048011240856216456,
      "loss": 5.841,
      "mean_token_accuracy": 0.19098050594329835,
      "num_tokens": 31814260.0,
      "step": 12550
    },
    {
      "entropy": 6.253865146636963,
      "epoch": 1.448932487016734,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048009544320924154,
      "loss": 5.8306,
      "mean_token_accuracy": 0.1913675770163536,
      "num_tokens": 31826674.0,
      "step": 12555
    },
    {
      "entropy": 6.233391857147216,
      "epoch": 1.4495095210617426,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004800784709580351,
      "loss": 5.9593,
      "mean_token_accuracy": 0.17964572161436082,
      "num_tokens": 31839951.0,
      "step": 12560
    },
    {
      "entropy": 6.229328966140747,
      "epoch": 1.4500865551067512,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004800614918091161,
      "loss": 5.7975,
      "mean_token_accuracy": 0.19147899448871614,
      "num_tokens": 31852344.0,
      "step": 12565
    },
    {
      "entropy": 6.162231969833374,
      "epoch": 1.4506635891517599,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800445057630558,
      "loss": 5.7232,
      "mean_token_accuracy": 0.1969872385263443,
      "num_tokens": 31865134.0,
      "step": 12570
    },
    {
      "entropy": 6.166445922851563,
      "epoch": 1.4512406231967687,
      "grad_norm": 0.875,
      "learning_rate": 0.0004800275128204254,
      "loss": 5.7884,
      "mean_token_accuracy": 0.1930304616689682,
      "num_tokens": 31877395.0,
      "step": 12575
    },
    {
      "entropy": 6.238461065292358,
      "epoch": 1.4518176572417771,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048001051298179637,
      "loss": 5.8195,
      "mean_token_accuracy": 0.18899209946393966,
      "num_tokens": 31889421.0,
      "step": 12580
    },
    {
      "entropy": 6.2282778263092045,
      "epoch": 1.452394691286786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004799935062477407,
      "loss": 5.8288,
      "mean_token_accuracy": 0.19454032629728318,
      "num_tokens": 31903004.0,
      "step": 12585
    },
    {
      "entropy": 6.256877613067627,
      "epoch": 1.4529717253317946,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047997649261883013,
      "loss": 5.8615,
      "mean_token_accuracy": 0.18735668510198594,
      "num_tokens": 31914682.0,
      "step": 12590
    },
    {
      "entropy": 6.271214580535888,
      "epoch": 1.4535487593768033,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004799594720956371,
      "loss": 5.8735,
      "mean_token_accuracy": 0.18234769403934478,
      "num_tokens": 31928861.0,
      "step": 12595
    },
    {
      "entropy": 6.219674062728882,
      "epoch": 1.454125793421812,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047994244467873407,
      "loss": 5.8199,
      "mean_token_accuracy": 0.18914350420236586,
      "num_tokens": 31942043.0,
      "step": 12600
    },
    {
      "entropy": 6.3054986000061035,
      "epoch": 1.4547028274668206,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047992541036869364,
      "loss": 5.9115,
      "mean_token_accuracy": 0.18517610132694245,
      "num_tokens": 31954453.0,
      "step": 12605
    },
    {
      "entropy": 6.214066410064698,
      "epoch": 1.4552798615118292,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004799083691660889,
      "loss": 5.79,
      "mean_token_accuracy": 0.18852175921201705,
      "num_tokens": 31967998.0,
      "step": 12610
    },
    {
      "entropy": 6.210002374649048,
      "epoch": 1.4558568955568378,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004798913210714929,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1793137639760971,
      "num_tokens": 31980445.0,
      "step": 12615
    },
    {
      "entropy": 6.150847005844116,
      "epoch": 1.4564339296018465,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047987426608547915,
      "loss": 5.8243,
      "mean_token_accuracy": 0.19338324815034866,
      "num_tokens": 31993706.0,
      "step": 12620
    },
    {
      "entropy": 6.24650821685791,
      "epoch": 1.4570109636468551,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004798572042086212,
      "loss": 5.818,
      "mean_token_accuracy": 0.1904278039932251,
      "num_tokens": 32005189.0,
      "step": 12625
    },
    {
      "entropy": 6.198037004470825,
      "epoch": 1.4575879976918638,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047984013544149286,
      "loss": 5.8779,
      "mean_token_accuracy": 0.1868915855884552,
      "num_tokens": 32017883.0,
      "step": 12630
    },
    {
      "entropy": 6.173274040222168,
      "epoch": 1.4581650317368724,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047982305978466836,
      "loss": 5.8172,
      "mean_token_accuracy": 0.18823535442352296,
      "num_tokens": 32030701.0,
      "step": 12635
    },
    {
      "entropy": 6.19607892036438,
      "epoch": 1.4587420657818813,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047980597723872205,
      "loss": 5.8608,
      "mean_token_accuracy": 0.18068182170391084,
      "num_tokens": 32042726.0,
      "step": 12640
    },
    {
      "entropy": 6.250464677810669,
      "epoch": 1.4593190998268897,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004797888878042283,
      "loss": 5.7939,
      "mean_token_accuracy": 0.1887376293540001,
      "num_tokens": 32055158.0,
      "step": 12645
    },
    {
      "entropy": 6.2482497215271,
      "epoch": 1.4598961338718985,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047977179148176217,
      "loss": 5.8465,
      "mean_token_accuracy": 0.1870214119553566,
      "num_tokens": 32068416.0,
      "step": 12650
    },
    {
      "entropy": 6.267807340621948,
      "epoch": 1.4604731679169072,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004797546882718985,
      "loss": 5.8734,
      "mean_token_accuracy": 0.18503910601139067,
      "num_tokens": 32081800.0,
      "step": 12655
    },
    {
      "entropy": 6.257523679733277,
      "epoch": 1.4610502019619158,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00047973757817521256,
      "loss": 5.8951,
      "mean_token_accuracy": 0.18366579264402388,
      "num_tokens": 32095134.0,
      "step": 12660
    },
    {
      "entropy": 6.140337419509888,
      "epoch": 1.4616272360069245,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004797204611922799,
      "loss": 5.7981,
      "mean_token_accuracy": 0.18688902556896209,
      "num_tokens": 32106685.0,
      "step": 12665
    },
    {
      "entropy": 6.24197998046875,
      "epoch": 1.462204270051933,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047970333732367626,
      "loss": 5.8818,
      "mean_token_accuracy": 0.1857426643371582,
      "num_tokens": 32120072.0,
      "step": 12670
    },
    {
      "entropy": 6.273858261108399,
      "epoch": 1.4627813040969417,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047968620656997754,
      "loss": 5.8336,
      "mean_token_accuracy": 0.18685563951730727,
      "num_tokens": 32133205.0,
      "step": 12675
    },
    {
      "entropy": 6.162689733505249,
      "epoch": 1.4633583381419504,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047966906893175994,
      "loss": 5.7726,
      "mean_token_accuracy": 0.19652485847473145,
      "num_tokens": 32147674.0,
      "step": 12680
    },
    {
      "entropy": 6.2253711223602295,
      "epoch": 1.463935372186959,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004796519244095998,
      "loss": 5.8332,
      "mean_token_accuracy": 0.1868842676281929,
      "num_tokens": 32159689.0,
      "step": 12685
    },
    {
      "entropy": 6.231762504577636,
      "epoch": 1.4645124062319677,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047963477300407394,
      "loss": 5.88,
      "mean_token_accuracy": 0.1852923259139061,
      "num_tokens": 32172641.0,
      "step": 12690
    },
    {
      "entropy": 6.186445713043213,
      "epoch": 1.4650894402769763,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047961761471575903,
      "loss": 5.8562,
      "mean_token_accuracy": 0.18955678045749663,
      "num_tokens": 32184512.0,
      "step": 12695
    },
    {
      "entropy": 6.208684396743775,
      "epoch": 1.465666474321985,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047960044954523237,
      "loss": 5.8444,
      "mean_token_accuracy": 0.19104921221733093,
      "num_tokens": 32195908.0,
      "step": 12700
    },
    {
      "entropy": 6.22921404838562,
      "epoch": 1.4662435083669936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047958327749307117,
      "loss": 5.8226,
      "mean_token_accuracy": 0.18249177634716035,
      "num_tokens": 32208855.0,
      "step": 12705
    },
    {
      "entropy": 6.207791137695312,
      "epoch": 1.4668205424120022,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047956609855985305,
      "loss": 5.8073,
      "mean_token_accuracy": 0.19072900712490082,
      "num_tokens": 32222769.0,
      "step": 12710
    },
    {
      "entropy": 6.298447751998902,
      "epoch": 1.467397576457011,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004795489127461559,
      "loss": 5.8722,
      "mean_token_accuracy": 0.18330584615468978,
      "num_tokens": 32235192.0,
      "step": 12715
    },
    {
      "entropy": 6.170766973495484,
      "epoch": 1.4679746105020195,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047953172005255756,
      "loss": 5.8502,
      "mean_token_accuracy": 0.19227186292409898,
      "num_tokens": 32249196.0,
      "step": 12720
    },
    {
      "entropy": 6.275102329254151,
      "epoch": 1.4685516445470284,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004795145204796365,
      "loss": 5.9046,
      "mean_token_accuracy": 0.18350826054811478,
      "num_tokens": 32261591.0,
      "step": 12725
    },
    {
      "entropy": 6.225742959976197,
      "epoch": 1.469128678592037,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004794973140279711,
      "loss": 5.9135,
      "mean_token_accuracy": 0.1830937996506691,
      "num_tokens": 32275409.0,
      "step": 12730
    },
    {
      "entropy": 6.21242389678955,
      "epoch": 1.4697057126370456,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047948010069814005,
      "loss": 5.7261,
      "mean_token_accuracy": 0.19143961519002914,
      "num_tokens": 32287897.0,
      "step": 12735
    },
    {
      "entropy": 6.229421234130859,
      "epoch": 1.4702827466820543,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004794628804907225,
      "loss": 5.9411,
      "mean_token_accuracy": 0.18070883750915528,
      "num_tokens": 32300595.0,
      "step": 12740
    },
    {
      "entropy": 6.2105930805206295,
      "epoch": 1.470859780727063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047944565340629755,
      "loss": 5.8604,
      "mean_token_accuracy": 0.1899055391550064,
      "num_tokens": 32312374.0,
      "step": 12745
    },
    {
      "entropy": 6.23991847038269,
      "epoch": 1.4714368147720716,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047942841944544453,
      "loss": 5.7803,
      "mean_token_accuracy": 0.19597933292388917,
      "num_tokens": 32326526.0,
      "step": 12750
    },
    {
      "entropy": 6.244479513168335,
      "epoch": 1.4720138488170802,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004794111786087431,
      "loss": 5.9128,
      "mean_token_accuracy": 0.1847749575972557,
      "num_tokens": 32339591.0,
      "step": 12755
    },
    {
      "entropy": 6.253715229034424,
      "epoch": 1.4725908828620888,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004793939308967733,
      "loss": 5.8651,
      "mean_token_accuracy": 0.18412415832281112,
      "num_tokens": 32352802.0,
      "step": 12760
    },
    {
      "entropy": 6.175605773925781,
      "epoch": 1.4731679169070975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793766763101152,
      "loss": 5.836,
      "mean_token_accuracy": 0.18129383474588395,
      "num_tokens": 32365951.0,
      "step": 12765
    },
    {
      "entropy": 6.256710958480835,
      "epoch": 1.4737449509521061,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047935941484934905,
      "loss": 5.8717,
      "mean_token_accuracy": 0.18162050247192382,
      "num_tokens": 32378573.0,
      "step": 12770
    },
    {
      "entropy": 6.222249698638916,
      "epoch": 1.4743219849971148,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047934214651505547,
      "loss": 5.8042,
      "mean_token_accuracy": 0.19582674652338028,
      "num_tokens": 32392135.0,
      "step": 12775
    },
    {
      "entropy": 6.225774049758911,
      "epoch": 1.4748990190421236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047932487130781533,
      "loss": 5.9057,
      "mean_token_accuracy": 0.18266624957323074,
      "num_tokens": 32404821.0,
      "step": 12780
    },
    {
      "entropy": 6.191546440124512,
      "epoch": 1.475476053087132,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004793075892282097,
      "loss": 5.7929,
      "mean_token_accuracy": 0.18894983530044557,
      "num_tokens": 32417118.0,
      "step": 12785
    },
    {
      "entropy": 6.229408884048462,
      "epoch": 1.476053087132141,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792903002768197,
      "loss": 5.799,
      "mean_token_accuracy": 0.19217437952756883,
      "num_tokens": 32428272.0,
      "step": 12790
    },
    {
      "entropy": 6.20485668182373,
      "epoch": 1.4766301211771495,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047927300445422687,
      "loss": 5.8709,
      "mean_token_accuracy": 0.18851347118616105,
      "num_tokens": 32441038.0,
      "step": 12795
    },
    {
      "entropy": 6.157735919952392,
      "epoch": 1.4772071552221582,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004792557017610131,
      "loss": 5.7644,
      "mean_token_accuracy": 0.18444529175758362,
      "num_tokens": 32452959.0,
      "step": 12800
    },
    {
      "entropy": 6.2351813316345215,
      "epoch": 1.4777841892671668,
      "grad_norm": 0.875,
      "learning_rate": 0.00047923839219776027,
      "loss": 5.8674,
      "mean_token_accuracy": 0.18899247646331788,
      "num_tokens": 32465398.0,
      "step": 12805
    },
    {
      "entropy": 6.215856409072876,
      "epoch": 1.4783612233121755,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004792210757650506,
      "loss": 5.7686,
      "mean_token_accuracy": 0.18553013503551483,
      "num_tokens": 32477128.0,
      "step": 12810
    },
    {
      "entropy": 6.198601293563843,
      "epoch": 1.478938257357184,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047920375246346636,
      "loss": 5.8599,
      "mean_token_accuracy": 0.1865470290184021,
      "num_tokens": 32489831.0,
      "step": 12815
    },
    {
      "entropy": 6.2368574142456055,
      "epoch": 1.4795152914021927,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047918642229359044,
      "loss": 5.9092,
      "mean_token_accuracy": 0.1766751751303673,
      "num_tokens": 32502239.0,
      "step": 12820
    },
    {
      "entropy": 6.258724117279053,
      "epoch": 1.4800923254472014,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004791690852560056,
      "loss": 5.8484,
      "mean_token_accuracy": 0.18253364115953447,
      "num_tokens": 32515374.0,
      "step": 12825
    },
    {
      "entropy": 6.1911149501800535,
      "epoch": 1.48066935949221,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000479151741351295,
      "loss": 5.8484,
      "mean_token_accuracy": 0.18699524253606797,
      "num_tokens": 32527208.0,
      "step": 12830
    },
    {
      "entropy": 6.256737089157104,
      "epoch": 1.4812463935372187,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000479134390580042,
      "loss": 5.9473,
      "mean_token_accuracy": 0.1810898333787918,
      "num_tokens": 32540834.0,
      "step": 12835
    },
    {
      "entropy": 6.211868619918823,
      "epoch": 1.4818234275822273,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047911703294283015,
      "loss": 5.786,
      "mean_token_accuracy": 0.18922590464353561,
      "num_tokens": 32552833.0,
      "step": 12840
    },
    {
      "entropy": 6.132483673095703,
      "epoch": 1.482400461627236,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047909966844024334,
      "loss": 5.7914,
      "mean_token_accuracy": 0.18970187604427338,
      "num_tokens": 32565484.0,
      "step": 12845
    },
    {
      "entropy": 6.280906915664673,
      "epoch": 1.4829774956722446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047908229707286547,
      "loss": 5.906,
      "mean_token_accuracy": 0.1820051908493042,
      "num_tokens": 32579099.0,
      "step": 12850
    },
    {
      "entropy": 6.243012237548828,
      "epoch": 1.4835545297172534,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000479064918841281,
      "loss": 5.7967,
      "mean_token_accuracy": 0.19242032319307328,
      "num_tokens": 32591530.0,
      "step": 12855
    },
    {
      "entropy": 6.254906415939331,
      "epoch": 1.4841315637622619,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047904753374607427,
      "loss": 5.9273,
      "mean_token_accuracy": 0.18533624112606048,
      "num_tokens": 32603980.0,
      "step": 12860
    },
    {
      "entropy": 6.229150390625,
      "epoch": 1.4847085978072707,
      "grad_norm": 1.0,
      "learning_rate": 0.00047903014178783015,
      "loss": 5.8417,
      "mean_token_accuracy": 0.18629190921783448,
      "num_tokens": 32617593.0,
      "step": 12865
    },
    {
      "entropy": 6.197377157211304,
      "epoch": 1.4852856318522794,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004790127429671335,
      "loss": 5.834,
      "mean_token_accuracy": 0.19327289313077928,
      "num_tokens": 32629813.0,
      "step": 12870
    },
    {
      "entropy": 6.262298583984375,
      "epoch": 1.485862665897288,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004789953372845697,
      "loss": 5.8929,
      "mean_token_accuracy": 0.1843281492590904,
      "num_tokens": 32640982.0,
      "step": 12875
    },
    {
      "entropy": 6.172667217254639,
      "epoch": 1.4864396999422966,
      "grad_norm": 0.9375,
      "learning_rate": 0.000478977924740724,
      "loss": 5.769,
      "mean_token_accuracy": 0.19383094906806947,
      "num_tokens": 32652717.0,
      "step": 12880
    },
    {
      "entropy": 6.213776636123657,
      "epoch": 1.4870167339873053,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004789605053361821,
      "loss": 5.8442,
      "mean_token_accuracy": 0.18912729173898696,
      "num_tokens": 32665531.0,
      "step": 12885
    },
    {
      "entropy": 6.26261568069458,
      "epoch": 1.487593768032314,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004789430790715299,
      "loss": 5.8297,
      "mean_token_accuracy": 0.1862094968557358,
      "num_tokens": 32679296.0,
      "step": 12890
    },
    {
      "entropy": 6.2569139957427975,
      "epoch": 1.4881708020773226,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047892564594735355,
      "loss": 5.8321,
      "mean_token_accuracy": 0.19597998708486558,
      "num_tokens": 32691806.0,
      "step": 12895
    },
    {
      "entropy": 6.16771936416626,
      "epoch": 1.4887478361223312,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047890820596423943,
      "loss": 5.8364,
      "mean_token_accuracy": 0.18944050669670104,
      "num_tokens": 32704280.0,
      "step": 12900
    },
    {
      "entropy": 6.183144998550415,
      "epoch": 1.4893248701673398,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000478890759122774,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1874974936246872,
      "num_tokens": 32716101.0,
      "step": 12905
    },
    {
      "entropy": 6.2596136093139645,
      "epoch": 1.4899019042123485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004788733054235443,
      "loss": 5.8674,
      "mean_token_accuracy": 0.18004628419876098,
      "num_tokens": 32728063.0,
      "step": 12910
    },
    {
      "entropy": 6.227377843856812,
      "epoch": 1.4904789382573571,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047885584486713717,
      "loss": 5.7787,
      "mean_token_accuracy": 0.19393699765205383,
      "num_tokens": 32742084.0,
      "step": 12915
    },
    {
      "entropy": 6.193652105331421,
      "epoch": 1.491055972302366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004788383774541399,
      "loss": 5.8288,
      "mean_token_accuracy": 0.19240004420280457,
      "num_tokens": 32754708.0,
      "step": 12920
    },
    {
      "entropy": 6.196392679214478,
      "epoch": 1.4916330063473744,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004788209031851402,
      "loss": 5.8155,
      "mean_token_accuracy": 0.1869417428970337,
      "num_tokens": 32767326.0,
      "step": 12925
    },
    {
      "entropy": 6.250534582138061,
      "epoch": 1.4922100403923833,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004788034220607256,
      "loss": 5.8893,
      "mean_token_accuracy": 0.18714260458946227,
      "num_tokens": 32781334.0,
      "step": 12930
    },
    {
      "entropy": 6.190632104873657,
      "epoch": 1.4927870744373917,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047878593408148405,
      "loss": 5.9132,
      "mean_token_accuracy": 0.1968586578965187,
      "num_tokens": 32793795.0,
      "step": 12935
    },
    {
      "entropy": 6.203883361816406,
      "epoch": 1.4933641084824005,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047876843924800393,
      "loss": 5.749,
      "mean_token_accuracy": 0.19471774995326996,
      "num_tokens": 32807484.0,
      "step": 12940
    },
    {
      "entropy": 6.2029859066009525,
      "epoch": 1.4939411425274092,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004787509375608735,
      "loss": 5.8228,
      "mean_token_accuracy": 0.19489093869924545,
      "num_tokens": 32819129.0,
      "step": 12945
    },
    {
      "entropy": 6.186358690261841,
      "epoch": 1.4945181765724178,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047873342902068157,
      "loss": 5.8536,
      "mean_token_accuracy": 0.18467129319906234,
      "num_tokens": 32832077.0,
      "step": 12950
    },
    {
      "entropy": 6.198476409912109,
      "epoch": 1.4950952106174265,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047871591362801694,
      "loss": 5.8552,
      "mean_token_accuracy": 0.1873326539993286,
      "num_tokens": 32845629.0,
      "step": 12955
    },
    {
      "entropy": 6.225423955917359,
      "epoch": 1.495672244662435,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004786983913834687,
      "loss": 5.8204,
      "mean_token_accuracy": 0.19441826343536378,
      "num_tokens": 32857233.0,
      "step": 12960
    },
    {
      "entropy": 6.213121604919434,
      "epoch": 1.4962492787074437,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047868086228762633,
      "loss": 5.862,
      "mean_token_accuracy": 0.18557816445827485,
      "num_tokens": 32869210.0,
      "step": 12965
    },
    {
      "entropy": 6.279606199264526,
      "epoch": 1.4968263127524524,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047866332634107926,
      "loss": 5.9093,
      "mean_token_accuracy": 0.18809690475463867,
      "num_tokens": 32881633.0,
      "step": 12970
    },
    {
      "entropy": 6.282651138305664,
      "epoch": 1.497403346797461,
      "grad_norm": 0.875,
      "learning_rate": 0.00047864578354441743,
      "loss": 5.7917,
      "mean_token_accuracy": 0.19121649116277695,
      "num_tokens": 32893904.0,
      "step": 12975
    },
    {
      "entropy": 6.19729437828064,
      "epoch": 1.4979803808424696,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004786282338982308,
      "loss": 5.845,
      "mean_token_accuracy": 0.19188052266836167,
      "num_tokens": 32906612.0,
      "step": 12980
    },
    {
      "entropy": 6.298680114746094,
      "epoch": 1.4985574148874783,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004786106774031096,
      "loss": 5.8951,
      "mean_token_accuracy": 0.18667923361063005,
      "num_tokens": 32919677.0,
      "step": 12985
    },
    {
      "entropy": 6.1812975883483885,
      "epoch": 1.499134448932487,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004785931140596445,
      "loss": 5.7911,
      "mean_token_accuracy": 0.1890546813607216,
      "num_tokens": 32932385.0,
      "step": 12990
    },
    {
      "entropy": 6.203750133514404,
      "epoch": 1.4997114829774958,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047857554386842606,
      "loss": 5.8171,
      "mean_token_accuracy": 0.192004032433033,
      "num_tokens": 32945524.0,
      "step": 12995
    },
    {
      "entropy": 6.218576574325562,
      "epoch": 1.5002885170225042,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047855796683004534,
      "loss": 5.8257,
      "mean_token_accuracy": 0.1879967123270035,
      "num_tokens": 32957871.0,
      "step": 13000
    },
    {
      "entropy": 6.212220954895019,
      "epoch": 1.500865551067513,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047854038294509356,
      "loss": 5.8646,
      "mean_token_accuracy": 0.18789520859718323,
      "num_tokens": 32971727.0,
      "step": 13005
    },
    {
      "entropy": 6.205683469772339,
      "epoch": 1.5014425851125215,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004785227922141621,
      "loss": 5.7694,
      "mean_token_accuracy": 0.1907043367624283,
      "num_tokens": 32984884.0,
      "step": 13010
    },
    {
      "entropy": 6.196247911453247,
      "epoch": 1.5020196191575304,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047850519463784263,
      "loss": 5.8664,
      "mean_token_accuracy": 0.18877162635326386,
      "num_tokens": 32997475.0,
      "step": 13015
    },
    {
      "entropy": 6.223559379577637,
      "epoch": 1.502596653202539,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047848759021672693,
      "loss": 5.8804,
      "mean_token_accuracy": 0.18658992201089858,
      "num_tokens": 33010466.0,
      "step": 13020
    },
    {
      "entropy": 6.2706708908081055,
      "epoch": 1.5031736872475476,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004784699789514073,
      "loss": 5.858,
      "mean_token_accuracy": 0.18523260354995727,
      "num_tokens": 33023806.0,
      "step": 13025
    },
    {
      "entropy": 6.2270934104919435,
      "epoch": 1.5037507212925563,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000478452360842476,
      "loss": 5.8134,
      "mean_token_accuracy": 0.1850523442029953,
      "num_tokens": 33036132.0,
      "step": 13030
    },
    {
      "entropy": 6.228479051589966,
      "epoch": 1.504327755337565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047843473589052557,
      "loss": 5.8994,
      "mean_token_accuracy": 0.1839058592915535,
      "num_tokens": 33049706.0,
      "step": 13035
    },
    {
      "entropy": 6.27629656791687,
      "epoch": 1.5049047893825735,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047841710409614893,
      "loss": 5.7968,
      "mean_token_accuracy": 0.19325800091028214,
      "num_tokens": 33061725.0,
      "step": 13040
    },
    {
      "entropy": 6.209563779830932,
      "epoch": 1.5054818234275822,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004783994654599389,
      "loss": 5.8219,
      "mean_token_accuracy": 0.1912074312567711,
      "num_tokens": 33074908.0,
      "step": 13045
    },
    {
      "entropy": 6.182907342910767,
      "epoch": 1.5060588574725908,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047838181998248897,
      "loss": 5.8316,
      "mean_token_accuracy": 0.18455471843481064,
      "num_tokens": 33086631.0,
      "step": 13050
    },
    {
      "entropy": 6.249213647842407,
      "epoch": 1.5066358915175995,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004783641676643925,
      "loss": 5.8451,
      "mean_token_accuracy": 0.18560626059770585,
      "num_tokens": 33100109.0,
      "step": 13055
    },
    {
      "entropy": 6.189642000198364,
      "epoch": 1.5072129255626083,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047834650850624334,
      "loss": 5.7714,
      "mean_token_accuracy": 0.190520241856575,
      "num_tokens": 33111953.0,
      "step": 13060
    },
    {
      "entropy": 6.187156867980957,
      "epoch": 1.5077899596076167,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004783288425086353,
      "loss": 5.8051,
      "mean_token_accuracy": 0.1944757506251335,
      "num_tokens": 33124563.0,
      "step": 13065
    },
    {
      "entropy": 6.288137340545655,
      "epoch": 1.5083669936526256,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004783111696721627,
      "loss": 5.8739,
      "mean_token_accuracy": 0.18941331803798675,
      "num_tokens": 33137636.0,
      "step": 13070
    },
    {
      "entropy": 6.25904688835144,
      "epoch": 1.508944027697634,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004782934899974199,
      "loss": 5.9384,
      "mean_token_accuracy": 0.18362302780151368,
      "num_tokens": 33150209.0,
      "step": 13075
    },
    {
      "entropy": 6.1227422714233395,
      "epoch": 1.5095210617426429,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047827580348500147,
      "loss": 5.7493,
      "mean_token_accuracy": 0.1959556981921196,
      "num_tokens": 33162424.0,
      "step": 13080
    },
    {
      "entropy": 6.201741361618042,
      "epoch": 1.5100980957876513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047825811013550246,
      "loss": 5.8272,
      "mean_token_accuracy": 0.19274757653474808,
      "num_tokens": 33174923.0,
      "step": 13085
    },
    {
      "entropy": 6.176445198059082,
      "epoch": 1.5106751298326602,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00047824040994951786,
      "loss": 5.7399,
      "mean_token_accuracy": 0.18690399527549745,
      "num_tokens": 33188022.0,
      "step": 13090
    },
    {
      "entropy": 6.19749345779419,
      "epoch": 1.5112521638776688,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004782227029276429,
      "loss": 5.8245,
      "mean_token_accuracy": 0.18717896640300752,
      "num_tokens": 33201657.0,
      "step": 13095
    },
    {
      "entropy": 6.234110307693482,
      "epoch": 1.5118291979226774,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047820498907047345,
      "loss": 5.8115,
      "mean_token_accuracy": 0.19088313430547715,
      "num_tokens": 33214692.0,
      "step": 13100
    },
    {
      "entropy": 6.264095163345337,
      "epoch": 1.512406231967686,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000478187268378605,
      "loss": 5.8599,
      "mean_token_accuracy": 0.1901498705148697,
      "num_tokens": 33227419.0,
      "step": 13105
    },
    {
      "entropy": 6.251426696777344,
      "epoch": 1.5129832660126947,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047816954085263375,
      "loss": 5.8645,
      "mean_token_accuracy": 0.1884318321943283,
      "num_tokens": 33239859.0,
      "step": 13110
    },
    {
      "entropy": 6.2335686683654785,
      "epoch": 1.5135603000577034,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004781518064931559,
      "loss": 5.8056,
      "mean_token_accuracy": 0.1961516708135605,
      "num_tokens": 33252872.0,
      "step": 13115
    },
    {
      "entropy": 6.23302149772644,
      "epoch": 1.514137334102712,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000478134065300768,
      "loss": 5.7683,
      "mean_token_accuracy": 0.1978388175368309,
      "num_tokens": 33264220.0,
      "step": 13120
    },
    {
      "entropy": 6.198013591766357,
      "epoch": 1.5147143681477209,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004781163172760667,
      "loss": 5.9088,
      "mean_token_accuracy": 0.17742422223091125,
      "num_tokens": 33276379.0,
      "step": 13125
    },
    {
      "entropy": 6.1771632671356205,
      "epoch": 1.5152914021927293,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047809856241964893,
      "loss": 5.7782,
      "mean_token_accuracy": 0.19086233228445054,
      "num_tokens": 33290487.0,
      "step": 13130
    },
    {
      "entropy": 6.285802507400513,
      "epoch": 1.5158684362377381,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004780808007321119,
      "loss": 5.925,
      "mean_token_accuracy": 0.17844678461551666,
      "num_tokens": 33302560.0,
      "step": 13135
    },
    {
      "entropy": 6.29577865600586,
      "epoch": 1.5164454702827466,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004780630322140531,
      "loss": 5.8975,
      "mean_token_accuracy": 0.18178286999464036,
      "num_tokens": 33315077.0,
      "step": 13140
    },
    {
      "entropy": 6.21107120513916,
      "epoch": 1.5170225043277554,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047804525686607,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1888168439269066,
      "num_tokens": 33326769.0,
      "step": 13145
    },
    {
      "entropy": 6.028330612182617,
      "epoch": 1.5175995383727638,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004780274746887606,
      "loss": 5.6805,
      "mean_token_accuracy": 0.19837866723537445,
      "num_tokens": 33340723.0,
      "step": 13150
    },
    {
      "entropy": 6.28200397491455,
      "epoch": 1.5181765724177727,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047800968568272284,
      "loss": 5.8665,
      "mean_token_accuracy": 0.19180553406476974,
      "num_tokens": 33353894.0,
      "step": 13155
    },
    {
      "entropy": 6.304289817810059,
      "epoch": 1.5187536064627813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004779918898485551,
      "loss": 5.9357,
      "mean_token_accuracy": 0.18196574747562408,
      "num_tokens": 33366134.0,
      "step": 13160
    },
    {
      "entropy": 6.253716087341308,
      "epoch": 1.51933064050779,
      "grad_norm": 0.875,
      "learning_rate": 0.00047797408718685614,
      "loss": 5.8386,
      "mean_token_accuracy": 0.18771272748708726,
      "num_tokens": 33377604.0,
      "step": 13165
    },
    {
      "entropy": 6.240119075775146,
      "epoch": 1.5199076745527986,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047795627769822447,
      "loss": 5.8911,
      "mean_token_accuracy": 0.1798719957470894,
      "num_tokens": 33389616.0,
      "step": 13170
    },
    {
      "entropy": 6.218089389801025,
      "epoch": 1.5204847085978073,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047793846138325925,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1870667278766632,
      "num_tokens": 33401493.0,
      "step": 13175
    },
    {
      "entropy": 6.207758283615112,
      "epoch": 1.521061742642816,
      "grad_norm": 0.875,
      "learning_rate": 0.0004779206382425598,
      "loss": 5.7808,
      "mean_token_accuracy": 0.19917986243963243,
      "num_tokens": 33414088.0,
      "step": 13180
    },
    {
      "entropy": 6.280774784088135,
      "epoch": 1.5216387766878245,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004779028082767253,
      "loss": 5.8552,
      "mean_token_accuracy": 0.1854931503534317,
      "num_tokens": 33426481.0,
      "step": 13185
    },
    {
      "entropy": 6.239508438110351,
      "epoch": 1.5222158107328332,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004778849714863557,
      "loss": 5.8284,
      "mean_token_accuracy": 0.1831372782588005,
      "num_tokens": 33437876.0,
      "step": 13190
    },
    {
      "entropy": 6.273824310302734,
      "epoch": 1.5227928447778418,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004778671278720508,
      "loss": 5.8812,
      "mean_token_accuracy": 0.18227548003196717,
      "num_tokens": 33450164.0,
      "step": 13195
    },
    {
      "entropy": 6.179072284698487,
      "epoch": 1.5233698788228507,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004778492774344109,
      "loss": 5.7814,
      "mean_token_accuracy": 0.19670782685279847,
      "num_tokens": 33462583.0,
      "step": 13200
    },
    {
      "entropy": 6.148195743560791,
      "epoch": 1.523946912867859,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004778314201740363,
      "loss": 5.8499,
      "mean_token_accuracy": 0.19271344989538192,
      "num_tokens": 33474099.0,
      "step": 13205
    },
    {
      "entropy": 6.244106960296631,
      "epoch": 1.524523946912868,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047781355609152764,
      "loss": 5.8247,
      "mean_token_accuracy": 0.1880191370844841,
      "num_tokens": 33486565.0,
      "step": 13210
    },
    {
      "entropy": 6.2393872261047365,
      "epoch": 1.5251009809578764,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004777956851874858,
      "loss": 5.7764,
      "mean_token_accuracy": 0.18627868443727494,
      "num_tokens": 33498703.0,
      "step": 13215
    },
    {
      "entropy": 6.2635313987731935,
      "epoch": 1.5256780150028852,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047777780746251176,
      "loss": 5.8641,
      "mean_token_accuracy": 0.18812676668167114,
      "num_tokens": 33511431.0,
      "step": 13220
    },
    {
      "entropy": 6.1785430908203125,
      "epoch": 1.5262550490478937,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047775992291720687,
      "loss": 5.7763,
      "mean_token_accuracy": 0.18891059458255768,
      "num_tokens": 33524771.0,
      "step": 13225
    },
    {
      "entropy": 6.196375751495362,
      "epoch": 1.5268320830929025,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004777420315521728,
      "loss": 5.7814,
      "mean_token_accuracy": 0.19751103222370148,
      "num_tokens": 33538362.0,
      "step": 13230
    },
    {
      "entropy": 6.207241725921631,
      "epoch": 1.5274091171379112,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004777241333680111,
      "loss": 5.8788,
      "mean_token_accuracy": 0.18835172653198243,
      "num_tokens": 33550905.0,
      "step": 13235
    },
    {
      "entropy": 6.270090198516845,
      "epoch": 1.5279861511829198,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004777062283653239,
      "loss": 5.9084,
      "mean_token_accuracy": 0.18963317275047303,
      "num_tokens": 33563665.0,
      "step": 13240
    },
    {
      "entropy": 6.199642372131348,
      "epoch": 1.5285631852279284,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047768831654471333,
      "loss": 5.8013,
      "mean_token_accuracy": 0.19605089575052262,
      "num_tokens": 33576957.0,
      "step": 13245
    },
    {
      "entropy": 6.232164764404297,
      "epoch": 1.529140219272937,
      "grad_norm": 0.875,
      "learning_rate": 0.00047767039790678204,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18266996890306472,
      "num_tokens": 33589641.0,
      "step": 13250
    },
    {
      "entropy": 6.223003244400024,
      "epoch": 1.5297172533179457,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047765247245213255,
      "loss": 5.7598,
      "mean_token_accuracy": 0.19582873582839966,
      "num_tokens": 33602377.0,
      "step": 13255
    },
    {
      "entropy": 6.210393571853638,
      "epoch": 1.5302942873629544,
      "grad_norm": 0.875,
      "learning_rate": 0.0004776345401813678,
      "loss": 5.7726,
      "mean_token_accuracy": 0.1928178310394287,
      "num_tokens": 33614796.0,
      "step": 13260
    },
    {
      "entropy": 6.281708812713623,
      "epoch": 1.5308713214079632,
      "grad_norm": 0.875,
      "learning_rate": 0.00047761660109509095,
      "loss": 5.9266,
      "mean_token_accuracy": 0.18113744407892227,
      "num_tokens": 33626988.0,
      "step": 13265
    },
    {
      "entropy": 6.164075231552124,
      "epoch": 1.5314483554529716,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004775986551939054,
      "loss": 5.806,
      "mean_token_accuracy": 0.18676834404468537,
      "num_tokens": 33641376.0,
      "step": 13270
    },
    {
      "entropy": 6.26697473526001,
      "epoch": 1.5320253894979805,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047758070247841465,
      "loss": 5.8363,
      "mean_token_accuracy": 0.18552766591310502,
      "num_tokens": 33655155.0,
      "step": 13275
    },
    {
      "entropy": 6.253081607818603,
      "epoch": 1.532602423542989,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047756274294922266,
      "loss": 5.7956,
      "mean_token_accuracy": 0.19269047379493714,
      "num_tokens": 33668480.0,
      "step": 13280
    },
    {
      "entropy": 6.1516824722290036,
      "epoch": 1.5331794575879978,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004775447766069334,
      "loss": 5.7454,
      "mean_token_accuracy": 0.19429850131273269,
      "num_tokens": 33680919.0,
      "step": 13285
    },
    {
      "entropy": 6.141225433349609,
      "epoch": 1.5337564916330062,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047752680345215115,
      "loss": 5.8257,
      "mean_token_accuracy": 0.18666446059942246,
      "num_tokens": 33693161.0,
      "step": 13290
    },
    {
      "entropy": 6.29141788482666,
      "epoch": 1.534333525678015,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004775088234854805,
      "loss": 5.8906,
      "mean_token_accuracy": 0.18971278816461562,
      "num_tokens": 33707537.0,
      "step": 13295
    },
    {
      "entropy": 6.234605407714843,
      "epoch": 1.5349105597230237,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004774908367075262,
      "loss": 5.788,
      "mean_token_accuracy": 0.18784658908843993,
      "num_tokens": 33719501.0,
      "step": 13300
    },
    {
      "entropy": 6.2662890434265135,
      "epoch": 1.5354875937680323,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004774728431188931,
      "loss": 5.9254,
      "mean_token_accuracy": 0.18093251138925553,
      "num_tokens": 33731938.0,
      "step": 13305
    },
    {
      "entropy": 6.258570003509521,
      "epoch": 1.536064627813041,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047745484272018664,
      "loss": 5.8802,
      "mean_token_accuracy": 0.18894830495119094,
      "num_tokens": 33745509.0,
      "step": 13310
    },
    {
      "entropy": 6.260461044311524,
      "epoch": 1.5366416618580496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004774368355120119,
      "loss": 5.8185,
      "mean_token_accuracy": 0.18980655521154405,
      "num_tokens": 33757818.0,
      "step": 13315
    },
    {
      "entropy": 6.212393951416016,
      "epoch": 1.5372186959030583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000477418821494975,
      "loss": 5.7971,
      "mean_token_accuracy": 0.18991858512163162,
      "num_tokens": 33769840.0,
      "step": 13320
    },
    {
      "entropy": 6.236335372924804,
      "epoch": 1.537795729948067,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004774008006696814,
      "loss": 5.8159,
      "mean_token_accuracy": 0.19045742303133012,
      "num_tokens": 33783713.0,
      "step": 13325
    },
    {
      "entropy": 6.1843362808227536,
      "epoch": 1.5383727639930755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004773827730367375,
      "loss": 5.8239,
      "mean_token_accuracy": 0.19330348819494247,
      "num_tokens": 33796005.0,
      "step": 13330
    },
    {
      "entropy": 6.19154167175293,
      "epoch": 1.5389497980380842,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047736473859674955,
      "loss": 5.7677,
      "mean_token_accuracy": 0.1939645916223526,
      "num_tokens": 33807295.0,
      "step": 13335
    },
    {
      "entropy": 6.2215172290802006,
      "epoch": 1.539526832083093,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047734669735032404,
      "loss": 5.7548,
      "mean_token_accuracy": 0.19064207077026368,
      "num_tokens": 33819772.0,
      "step": 13340
    },
    {
      "entropy": 6.198299407958984,
      "epoch": 1.5401038661281015,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047732864929806796,
      "loss": 5.8536,
      "mean_token_accuracy": 0.1890008732676506,
      "num_tokens": 33831820.0,
      "step": 13345
    },
    {
      "entropy": 6.26213755607605,
      "epoch": 1.5406809001731103,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004773105944405883,
      "loss": 5.9056,
      "mean_token_accuracy": 0.18243135660886764,
      "num_tokens": 33845753.0,
      "step": 13350
    },
    {
      "entropy": 6.312659883499146,
      "epoch": 1.5412579342181187,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047729253277849226,
      "loss": 5.8243,
      "mean_token_accuracy": 0.1889658346772194,
      "num_tokens": 33859438.0,
      "step": 13355
    },
    {
      "entropy": 6.2517393112182615,
      "epoch": 1.5418349682631276,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047727446431238734,
      "loss": 5.8543,
      "mean_token_accuracy": 0.1916137233376503,
      "num_tokens": 33872062.0,
      "step": 13360
    },
    {
      "entropy": 6.150335121154785,
      "epoch": 1.542412002308136,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004772563890428813,
      "loss": 5.7885,
      "mean_token_accuracy": 0.19504280835390092,
      "num_tokens": 33883840.0,
      "step": 13365
    },
    {
      "entropy": 6.2797259330749515,
      "epoch": 1.5429890363531449,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004772383069705821,
      "loss": 5.8811,
      "mean_token_accuracy": 0.1802084416151047,
      "num_tokens": 33896907.0,
      "step": 13370
    },
    {
      "entropy": 6.1925756454467775,
      "epoch": 1.5435660703981535,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004772202180960978,
      "loss": 5.824,
      "mean_token_accuracy": 0.19038461595773698,
      "num_tokens": 33912135.0,
      "step": 13375
    },
    {
      "entropy": 6.2145771980285645,
      "epoch": 1.5441431044431622,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047720212242003703,
      "loss": 5.7761,
      "mean_token_accuracy": 0.19460065215826033,
      "num_tokens": 33925014.0,
      "step": 13380
    },
    {
      "entropy": 6.293187427520752,
      "epoch": 1.5447201384881708,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047718401994300825,
      "loss": 5.8977,
      "mean_token_accuracy": 0.18359797745943068,
      "num_tokens": 33937429.0,
      "step": 13385
    },
    {
      "entropy": 6.264120388031006,
      "epoch": 1.5452971725331794,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047716591066562043,
      "loss": 5.8334,
      "mean_token_accuracy": 0.1919166624546051,
      "num_tokens": 33951030.0,
      "step": 13390
    },
    {
      "entropy": 6.271553564071655,
      "epoch": 1.545874206578188,
      "grad_norm": 0.875,
      "learning_rate": 0.00047714779458848255,
      "loss": 5.9079,
      "mean_token_accuracy": 0.18155051916837692,
      "num_tokens": 33962930.0,
      "step": 13395
    },
    {
      "entropy": 6.241593360900879,
      "epoch": 1.5464512406231967,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004771296717122041,
      "loss": 5.8077,
      "mean_token_accuracy": 0.19454100579023362,
      "num_tokens": 33977113.0,
      "step": 13400
    },
    {
      "entropy": 6.094107007980346,
      "epoch": 1.5470282746682056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004771115420373945,
      "loss": 5.561,
      "mean_token_accuracy": 0.20948735028505325,
      "num_tokens": 33989091.0,
      "step": 13405
    },
    {
      "entropy": 6.213526678085327,
      "epoch": 1.547605308713214,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047709340556466366,
      "loss": 5.9422,
      "mean_token_accuracy": 0.18275455832481385,
      "num_tokens": 34002051.0,
      "step": 13410
    },
    {
      "entropy": 6.288247871398926,
      "epoch": 1.5481823427582229,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047707526229462144,
      "loss": 5.868,
      "mean_token_accuracy": 0.1878260299563408,
      "num_tokens": 34014482.0,
      "step": 13415
    },
    {
      "entropy": 6.256914234161377,
      "epoch": 1.5487593768032313,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047705711222787816,
      "loss": 5.8322,
      "mean_token_accuracy": 0.19088772386312486,
      "num_tokens": 34025665.0,
      "step": 13420
    },
    {
      "entropy": 6.219833898544311,
      "epoch": 1.5493364108482401,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047703895536504423,
      "loss": 5.8105,
      "mean_token_accuracy": 0.1916133299469948,
      "num_tokens": 34038163.0,
      "step": 13425
    },
    {
      "entropy": 6.218587446212768,
      "epoch": 1.5499134448932486,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004770207917067305,
      "loss": 5.8664,
      "mean_token_accuracy": 0.19042722880840302,
      "num_tokens": 34050774.0,
      "step": 13430
    },
    {
      "entropy": 6.31106653213501,
      "epoch": 1.5504904789382574,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047700262125354765,
      "loss": 5.8769,
      "mean_token_accuracy": 0.18001208901405336,
      "num_tokens": 34062480.0,
      "step": 13435
    },
    {
      "entropy": 6.206598854064941,
      "epoch": 1.551067512983266,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047698444400610707,
      "loss": 5.8158,
      "mean_token_accuracy": 0.1934734344482422,
      "num_tokens": 34075314.0,
      "step": 13440
    },
    {
      "entropy": 6.156052923202514,
      "epoch": 1.5516445470282747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047696625996502,
      "loss": 5.7582,
      "mean_token_accuracy": 0.19048720300197602,
      "num_tokens": 34087528.0,
      "step": 13445
    },
    {
      "entropy": 6.1763464450836185,
      "epoch": 1.5522215810732833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047694806913089815,
      "loss": 5.7276,
      "mean_token_accuracy": 0.19297634363174437,
      "num_tokens": 34101740.0,
      "step": 13450
    },
    {
      "entropy": 6.294735431671143,
      "epoch": 1.552798615118292,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004769298715043533,
      "loss": 5.8442,
      "mean_token_accuracy": 0.18769462704658507,
      "num_tokens": 34113042.0,
      "step": 13455
    },
    {
      "entropy": 6.226419305801391,
      "epoch": 1.5533756491633006,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004769116670859975,
      "loss": 5.907,
      "mean_token_accuracy": 0.18598962128162383,
      "num_tokens": 34126801.0,
      "step": 13460
    },
    {
      "entropy": 6.233967208862305,
      "epoch": 1.5539526832083093,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047689345587644314,
      "loss": 5.9031,
      "mean_token_accuracy": 0.1880720391869545,
      "num_tokens": 34139469.0,
      "step": 13465
    },
    {
      "entropy": 6.267644023895263,
      "epoch": 1.554529717253318,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047687523787630256,
      "loss": 5.8374,
      "mean_token_accuracy": 0.18743398189544677,
      "num_tokens": 34151940.0,
      "step": 13470
    },
    {
      "entropy": 6.192278242111206,
      "epoch": 1.5551067512983265,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004768570130861887,
      "loss": 5.7857,
      "mean_token_accuracy": 0.18774448335170746,
      "num_tokens": 34164375.0,
      "step": 13475
    },
    {
      "entropy": 6.236714220046997,
      "epoch": 1.5556837853433354,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004768387815067144,
      "loss": 5.8694,
      "mean_token_accuracy": 0.1834099307656288,
      "num_tokens": 34176585.0,
      "step": 13480
    },
    {
      "entropy": 6.27178201675415,
      "epoch": 1.5562608193883438,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047682054313849304,
      "loss": 5.8508,
      "mean_token_accuracy": 0.18469847589731217,
      "num_tokens": 34188518.0,
      "step": 13485
    },
    {
      "entropy": 6.190626430511474,
      "epoch": 1.5568378534333527,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004768022979821379,
      "loss": 5.7454,
      "mean_token_accuracy": 0.19676989316940308,
      "num_tokens": 34200865.0,
      "step": 13490
    },
    {
      "entropy": 6.260462808609009,
      "epoch": 1.557414887478361,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004767840460382628,
      "loss": 5.8076,
      "mean_token_accuracy": 0.19487357288599014,
      "num_tokens": 34212766.0,
      "step": 13495
    },
    {
      "entropy": 6.239464950561524,
      "epoch": 1.55799192152337,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004767657873074815,
      "loss": 5.8346,
      "mean_token_accuracy": 0.18853898644447326,
      "num_tokens": 34225401.0,
      "step": 13500
    },
    {
      "entropy": 6.214106559753418,
      "epoch": 1.5585689555683784,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004767475217904081,
      "loss": 5.7902,
      "mean_token_accuracy": 0.19690571129322051,
      "num_tokens": 34237474.0,
      "step": 13505
    },
    {
      "entropy": 6.235998296737671,
      "epoch": 1.5591459896133872,
      "grad_norm": 0.875,
      "learning_rate": 0.00047672924948765705,
      "loss": 5.8904,
      "mean_token_accuracy": 0.1807642325758934,
      "num_tokens": 34249598.0,
      "step": 13510
    },
    {
      "entropy": 6.230290842056275,
      "epoch": 1.5597230236583959,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047671097039984293,
      "loss": 5.7697,
      "mean_token_accuracy": 0.19288739711046218,
      "num_tokens": 34261773.0,
      "step": 13515
    },
    {
      "entropy": 6.253873109817505,
      "epoch": 1.5603000577034045,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047669268452758053,
      "loss": 5.8537,
      "mean_token_accuracy": 0.19012995064258575,
      "num_tokens": 34275398.0,
      "step": 13520
    },
    {
      "entropy": 6.197601842880249,
      "epoch": 1.5608770917484132,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047667439187148476,
      "loss": 5.7916,
      "mean_token_accuracy": 0.19038860499858856,
      "num_tokens": 34288934.0,
      "step": 13525
    },
    {
      "entropy": 6.181307172775268,
      "epoch": 1.5614541257934218,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004766560924321711,
      "loss": 5.8186,
      "mean_token_accuracy": 0.1892324984073639,
      "num_tokens": 34302187.0,
      "step": 13530
    },
    {
      "entropy": 6.241639089584351,
      "epoch": 1.5620311598384304,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047663778621025496,
      "loss": 5.8261,
      "mean_token_accuracy": 0.18185652941465377,
      "num_tokens": 34314745.0,
      "step": 13535
    },
    {
      "entropy": 6.228024053573608,
      "epoch": 1.562608193883439,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004766194732063519,
      "loss": 5.7823,
      "mean_token_accuracy": 0.19383231699466705,
      "num_tokens": 34326771.0,
      "step": 13540
    },
    {
      "entropy": 6.21107964515686,
      "epoch": 1.563185227928448,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047660115342107814,
      "loss": 5.8074,
      "mean_token_accuracy": 0.18649399429559707,
      "num_tokens": 34339452.0,
      "step": 13545
    },
    {
      "entropy": 6.2527093410491945,
      "epoch": 1.5637622619734564,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047658282685504973,
      "loss": 5.8561,
      "mean_token_accuracy": 0.17872317433357238,
      "num_tokens": 34351872.0,
      "step": 13550
    },
    {
      "entropy": 6.283057641983032,
      "epoch": 1.5643392960184652,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000476564493508883,
      "loss": 5.8179,
      "mean_token_accuracy": 0.18874061554670335,
      "num_tokens": 34363963.0,
      "step": 13555
    },
    {
      "entropy": 6.295781183242798,
      "epoch": 1.5649163300634736,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047654615338319466,
      "loss": 5.8608,
      "mean_token_accuracy": 0.18437285274267196,
      "num_tokens": 34377245.0,
      "step": 13560
    },
    {
      "entropy": 6.188222980499267,
      "epoch": 1.5654933641084825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004765278064786016,
      "loss": 5.7623,
      "mean_token_accuracy": 0.19033338874578476,
      "num_tokens": 34391047.0,
      "step": 13565
    },
    {
      "entropy": 6.245954418182373,
      "epoch": 1.566070398153491,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047650945279572085,
      "loss": 5.8309,
      "mean_token_accuracy": 0.18295957744121552,
      "num_tokens": 34404655.0,
      "step": 13570
    },
    {
      "entropy": 6.261426258087158,
      "epoch": 1.5666474321984998,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004764910923351698,
      "loss": 5.7627,
      "mean_token_accuracy": 0.19444448053836821,
      "num_tokens": 34417902.0,
      "step": 13575
    },
    {
      "entropy": 6.280911588668824,
      "epoch": 1.5672244662435084,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047647272509756584,
      "loss": 5.8112,
      "mean_token_accuracy": 0.18834198713302613,
      "num_tokens": 34430630.0,
      "step": 13580
    },
    {
      "entropy": 6.2056910514831545,
      "epoch": 1.567801500288517,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004764543510835269,
      "loss": 5.8458,
      "mean_token_accuracy": 0.18694427311420442,
      "num_tokens": 34444320.0,
      "step": 13585
    },
    {
      "entropy": 6.18598165512085,
      "epoch": 1.5683785343335257,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000476435970293671,
      "loss": 5.7221,
      "mean_token_accuracy": 0.20084212124347686,
      "num_tokens": 34457081.0,
      "step": 13590
    },
    {
      "entropy": 6.163789844512939,
      "epoch": 1.5689555683785343,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047641758272861626,
      "loss": 5.7798,
      "mean_token_accuracy": 0.19682853668928146,
      "num_tokens": 34469560.0,
      "step": 13595
    },
    {
      "entropy": 6.2018592834472654,
      "epoch": 1.569532602423543,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004763991883889811,
      "loss": 5.8292,
      "mean_token_accuracy": 0.1882767468690872,
      "num_tokens": 34482036.0,
      "step": 13600
    },
    {
      "entropy": 6.249498462677002,
      "epoch": 1.5701096364685516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004763807872753843,
      "loss": 5.7953,
      "mean_token_accuracy": 0.18720098435878754,
      "num_tokens": 34494316.0,
      "step": 13605
    },
    {
      "entropy": 6.1701406955719,
      "epoch": 1.5706866705135603,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047636237938844484,
      "loss": 5.7674,
      "mean_token_accuracy": 0.18870268166065216,
      "num_tokens": 34507026.0,
      "step": 13610
    },
    {
      "entropy": 6.198963785171509,
      "epoch": 1.571263704558569,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004763439647287817,
      "loss": 5.7828,
      "mean_token_accuracy": 0.1943471163511276,
      "num_tokens": 34519910.0,
      "step": 13615
    },
    {
      "entropy": 6.214790201187133,
      "epoch": 1.5718407386035778,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004763255432970144,
      "loss": 5.8279,
      "mean_token_accuracy": 0.19143639057874678,
      "num_tokens": 34534832.0,
      "step": 13620
    },
    {
      "entropy": 6.256109666824341,
      "epoch": 1.5724177726485862,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047630711509376235,
      "loss": 5.8398,
      "mean_token_accuracy": 0.18735139071941376,
      "num_tokens": 34548559.0,
      "step": 13625
    },
    {
      "entropy": 6.218293476104736,
      "epoch": 1.572994806693595,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004762886801196455,
      "loss": 5.8675,
      "mean_token_accuracy": 0.1887466862797737,
      "num_tokens": 34561131.0,
      "step": 13630
    },
    {
      "entropy": 6.257176303863526,
      "epoch": 1.5735718407386035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047627023837528386,
      "loss": 5.8164,
      "mean_token_accuracy": 0.19079274833202362,
      "num_tokens": 34572863.0,
      "step": 13635
    },
    {
      "entropy": 6.188808870315552,
      "epoch": 1.5741488747836123,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047625178986129775,
      "loss": 5.6634,
      "mean_token_accuracy": 0.20015804767608641,
      "num_tokens": 34586277.0,
      "step": 13640
    },
    {
      "entropy": 6.137176370620727,
      "epoch": 1.5747259088286207,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004762333345783078,
      "loss": 5.7519,
      "mean_token_accuracy": 0.18951563835144042,
      "num_tokens": 34598112.0,
      "step": 13645
    },
    {
      "entropy": 6.160854482650757,
      "epoch": 1.5753029428736296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047621487252693436,
      "loss": 5.7643,
      "mean_token_accuracy": 0.19611763060092927,
      "num_tokens": 34610569.0,
      "step": 13650
    },
    {
      "entropy": 6.267647933959961,
      "epoch": 1.5758799769186382,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047619640370779876,
      "loss": 5.8402,
      "mean_token_accuracy": 0.19009443670511245,
      "num_tokens": 34623148.0,
      "step": 13655
    },
    {
      "entropy": 6.224954843521118,
      "epoch": 1.5764570109636469,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047617792812152207,
      "loss": 5.8215,
      "mean_token_accuracy": 0.189132522046566,
      "num_tokens": 34635260.0,
      "step": 13660
    },
    {
      "entropy": 6.095504665374756,
      "epoch": 1.5770340450086555,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004761594457687256,
      "loss": 5.7163,
      "mean_token_accuracy": 0.19864651560783386,
      "num_tokens": 34648717.0,
      "step": 13665
    },
    {
      "entropy": 6.228409051895142,
      "epoch": 1.5776110790536642,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004761409566500313,
      "loss": 5.7931,
      "mean_token_accuracy": 0.19123381078243257,
      "num_tokens": 34662442.0,
      "step": 13670
    },
    {
      "entropy": 6.279564237594604,
      "epoch": 1.5781881130986728,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047612246076606066,
      "loss": 5.7925,
      "mean_token_accuracy": 0.19206460267305375,
      "num_tokens": 34675213.0,
      "step": 13675
    },
    {
      "entropy": 6.234532403945923,
      "epoch": 1.5787651471436814,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047610395811743594,
      "loss": 5.7794,
      "mean_token_accuracy": 0.18979280889034272,
      "num_tokens": 34687959.0,
      "step": 13680
    },
    {
      "entropy": 6.146605682373047,
      "epoch": 1.5793421811886903,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047608544870477956,
      "loss": 5.7431,
      "mean_token_accuracy": 0.19873310774564742,
      "num_tokens": 34700242.0,
      "step": 13685
    },
    {
      "entropy": 6.130990791320801,
      "epoch": 1.5799192152336987,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047606693252871395,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19941075146198273,
      "num_tokens": 34713245.0,
      "step": 13690
    },
    {
      "entropy": 6.2891425609588625,
      "epoch": 1.5804962492787076,
      "grad_norm": 0.875,
      "learning_rate": 0.000476048409589862,
      "loss": 5.8846,
      "mean_token_accuracy": 0.18503808081150055,
      "num_tokens": 34725248.0,
      "step": 13695
    },
    {
      "entropy": 6.352126693725586,
      "epoch": 1.581073283323716,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004760298798888466,
      "loss": 5.9045,
      "mean_token_accuracy": 0.1809140607714653,
      "num_tokens": 34737801.0,
      "step": 13700
    },
    {
      "entropy": 6.226459789276123,
      "epoch": 1.5816503173687249,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004760113434262911,
      "loss": 5.8559,
      "mean_token_accuracy": 0.18107242584228517,
      "num_tokens": 34749504.0,
      "step": 13705
    },
    {
      "entropy": 6.220118379592895,
      "epoch": 1.5822273514137333,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047599280020281894,
      "loss": 5.8653,
      "mean_token_accuracy": 0.18708096593618392,
      "num_tokens": 34761238.0,
      "step": 13710
    },
    {
      "entropy": 6.298572683334351,
      "epoch": 1.5828043854587421,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047597425021905364,
      "loss": 5.8571,
      "mean_token_accuracy": 0.18732394725084306,
      "num_tokens": 34773662.0,
      "step": 13715
    },
    {
      "entropy": 6.256297302246094,
      "epoch": 1.5833814195037508,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004759556934756194,
      "loss": 5.8504,
      "mean_token_accuracy": 0.1847571462392807,
      "num_tokens": 34786774.0,
      "step": 13720
    },
    {
      "entropy": 6.172251987457275,
      "epoch": 1.5839584535487594,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047593712997314017,
      "loss": 5.7956,
      "mean_token_accuracy": 0.1935566782951355,
      "num_tokens": 34798169.0,
      "step": 13725
    },
    {
      "entropy": 6.186849927902221,
      "epoch": 1.584535487593768,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047591855971224035,
      "loss": 5.7539,
      "mean_token_accuracy": 0.19370856285095214,
      "num_tokens": 34809870.0,
      "step": 13730
    },
    {
      "entropy": 6.23154649734497,
      "epoch": 1.5851125216387767,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004758999826935446,
      "loss": 5.9241,
      "mean_token_accuracy": 0.18438879549503326,
      "num_tokens": 34821684.0,
      "step": 13735
    },
    {
      "entropy": 6.2451917171478275,
      "epoch": 1.5856895556837853,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004758813989176778,
      "loss": 5.8108,
      "mean_token_accuracy": 0.19293127506971358,
      "num_tokens": 34834905.0,
      "step": 13740
    },
    {
      "entropy": 6.206116104125977,
      "epoch": 1.586266589728794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047586280838526483,
      "loss": 5.7463,
      "mean_token_accuracy": 0.1910412311553955,
      "num_tokens": 34846797.0,
      "step": 13745
    },
    {
      "entropy": 6.253277254104614,
      "epoch": 1.5868436237738026,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004758442110969312,
      "loss": 5.8052,
      "mean_token_accuracy": 0.189196415245533,
      "num_tokens": 34859261.0,
      "step": 13750
    },
    {
      "entropy": 6.238417482376098,
      "epoch": 1.5874206578188113,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004758256070533022,
      "loss": 5.8556,
      "mean_token_accuracy": 0.18680881410837175,
      "num_tokens": 34871775.0,
      "step": 13755
    },
    {
      "entropy": 6.188628435134888,
      "epoch": 1.5879976918638201,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004758069962550037,
      "loss": 5.734,
      "mean_token_accuracy": 0.190068382024765,
      "num_tokens": 34884461.0,
      "step": 13760
    },
    {
      "entropy": 6.2870580673217775,
      "epoch": 1.5885747259088285,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047578837870266156,
      "loss": 5.8451,
      "mean_token_accuracy": 0.18817704170942307,
      "num_tokens": 34896772.0,
      "step": 13765
    },
    {
      "entropy": 6.2610846042633055,
      "epoch": 1.5891517599538374,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047576975439690206,
      "loss": 5.8662,
      "mean_token_accuracy": 0.17997754812240602,
      "num_tokens": 34909914.0,
      "step": 13770
    },
    {
      "entropy": 6.238013744354248,
      "epoch": 1.5897287939988458,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047575112333835164,
      "loss": 5.8365,
      "mean_token_accuracy": 0.19200855791568755,
      "num_tokens": 34922298.0,
      "step": 13775
    },
    {
      "entropy": 6.312156915664673,
      "epoch": 1.5903058280438547,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047573248552763684,
      "loss": 5.9469,
      "mean_token_accuracy": 0.18925797194242477,
      "num_tokens": 34935159.0,
      "step": 13780
    },
    {
      "entropy": 6.16183557510376,
      "epoch": 1.590882862088863,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047571384096538474,
      "loss": 5.6914,
      "mean_token_accuracy": 0.19654348492622375,
      "num_tokens": 34948347.0,
      "step": 13785
    },
    {
      "entropy": 6.2010454654693605,
      "epoch": 1.591459896133872,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004756951896522222,
      "loss": 5.8398,
      "mean_token_accuracy": 0.1894977867603302,
      "num_tokens": 34961489.0,
      "step": 13790
    },
    {
      "entropy": 6.36897292137146,
      "epoch": 1.5920369301788806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004756765315887766,
      "loss": 5.9546,
      "mean_token_accuracy": 0.17735608518123627,
      "num_tokens": 34974458.0,
      "step": 13795
    },
    {
      "entropy": 6.291191816329956,
      "epoch": 1.5926139642238892,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004756578667756756,
      "loss": 5.8072,
      "mean_token_accuracy": 0.19481315165758134,
      "num_tokens": 34988498.0,
      "step": 13800
    },
    {
      "entropy": 6.061761283874512,
      "epoch": 1.5931909982688979,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004756391952135469,
      "loss": 5.6165,
      "mean_token_accuracy": 0.2175813063979149,
      "num_tokens": 35002101.0,
      "step": 13805
    },
    {
      "entropy": 6.250816774368286,
      "epoch": 1.5937680323139065,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047562051690301855,
      "loss": 5.9095,
      "mean_token_accuracy": 0.18890601694583892,
      "num_tokens": 35014079.0,
      "step": 13810
    },
    {
      "entropy": 6.217612075805664,
      "epoch": 1.5943450663589152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047560183184471873,
      "loss": 5.7639,
      "mean_token_accuracy": 0.1935321345925331,
      "num_tokens": 35027455.0,
      "step": 13815
    },
    {
      "entropy": 6.152585363388061,
      "epoch": 1.5949221004039238,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000475583140039276,
      "loss": 5.7595,
      "mean_token_accuracy": 0.19391000866889954,
      "num_tokens": 35040699.0,
      "step": 13820
    },
    {
      "entropy": 6.280860185623169,
      "epoch": 1.5954991344489324,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047556444148731897,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1840968072414398,
      "num_tokens": 35052758.0,
      "step": 13825
    },
    {
      "entropy": 6.237641620635986,
      "epoch": 1.596076168493941,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004755457361894766,
      "loss": 5.8522,
      "mean_token_accuracy": 0.186813285946846,
      "num_tokens": 35065393.0,
      "step": 13830
    },
    {
      "entropy": 6.265924692153931,
      "epoch": 1.59665320253895,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047552702414637793,
      "loss": 5.8551,
      "mean_token_accuracy": 0.1826876163482666,
      "num_tokens": 35077708.0,
      "step": 13835
    },
    {
      "entropy": 6.267186880111694,
      "epoch": 1.5972302365839584,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047550830535865245,
      "loss": 5.8747,
      "mean_token_accuracy": 0.18362477272748948,
      "num_tokens": 35088870.0,
      "step": 13840
    },
    {
      "entropy": 6.220412921905518,
      "epoch": 1.5978072706289672,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047548957982692974,
      "loss": 5.7457,
      "mean_token_accuracy": 0.19290478825569152,
      "num_tokens": 35100128.0,
      "step": 13845
    },
    {
      "entropy": 6.251234769821167,
      "epoch": 1.5983843046739756,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004754708475518396,
      "loss": 5.9219,
      "mean_token_accuracy": 0.19032201617956163,
      "num_tokens": 35113539.0,
      "step": 13850
    },
    {
      "entropy": 6.300593900680542,
      "epoch": 1.5989613387189845,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047545210853401214,
      "loss": 5.8746,
      "mean_token_accuracy": 0.18361787647008895,
      "num_tokens": 35126897.0,
      "step": 13855
    },
    {
      "entropy": 6.218615531921387,
      "epoch": 1.599538372763993,
      "grad_norm": 0.875,
      "learning_rate": 0.00047543336277407753,
      "loss": 5.823,
      "mean_token_accuracy": 0.1940055415034294,
      "num_tokens": 35140197.0,
      "step": 13860
    },
    {
      "entropy": 6.249961090087891,
      "epoch": 1.6001154068090018,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047541461027266624,
      "loss": 5.8053,
      "mean_token_accuracy": 0.19152496308088302,
      "num_tokens": 35151928.0,
      "step": 13865
    },
    {
      "entropy": 6.270364475250244,
      "epoch": 1.6006924408540104,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004753958510304091,
      "loss": 5.8559,
      "mean_token_accuracy": 0.18470062911510468,
      "num_tokens": 35165045.0,
      "step": 13870
    },
    {
      "entropy": 6.238029813766479,
      "epoch": 1.601269474899019,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047537708504793714,
      "loss": 5.756,
      "mean_token_accuracy": 0.19575304985046388,
      "num_tokens": 35177631.0,
      "step": 13875
    },
    {
      "entropy": 6.1533914566040036,
      "epoch": 1.6018465089440277,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047535831232588146,
      "loss": 5.8082,
      "mean_token_accuracy": 0.18768918812274932,
      "num_tokens": 35190166.0,
      "step": 13880
    },
    {
      "entropy": 6.260870742797851,
      "epoch": 1.6024235429890363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047533953286487345,
      "loss": 5.8842,
      "mean_token_accuracy": 0.18845939338207246,
      "num_tokens": 35202604.0,
      "step": 13885
    },
    {
      "entropy": 6.179200315475464,
      "epoch": 1.603000577034045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004753207466655447,
      "loss": 5.6947,
      "mean_token_accuracy": 0.19846705645322799,
      "num_tokens": 35214686.0,
      "step": 13890
    },
    {
      "entropy": 6.258864402770996,
      "epoch": 1.6035776110790536,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004753019537285273,
      "loss": 5.8795,
      "mean_token_accuracy": 0.18916985690593718,
      "num_tokens": 35226230.0,
      "step": 13895
    },
    {
      "entropy": 6.264652872085572,
      "epoch": 1.6041546451240625,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047528315405445296,
      "loss": 5.8239,
      "mean_token_accuracy": 0.18555429130792617,
      "num_tokens": 35240273.0,
      "step": 13900
    },
    {
      "entropy": 6.24879641532898,
      "epoch": 1.604731679169071,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047526434764395435,
      "loss": 5.8946,
      "mean_token_accuracy": 0.18154291957616805,
      "num_tokens": 35253095.0,
      "step": 13905
    },
    {
      "entropy": 6.250284671783447,
      "epoch": 1.6053087132140798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047524553449766386,
      "loss": 5.7783,
      "mean_token_accuracy": 0.19508038759231566,
      "num_tokens": 35265770.0,
      "step": 13910
    },
    {
      "entropy": 6.175844049453735,
      "epoch": 1.6058857472590882,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047522671461621433,
      "loss": 5.8404,
      "mean_token_accuracy": 0.194582699239254,
      "num_tokens": 35278800.0,
      "step": 13915
    },
    {
      "entropy": 6.268348455429077,
      "epoch": 1.606462781304097,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004752078880002386,
      "loss": 5.7843,
      "mean_token_accuracy": 0.1936597228050232,
      "num_tokens": 35290423.0,
      "step": 13920
    },
    {
      "entropy": 6.252221393585205,
      "epoch": 1.6070398153491054,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047518905465037005,
      "loss": 5.8562,
      "mean_token_accuracy": 0.1911500036716461,
      "num_tokens": 35303679.0,
      "step": 13925
    },
    {
      "entropy": 6.257269477844238,
      "epoch": 1.6076168493941143,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047517021456724214,
      "loss": 5.917,
      "mean_token_accuracy": 0.18437270671129227,
      "num_tokens": 35316900.0,
      "step": 13930
    },
    {
      "entropy": 6.216302490234375,
      "epoch": 1.608193883439123,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047515136775148843,
      "loss": 5.8141,
      "mean_token_accuracy": 0.19048181772232056,
      "num_tokens": 35329952.0,
      "step": 13935
    },
    {
      "entropy": 6.2883673191070555,
      "epoch": 1.6087709174841316,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004751325142037429,
      "loss": 5.88,
      "mean_token_accuracy": 0.18210436254739762,
      "num_tokens": 35342107.0,
      "step": 13940
    },
    {
      "entropy": 6.247175645828247,
      "epoch": 1.6093479515291402,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047511365392463974,
      "loss": 5.7866,
      "mean_token_accuracy": 0.19133169800043107,
      "num_tokens": 35354996.0,
      "step": 13945
    },
    {
      "entropy": 6.170210552215576,
      "epoch": 1.6099249855741489,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047509478691481317,
      "loss": 5.8455,
      "mean_token_accuracy": 0.1855359211564064,
      "num_tokens": 35367332.0,
      "step": 13950
    },
    {
      "entropy": 6.251528263092041,
      "epoch": 1.6105020196191575,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047507591317489783,
      "loss": 5.783,
      "mean_token_accuracy": 0.19406746029853822,
      "num_tokens": 35379469.0,
      "step": 13955
    },
    {
      "entropy": 6.236357975006103,
      "epoch": 1.6110790536641661,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004750570327055286,
      "loss": 5.7362,
      "mean_token_accuracy": 0.19823227673768998,
      "num_tokens": 35392579.0,
      "step": 13960
    },
    {
      "entropy": 6.226956510543824,
      "epoch": 1.6116560877091748,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047503814550734034,
      "loss": 5.846,
      "mean_token_accuracy": 0.18353219628334044,
      "num_tokens": 35405421.0,
      "step": 13965
    },
    {
      "entropy": 6.283872604370117,
      "epoch": 1.6122331217541834,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004750192515809685,
      "loss": 5.885,
      "mean_token_accuracy": 0.18996184766292573,
      "num_tokens": 35418046.0,
      "step": 13970
    },
    {
      "entropy": 6.306869411468506,
      "epoch": 1.6128101557991923,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047500035092704843,
      "loss": 5.8078,
      "mean_token_accuracy": 0.19264616668224335,
      "num_tokens": 35430346.0,
      "step": 13975
    },
    {
      "entropy": 6.238875246047973,
      "epoch": 1.6133871898442007,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004749814435462159,
      "loss": 5.8271,
      "mean_token_accuracy": 0.18097743541002273,
      "num_tokens": 35443607.0,
      "step": 13980
    },
    {
      "entropy": 6.252057313919067,
      "epoch": 1.6139642238892096,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004749625294391069,
      "loss": 5.7894,
      "mean_token_accuracy": 0.1881336197257042,
      "num_tokens": 35455890.0,
      "step": 13985
    },
    {
      "entropy": 6.181204223632813,
      "epoch": 1.614541257934218,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047494360860635755,
      "loss": 5.825,
      "mean_token_accuracy": 0.1967889115214348,
      "num_tokens": 35470340.0,
      "step": 13990
    },
    {
      "entropy": 6.2516380786895756,
      "epoch": 1.6151182919792268,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004749246810486042,
      "loss": 5.7854,
      "mean_token_accuracy": 0.18956567198038102,
      "num_tokens": 35482827.0,
      "step": 13995
    },
    {
      "entropy": 6.311751413345337,
      "epoch": 1.6156953260242353,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004749057467664836,
      "loss": 5.894,
      "mean_token_accuracy": 0.18534394949674607,
      "num_tokens": 35494289.0,
      "step": 14000
    },
    {
      "entropy": 6.194520664215088,
      "epoch": 1.6162723600692441,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047488680576063247,
      "loss": 5.7588,
      "mean_token_accuracy": 0.187918958067894,
      "num_tokens": 35506541.0,
      "step": 14005
    },
    {
      "entropy": 6.1610032081604,
      "epoch": 1.6168493941142528,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004748678580316878,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1977929577231407,
      "num_tokens": 35519628.0,
      "step": 14010
    },
    {
      "entropy": 6.2593084335327145,
      "epoch": 1.6174264281592614,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047484890358028714,
      "loss": 5.7356,
      "mean_token_accuracy": 0.1913806453347206,
      "num_tokens": 35532525.0,
      "step": 14015
    },
    {
      "entropy": 6.26611180305481,
      "epoch": 1.61800346220427,
      "grad_norm": 0.875,
      "learning_rate": 0.0004748299424070678,
      "loss": 5.848,
      "mean_token_accuracy": 0.1889537274837494,
      "num_tokens": 35544046.0,
      "step": 14020
    },
    {
      "entropy": 6.246851444244385,
      "epoch": 1.6185804962492787,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004748109745126677,
      "loss": 5.8516,
      "mean_token_accuracy": 0.1884143978357315,
      "num_tokens": 35557784.0,
      "step": 14025
    },
    {
      "entropy": 6.304057216644287,
      "epoch": 1.6191575302942873,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047479199989772464,
      "loss": 5.8835,
      "mean_token_accuracy": 0.18515651673078537,
      "num_tokens": 35571472.0,
      "step": 14030
    },
    {
      "entropy": 6.156277990341186,
      "epoch": 1.619734564339296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000474773018562877,
      "loss": 5.6978,
      "mean_token_accuracy": 0.21420663744211196,
      "num_tokens": 35584992.0,
      "step": 14035
    },
    {
      "entropy": 6.183495092391968,
      "epoch": 1.6203115983843048,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047475403050876297,
      "loss": 5.7437,
      "mean_token_accuracy": 0.192563296854496,
      "num_tokens": 35596643.0,
      "step": 14040
    },
    {
      "entropy": 6.272517585754395,
      "epoch": 1.6208886324293132,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004747350357360214,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1842023178935051,
      "num_tokens": 35608796.0,
      "step": 14045
    },
    {
      "entropy": 6.187342023849487,
      "epoch": 1.621465666474322,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004747160342452912,
      "loss": 5.7593,
      "mean_token_accuracy": 0.19461818784475327,
      "num_tokens": 35620008.0,
      "step": 14050
    },
    {
      "entropy": 6.247169780731201,
      "epoch": 1.6220427005193305,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047469702603721134,
      "loss": 5.7945,
      "mean_token_accuracy": 0.18930141627788544,
      "num_tokens": 35632751.0,
      "step": 14055
    },
    {
      "entropy": 6.16063871383667,
      "epoch": 1.6226197345643394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004746780111124212,
      "loss": 5.7768,
      "mean_token_accuracy": 0.18984494656324385,
      "num_tokens": 35645031.0,
      "step": 14060
    },
    {
      "entropy": 6.227063131332398,
      "epoch": 1.6231967686093478,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047465898947156033,
      "loss": 5.8048,
      "mean_token_accuracy": 0.18854637444019318,
      "num_tokens": 35656731.0,
      "step": 14065
    },
    {
      "entropy": 6.256523561477661,
      "epoch": 1.6237738026543567,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047463996111526854,
      "loss": 5.7971,
      "mean_token_accuracy": 0.19205496311187745,
      "num_tokens": 35668371.0,
      "step": 14070
    },
    {
      "entropy": 6.239273166656494,
      "epoch": 1.6243508366993653,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047462092604418576,
      "loss": 5.8668,
      "mean_token_accuracy": 0.18932655900716783,
      "num_tokens": 35681339.0,
      "step": 14075
    },
    {
      "entropy": 6.272698259353637,
      "epoch": 1.624927870744374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047460188425895236,
      "loss": 5.8245,
      "mean_token_accuracy": 0.19035169631242752,
      "num_tokens": 35693775.0,
      "step": 14080
    },
    {
      "entropy": 6.156334304809571,
      "epoch": 1.6255049047893826,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047458283576020874,
      "loss": 5.7243,
      "mean_token_accuracy": 0.20742534846067429,
      "num_tokens": 35705707.0,
      "step": 14085
    },
    {
      "entropy": 6.1782386302948,
      "epoch": 1.6260819388343912,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047456378054859554,
      "loss": 5.7646,
      "mean_token_accuracy": 0.19366937428712844,
      "num_tokens": 35717793.0,
      "step": 14090
    },
    {
      "entropy": 6.189295673370362,
      "epoch": 1.6266589728793999,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047454471862475375,
      "loss": 5.782,
      "mean_token_accuracy": 0.1933282122015953,
      "num_tokens": 35730762.0,
      "step": 14095
    },
    {
      "entropy": 6.243408155441284,
      "epoch": 1.6272360069244085,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047452564998932446,
      "loss": 5.7814,
      "mean_token_accuracy": 0.18916840106248856,
      "num_tokens": 35742852.0,
      "step": 14100
    },
    {
      "entropy": 6.260479164123535,
      "epoch": 1.6278130409694171,
      "grad_norm": 0.90625,
      "learning_rate": 0.000474506574642949,
      "loss": 5.8689,
      "mean_token_accuracy": 0.18073659837245942,
      "num_tokens": 35756416.0,
      "step": 14105
    },
    {
      "entropy": 6.311232948303223,
      "epoch": 1.6283900750144258,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000474487492586269,
      "loss": 5.8848,
      "mean_token_accuracy": 0.18322248458862306,
      "num_tokens": 35768347.0,
      "step": 14110
    },
    {
      "entropy": 6.241309022903442,
      "epoch": 1.6289671090594346,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004744684038199263,
      "loss": 5.7598,
      "mean_token_accuracy": 0.1898166537284851,
      "num_tokens": 35781228.0,
      "step": 14115
    },
    {
      "entropy": 6.246235752105713,
      "epoch": 1.629544143104443,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047444930834456293,
      "loss": 5.7535,
      "mean_token_accuracy": 0.19290558397769927,
      "num_tokens": 35793757.0,
      "step": 14120
    },
    {
      "entropy": 6.284105634689331,
      "epoch": 1.630121177149452,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004744302061608212,
      "loss": 5.857,
      "mean_token_accuracy": 0.18671861737966539,
      "num_tokens": 35805321.0,
      "step": 14125
    },
    {
      "entropy": 6.202886009216309,
      "epoch": 1.6306982111944603,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047441109726934345,
      "loss": 5.7806,
      "mean_token_accuracy": 0.19577531963586808,
      "num_tokens": 35817755.0,
      "step": 14130
    },
    {
      "entropy": 6.243549394607544,
      "epoch": 1.6312752452394692,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00047439198167077256,
      "loss": 5.801,
      "mean_token_accuracy": 0.18682890981435776,
      "num_tokens": 35831856.0,
      "step": 14135
    },
    {
      "entropy": 6.210424566268921,
      "epoch": 1.6318522792844776,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004743728593657514,
      "loss": 5.8417,
      "mean_token_accuracy": 0.193044576048851,
      "num_tokens": 35844613.0,
      "step": 14140
    },
    {
      "entropy": 6.204087018966675,
      "epoch": 1.6324293133294865,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004743537303549231,
      "loss": 5.7498,
      "mean_token_accuracy": 0.19214850217103957,
      "num_tokens": 35856928.0,
      "step": 14145
    },
    {
      "entropy": 6.223127174377441,
      "epoch": 1.6330063473744951,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004743345946389311,
      "loss": 5.7902,
      "mean_token_accuracy": 0.18633997887372972,
      "num_tokens": 35870161.0,
      "step": 14150
    },
    {
      "entropy": 6.2789257049560545,
      "epoch": 1.6335833814195038,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047431545221841907,
      "loss": 5.8125,
      "mean_token_accuracy": 0.1887281894683838,
      "num_tokens": 35883176.0,
      "step": 14155
    },
    {
      "entropy": 6.233300065994262,
      "epoch": 1.6341604154645124,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047429630309403086,
      "loss": 5.7233,
      "mean_token_accuracy": 0.1939747527241707,
      "num_tokens": 35895741.0,
      "step": 14160
    },
    {
      "entropy": 6.194824123382569,
      "epoch": 1.634737449509521,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047427714726641044,
      "loss": 5.8272,
      "mean_token_accuracy": 0.18817137479782103,
      "num_tokens": 35908921.0,
      "step": 14165
    },
    {
      "entropy": 6.21655306816101,
      "epoch": 1.6353144835545297,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047425798473620215,
      "loss": 5.7788,
      "mean_token_accuracy": 0.20220478922128676,
      "num_tokens": 35921388.0,
      "step": 14170
    },
    {
      "entropy": 6.264094018936158,
      "epoch": 1.6358915175995383,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004742388155040505,
      "loss": 5.8642,
      "mean_token_accuracy": 0.18901003301143646,
      "num_tokens": 35933880.0,
      "step": 14175
    },
    {
      "entropy": 6.193027305603027,
      "epoch": 1.6364685516445472,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047421963957060026,
      "loss": 5.7949,
      "mean_token_accuracy": 0.18940312117338182,
      "num_tokens": 35946690.0,
      "step": 14180
    },
    {
      "entropy": 6.278568363189697,
      "epoch": 1.6370455856895556,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004742004569364964,
      "loss": 5.82,
      "mean_token_accuracy": 0.18678973168134688,
      "num_tokens": 35958437.0,
      "step": 14185
    },
    {
      "entropy": 6.275190353393555,
      "epoch": 1.6376226197345645,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047418126760238416,
      "loss": 5.8064,
      "mean_token_accuracy": 0.1920170560479164,
      "num_tokens": 35969734.0,
      "step": 14190
    },
    {
      "entropy": 6.1726771831512455,
      "epoch": 1.6381996537795729,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047416207156890887,
      "loss": 5.7596,
      "mean_token_accuracy": 0.199510158598423,
      "num_tokens": 35983059.0,
      "step": 14195
    },
    {
      "entropy": 6.290701389312744,
      "epoch": 1.6387766878245817,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004741428688367164,
      "loss": 5.7767,
      "mean_token_accuracy": 0.19395927786827089,
      "num_tokens": 35996207.0,
      "step": 14200
    },
    {
      "entropy": 6.301733160018921,
      "epoch": 1.6393537218695902,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047412365940645225,
      "loss": 5.8451,
      "mean_token_accuracy": 0.18671062588691711,
      "num_tokens": 36008328.0,
      "step": 14205
    },
    {
      "entropy": 6.235314989089966,
      "epoch": 1.639930755914599,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047410444327876286,
      "loss": 5.9228,
      "mean_token_accuracy": 0.18510375171899796,
      "num_tokens": 36021384.0,
      "step": 14210
    },
    {
      "entropy": 6.177199363708496,
      "epoch": 1.6405077899596077,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047408522045429426,
      "loss": 5.8048,
      "mean_token_accuracy": 0.19497798681259154,
      "num_tokens": 36034101.0,
      "step": 14215
    },
    {
      "entropy": 6.273762655258179,
      "epoch": 1.6410848240046163,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004740659909336933,
      "loss": 5.7431,
      "mean_token_accuracy": 0.19643744677305222,
      "num_tokens": 36046504.0,
      "step": 14220
    },
    {
      "entropy": 6.180414533615112,
      "epoch": 1.641661858049625,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047404675471760655,
      "loss": 5.8465,
      "mean_token_accuracy": 0.18784733563661576,
      "num_tokens": 36058759.0,
      "step": 14225
    },
    {
      "entropy": 6.199416923522949,
      "epoch": 1.6422388920946336,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004740275118066811,
      "loss": 5.7901,
      "mean_token_accuracy": 0.199339796602726,
      "num_tokens": 36070238.0,
      "step": 14230
    },
    {
      "entropy": 6.257013034820557,
      "epoch": 1.6428159261396422,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047400826220156416,
      "loss": 5.8375,
      "mean_token_accuracy": 0.18937379866838455,
      "num_tokens": 36083543.0,
      "step": 14235
    },
    {
      "entropy": 6.21707501411438,
      "epoch": 1.6433929601846509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047398900590290313,
      "loss": 5.691,
      "mean_token_accuracy": 0.20031681507825852,
      "num_tokens": 36095250.0,
      "step": 14240
    },
    {
      "entropy": 6.152253866195679,
      "epoch": 1.6439699942296595,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047396974291134575,
      "loss": 5.8258,
      "mean_token_accuracy": 0.19633052349090577,
      "num_tokens": 36107919.0,
      "step": 14245
    },
    {
      "entropy": 6.280467796325683,
      "epoch": 1.6445470282746681,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047395047322754,
      "loss": 5.8382,
      "mean_token_accuracy": 0.188953498005867,
      "num_tokens": 36120284.0,
      "step": 14250
    },
    {
      "entropy": 6.313269329071045,
      "epoch": 1.645124062319677,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047393119685213374,
      "loss": 5.8753,
      "mean_token_accuracy": 0.19067458510398866,
      "num_tokens": 36133710.0,
      "step": 14255
    },
    {
      "entropy": 6.191392135620117,
      "epoch": 1.6457010963646854,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004739119137857756,
      "loss": 5.776,
      "mean_token_accuracy": 0.19545062482357026,
      "num_tokens": 36147042.0,
      "step": 14260
    },
    {
      "entropy": 6.267407989501953,
      "epoch": 1.6462781304096943,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047389262402911404,
      "loss": 5.85,
      "mean_token_accuracy": 0.18643911331892013,
      "num_tokens": 36159053.0,
      "step": 14265
    },
    {
      "entropy": 6.179013586044311,
      "epoch": 1.6468551644547027,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047387332758279784,
      "loss": 5.7376,
      "mean_token_accuracy": 0.19240613728761674,
      "num_tokens": 36171321.0,
      "step": 14270
    },
    {
      "entropy": 6.1364357471466064,
      "epoch": 1.6474321984997116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047385402444747606,
      "loss": 5.6961,
      "mean_token_accuracy": 0.2022164449095726,
      "num_tokens": 36183341.0,
      "step": 14275
    },
    {
      "entropy": 6.193008613586426,
      "epoch": 1.64800923254472,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047383471462379803,
      "loss": 5.7887,
      "mean_token_accuracy": 0.18813028782606125,
      "num_tokens": 36195407.0,
      "step": 14280
    },
    {
      "entropy": 6.255166959762573,
      "epoch": 1.6485862665897288,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000473815398112413,
      "loss": 5.8769,
      "mean_token_accuracy": 0.18305103927850724,
      "num_tokens": 36208736.0,
      "step": 14285
    },
    {
      "entropy": 6.245682239532471,
      "epoch": 1.6491633006347375,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004737960749139708,
      "loss": 5.7994,
      "mean_token_accuracy": 0.19571252465248107,
      "num_tokens": 36220022.0,
      "step": 14290
    },
    {
      "entropy": 6.217038345336914,
      "epoch": 1.6497403346797461,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004737767450291215,
      "loss": 5.7082,
      "mean_token_accuracy": 0.19272302240133285,
      "num_tokens": 36232105.0,
      "step": 14295
    },
    {
      "entropy": 6.2115623474121096,
      "epoch": 1.6503173687247548,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047375740845851506,
      "loss": 5.7982,
      "mean_token_accuracy": 0.18983737379312515,
      "num_tokens": 36243892.0,
      "step": 14300
    },
    {
      "entropy": 6.207330656051636,
      "epoch": 1.6508944027697634,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004737380652028019,
      "loss": 5.8862,
      "mean_token_accuracy": 0.18646888732910155,
      "num_tokens": 36255365.0,
      "step": 14305
    },
    {
      "entropy": 6.262782382965088,
      "epoch": 1.651471436814772,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047371871526263263,
      "loss": 5.8562,
      "mean_token_accuracy": 0.18902252465486527,
      "num_tokens": 36268238.0,
      "step": 14310
    },
    {
      "entropy": 6.308174896240234,
      "epoch": 1.6520484708597807,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004736993586386581,
      "loss": 5.8443,
      "mean_token_accuracy": 0.1899453431367874,
      "num_tokens": 36280294.0,
      "step": 14315
    },
    {
      "entropy": 6.1829758644104,
      "epoch": 1.6526255049047895,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047367999533152934,
      "loss": 5.748,
      "mean_token_accuracy": 0.1936400294303894,
      "num_tokens": 36293159.0,
      "step": 14320
    },
    {
      "entropy": 6.306614923477173,
      "epoch": 1.653202538949798,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047366062534189756,
      "loss": 5.8801,
      "mean_token_accuracy": 0.1799183577299118,
      "num_tokens": 36306492.0,
      "step": 14325
    },
    {
      "entropy": 6.239577579498291,
      "epoch": 1.6537795729948068,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047364124867041446,
      "loss": 5.8316,
      "mean_token_accuracy": 0.18400402516126632,
      "num_tokens": 36319020.0,
      "step": 14330
    },
    {
      "entropy": 6.232713222503662,
      "epoch": 1.6543566070398152,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047362186531773156,
      "loss": 5.8573,
      "mean_token_accuracy": 0.18261852860450745,
      "num_tokens": 36332163.0,
      "step": 14335
    },
    {
      "entropy": 6.2519432544708256,
      "epoch": 1.654933641084824,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004736024752845008,
      "loss": 5.7874,
      "mean_token_accuracy": 0.200864514708519,
      "num_tokens": 36345880.0,
      "step": 14340
    },
    {
      "entropy": 6.216885375976562,
      "epoch": 1.6555106751298325,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004735830785713746,
      "loss": 5.7899,
      "mean_token_accuracy": 0.19531371742486953,
      "num_tokens": 36358927.0,
      "step": 14345
    },
    {
      "entropy": 6.237912559509278,
      "epoch": 1.6560877091748414,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004735636751790051,
      "loss": 5.8425,
      "mean_token_accuracy": 0.18543946593999863,
      "num_tokens": 36371009.0,
      "step": 14350
    },
    {
      "entropy": 6.293756341934204,
      "epoch": 1.65666474321985,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000473544265108045,
      "loss": 5.8674,
      "mean_token_accuracy": 0.18664560765028,
      "num_tokens": 36385538.0,
      "step": 14355
    },
    {
      "entropy": 6.186305046081543,
      "epoch": 1.6572417772648587,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047352484835914716,
      "loss": 5.7534,
      "mean_token_accuracy": 0.1942903593182564,
      "num_tokens": 36398265.0,
      "step": 14360
    },
    {
      "entropy": 6.257086133956909,
      "epoch": 1.6578188113098673,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004735054249329647,
      "loss": 5.8124,
      "mean_token_accuracy": 0.19226132333278656,
      "num_tokens": 36410691.0,
      "step": 14365
    },
    {
      "entropy": 6.240606355667114,
      "epoch": 1.658395845354876,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047348599483015087,
      "loss": 5.7354,
      "mean_token_accuracy": 0.19234289824962617,
      "num_tokens": 36423531.0,
      "step": 14370
    },
    {
      "entropy": 6.198148632049561,
      "epoch": 1.6589728793998846,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047346655805135916,
      "loss": 5.8287,
      "mean_token_accuracy": 0.1938243865966797,
      "num_tokens": 36437429.0,
      "step": 14375
    },
    {
      "entropy": 6.174578905105591,
      "epoch": 1.6595499134448932,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004734471145972434,
      "loss": 5.7064,
      "mean_token_accuracy": 0.20009214729070662,
      "num_tokens": 36449264.0,
      "step": 14380
    },
    {
      "entropy": 6.240557289123535,
      "epoch": 1.6601269474899019,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047342766446845753,
      "loss": 5.7395,
      "mean_token_accuracy": 0.19767534285783767,
      "num_tokens": 36461132.0,
      "step": 14385
    },
    {
      "entropy": 6.241312170028687,
      "epoch": 1.6607039815349105,
      "grad_norm": 0.875,
      "learning_rate": 0.0004734082076656557,
      "loss": 5.8462,
      "mean_token_accuracy": 0.19019540250301362,
      "num_tokens": 36474195.0,
      "step": 14390
    },
    {
      "entropy": 6.267101287841797,
      "epoch": 1.6612810155799194,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047338874418949235,
      "loss": 5.9338,
      "mean_token_accuracy": 0.18116652816534043,
      "num_tokens": 36486197.0,
      "step": 14395
    },
    {
      "entropy": 6.192219495773315,
      "epoch": 1.6618580496249278,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047336927404062213,
      "loss": 5.75,
      "mean_token_accuracy": 0.2016308605670929,
      "num_tokens": 36499215.0,
      "step": 14400
    },
    {
      "entropy": 6.235541534423828,
      "epoch": 1.6624350836699366,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047334979721969995,
      "loss": 5.8688,
      "mean_token_accuracy": 0.18444155752658845,
      "num_tokens": 36511872.0,
      "step": 14405
    },
    {
      "entropy": 6.328991079330445,
      "epoch": 1.663012117714945,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004733303137273808,
      "loss": 5.8706,
      "mean_token_accuracy": 0.18904015123844148,
      "num_tokens": 36524313.0,
      "step": 14410
    },
    {
      "entropy": 6.14520034790039,
      "epoch": 1.663589151759954,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047331082356432015,
      "loss": 5.7646,
      "mean_token_accuracy": 0.19537544548511504,
      "num_tokens": 36538532.0,
      "step": 14415
    },
    {
      "entropy": 6.266261434555053,
      "epoch": 1.6641661858049623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004732913267311734,
      "loss": 5.8433,
      "mean_token_accuracy": 0.18475206047296525,
      "num_tokens": 36551117.0,
      "step": 14420
    },
    {
      "entropy": 6.259034633636475,
      "epoch": 1.6647432198499712,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004732718232285964,
      "loss": 5.7114,
      "mean_token_accuracy": 0.19416361451148986,
      "num_tokens": 36563704.0,
      "step": 14425
    },
    {
      "entropy": 6.199155187606811,
      "epoch": 1.6653202538949798,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047325231305724507,
      "loss": 5.8852,
      "mean_token_accuracy": 0.18580249845981597,
      "num_tokens": 36575952.0,
      "step": 14430
    },
    {
      "entropy": 6.184304141998291,
      "epoch": 1.6658972879399885,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004732327962177757,
      "loss": 5.723,
      "mean_token_accuracy": 0.19661466032266617,
      "num_tokens": 36590181.0,
      "step": 14435
    },
    {
      "entropy": 6.228618192672729,
      "epoch": 1.6664743219849971,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004732132727108447,
      "loss": 5.7775,
      "mean_token_accuracy": 0.1873861864209175,
      "num_tokens": 36601986.0,
      "step": 14440
    },
    {
      "entropy": 6.245586681365967,
      "epoch": 1.6670513560300058,
      "grad_norm": 1.0,
      "learning_rate": 0.00047319374253710874,
      "loss": 5.8536,
      "mean_token_accuracy": 0.18380175083875655,
      "num_tokens": 36614135.0,
      "step": 14445
    },
    {
      "entropy": 6.273624610900879,
      "epoch": 1.6676283900750144,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004731742056972247,
      "loss": 5.8843,
      "mean_token_accuracy": 0.1917415127158165,
      "num_tokens": 36625952.0,
      "step": 14450
    },
    {
      "entropy": 6.304927206039428,
      "epoch": 1.668205424120023,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004731546621918497,
      "loss": 5.8396,
      "mean_token_accuracy": 0.1913016125559807,
      "num_tokens": 36638945.0,
      "step": 14455
    },
    {
      "entropy": 6.261803722381591,
      "epoch": 1.668782458165032,
      "grad_norm": 0.875,
      "learning_rate": 0.000473135112021641,
      "loss": 5.8646,
      "mean_token_accuracy": 0.18723485767841339,
      "num_tokens": 36652550.0,
      "step": 14460
    },
    {
      "entropy": 6.253680467605591,
      "epoch": 1.6693594922100403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047311555518725617,
      "loss": 5.7915,
      "mean_token_accuracy": 0.19200919270515443,
      "num_tokens": 36665260.0,
      "step": 14465
    },
    {
      "entropy": 6.243175077438354,
      "epoch": 1.6699365262550492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047309599168935323,
      "loss": 5.7992,
      "mean_token_accuracy": 0.19736746847629547,
      "num_tokens": 36679275.0,
      "step": 14470
    },
    {
      "entropy": 6.158420419692993,
      "epoch": 1.6705135603000576,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047307642152858993,
      "loss": 5.7816,
      "mean_token_accuracy": 0.1975673481822014,
      "num_tokens": 36691849.0,
      "step": 14475
    },
    {
      "entropy": 6.247414636611938,
      "epoch": 1.6710905943450665,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047305684470562453,
      "loss": 5.8333,
      "mean_token_accuracy": 0.19576337784528733,
      "num_tokens": 36703618.0,
      "step": 14480
    },
    {
      "entropy": 6.204895687103272,
      "epoch": 1.6716676283900749,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004730372612211156,
      "loss": 5.7889,
      "mean_token_accuracy": 0.18904801458120346,
      "num_tokens": 36716332.0,
      "step": 14485
    },
    {
      "entropy": 6.106814336776734,
      "epoch": 1.6722446624350837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047301767107572174,
      "loss": 5.7217,
      "mean_token_accuracy": 0.19715412557125092,
      "num_tokens": 36729136.0,
      "step": 14490
    },
    {
      "entropy": 6.218730115890503,
      "epoch": 1.6728216964800924,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004729980742701018,
      "loss": 5.8093,
      "mean_token_accuracy": 0.19435659795999527,
      "num_tokens": 36743061.0,
      "step": 14495
    },
    {
      "entropy": 6.3000226497650145,
      "epoch": 1.673398730525101,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004729784708049151,
      "loss": 5.8048,
      "mean_token_accuracy": 0.19730500727891923,
      "num_tokens": 36756151.0,
      "step": 14500
    },
    {
      "entropy": 6.221712255477906,
      "epoch": 1.6739757645701097,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004729588606808209,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1904413565993309,
      "num_tokens": 36768008.0,
      "step": 14505
    },
    {
      "entropy": 6.1381079196929935,
      "epoch": 1.6745527986151183,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047293924389847864,
      "loss": 5.688,
      "mean_token_accuracy": 0.20036934912204743,
      "num_tokens": 36780694.0,
      "step": 14510
    },
    {
      "entropy": 6.138684177398682,
      "epoch": 1.675129832660127,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004729196204585483,
      "loss": 5.768,
      "mean_token_accuracy": 0.1954972490668297,
      "num_tokens": 36793040.0,
      "step": 14515
    },
    {
      "entropy": 6.2393676280975345,
      "epoch": 1.6757068667051356,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047289999036168983,
      "loss": 5.8554,
      "mean_token_accuracy": 0.18897452503442763,
      "num_tokens": 36806293.0,
      "step": 14520
    },
    {
      "entropy": 6.229583311080932,
      "epoch": 1.6762839007501442,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004728803536085634,
      "loss": 5.77,
      "mean_token_accuracy": 0.19155987054109574,
      "num_tokens": 36818717.0,
      "step": 14525
    },
    {
      "entropy": 6.291009759902954,
      "epoch": 1.6768609347951529,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047286071019982974,
      "loss": 5.7724,
      "mean_token_accuracy": 0.18907688707113265,
      "num_tokens": 36832984.0,
      "step": 14530
    },
    {
      "entropy": 6.219508409500122,
      "epoch": 1.6774379688401617,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047284106013614926,
      "loss": 5.7728,
      "mean_token_accuracy": 0.19110193699598313,
      "num_tokens": 36844088.0,
      "step": 14535
    },
    {
      "entropy": 6.278071403503418,
      "epoch": 1.6780150028851701,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000472821403418183,
      "loss": 5.8966,
      "mean_token_accuracy": 0.18688446432352065,
      "num_tokens": 36856846.0,
      "step": 14540
    },
    {
      "entropy": 6.246147108078003,
      "epoch": 1.678592036930179,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004728017400465921,
      "loss": 5.7115,
      "mean_token_accuracy": 0.20220181941986085,
      "num_tokens": 36870023.0,
      "step": 14545
    },
    {
      "entropy": 6.240050840377807,
      "epoch": 1.6791690709751874,
      "grad_norm": 1.0,
      "learning_rate": 0.00047278207002203796,
      "loss": 5.8401,
      "mean_token_accuracy": 0.1875194251537323,
      "num_tokens": 36883298.0,
      "step": 14550
    },
    {
      "entropy": 6.253798770904541,
      "epoch": 1.6797461050201963,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047276239334518216,
      "loss": 5.8246,
      "mean_token_accuracy": 0.18980590254068375,
      "num_tokens": 36894714.0,
      "step": 14555
    },
    {
      "entropy": 6.142205381393433,
      "epoch": 1.6803231390652047,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047274271001668646,
      "loss": 5.6506,
      "mean_token_accuracy": 0.20807466059923171,
      "num_tokens": 36908332.0,
      "step": 14560
    },
    {
      "entropy": 6.215235567092895,
      "epoch": 1.6809001731102136,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047272302003721286,
      "loss": 5.7853,
      "mean_token_accuracy": 0.19308234751224518,
      "num_tokens": 36921235.0,
      "step": 14565
    },
    {
      "entropy": 6.248942708969116,
      "epoch": 1.6814772071552222,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047270332340742377,
      "loss": 5.8692,
      "mean_token_accuracy": 0.18347352147102355,
      "num_tokens": 36934580.0,
      "step": 14570
    },
    {
      "entropy": 6.274139738082885,
      "epoch": 1.6820542412002308,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047268362012798157,
      "loss": 5.8301,
      "mean_token_accuracy": 0.19177044928073883,
      "num_tokens": 36946936.0,
      "step": 14575
    },
    {
      "entropy": 6.1152112007141115,
      "epoch": 1.6826312752452395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004726639101995491,
      "loss": 5.7279,
      "mean_token_accuracy": 0.1975792646408081,
      "num_tokens": 36960179.0,
      "step": 14580
    },
    {
      "entropy": 6.269066095352173,
      "epoch": 1.6832083092902481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047264419362278906,
      "loss": 5.8378,
      "mean_token_accuracy": 0.19260501563549043,
      "num_tokens": 36972739.0,
      "step": 14585
    },
    {
      "entropy": 6.161876106262207,
      "epoch": 1.6837853433352568,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047262447039836484,
      "loss": 5.7357,
      "mean_token_accuracy": 0.19397418200969696,
      "num_tokens": 36987314.0,
      "step": 14590
    },
    {
      "entropy": 6.19968957901001,
      "epoch": 1.6843623773802654,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047260474052693963,
      "loss": 5.8045,
      "mean_token_accuracy": 0.19401441514492035,
      "num_tokens": 36999050.0,
      "step": 14595
    },
    {
      "entropy": 6.233952379226684,
      "epoch": 1.684939411425274,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047258500400917724,
      "loss": 5.7618,
      "mean_token_accuracy": 0.19368332028388976,
      "num_tokens": 37011905.0,
      "step": 14600
    },
    {
      "entropy": 6.258266973495483,
      "epoch": 1.6855164454702827,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047256526084574137,
      "loss": 5.8457,
      "mean_token_accuracy": 0.1914002075791359,
      "num_tokens": 37023926.0,
      "step": 14605
    },
    {
      "entropy": 6.231887722015381,
      "epoch": 1.6860934795152915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047254551103729597,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18329426646232605,
      "num_tokens": 37037734.0,
      "step": 14610
    },
    {
      "entropy": 6.275544786453247,
      "epoch": 1.6866705135603,
      "grad_norm": 1.75,
      "learning_rate": 0.0004725257545845055,
      "loss": 5.8801,
      "mean_token_accuracy": 0.18693422228097917,
      "num_tokens": 37049689.0,
      "step": 14615
    },
    {
      "entropy": 6.337009334564209,
      "epoch": 1.6872475476053088,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047250599148803443,
      "loss": 5.9037,
      "mean_token_accuracy": 0.1846860945224762,
      "num_tokens": 37063999.0,
      "step": 14620
    },
    {
      "entropy": 6.170462989807129,
      "epoch": 1.6878245816503172,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047248622174854746,
      "loss": 5.7165,
      "mean_token_accuracy": 0.2008201763033867,
      "num_tokens": 37076855.0,
      "step": 14625
    },
    {
      "entropy": 6.269242763519287,
      "epoch": 1.688401615695326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004724664453667094,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1932467222213745,
      "num_tokens": 37089235.0,
      "step": 14630
    },
    {
      "entropy": 6.2112260341644285,
      "epoch": 1.6889786497403345,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004724466623431857,
      "loss": 5.8663,
      "mean_token_accuracy": 0.18881309628486634,
      "num_tokens": 37102011.0,
      "step": 14635
    },
    {
      "entropy": 6.1775877475738525,
      "epoch": 1.6895556837853434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004724268726786415,
      "loss": 5.725,
      "mean_token_accuracy": 0.20206808298826218,
      "num_tokens": 37115246.0,
      "step": 14640
    },
    {
      "entropy": 6.293998289108276,
      "epoch": 1.690132717830352,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004724070763737424,
      "loss": 5.9318,
      "mean_token_accuracy": 0.1836265429854393,
      "num_tokens": 37127705.0,
      "step": 14645
    },
    {
      "entropy": 6.243282175064087,
      "epoch": 1.6907097518753607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004723872734291545,
      "loss": 5.7469,
      "mean_token_accuracy": 0.19706328064203263,
      "num_tokens": 37141136.0,
      "step": 14650
    },
    {
      "entropy": 6.191778326034546,
      "epoch": 1.6912867859203693,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047236746384554364,
      "loss": 5.7419,
      "mean_token_accuracy": 0.19102190136909486,
      "num_tokens": 37154242.0,
      "step": 14655
    },
    {
      "entropy": 6.204723262786866,
      "epoch": 1.691863819965378,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004723476476235762,
      "loss": 5.7855,
      "mean_token_accuracy": 0.19343788027763367,
      "num_tokens": 37166254.0,
      "step": 14660
    },
    {
      "entropy": 6.293790245056153,
      "epoch": 1.6924408540103866,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004723278247639186,
      "loss": 5.8462,
      "mean_token_accuracy": 0.19213422238826752,
      "num_tokens": 37179579.0,
      "step": 14665
    },
    {
      "entropy": 6.169289064407349,
      "epoch": 1.6930178880553952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004723079952672377,
      "loss": 5.7499,
      "mean_token_accuracy": 0.19856800884008408,
      "num_tokens": 37191478.0,
      "step": 14670
    },
    {
      "entropy": 6.1597755432128904,
      "epoch": 1.693594922100404,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047228815913420035,
      "loss": 5.6494,
      "mean_token_accuracy": 0.20338428169488906,
      "num_tokens": 37203785.0,
      "step": 14675
    },
    {
      "entropy": 6.248442697525024,
      "epoch": 1.6941719561454125,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004722683163654738,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20013769268989562,
      "num_tokens": 37216298.0,
      "step": 14680
    },
    {
      "entropy": 6.255555438995361,
      "epoch": 1.6947489901904214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004722484669617254,
      "loss": 5.8432,
      "mean_token_accuracy": 0.1991363987326622,
      "num_tokens": 37228659.0,
      "step": 14685
    },
    {
      "entropy": 6.205869150161743,
      "epoch": 1.6953260242354298,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047222861092362277,
      "loss": 5.7892,
      "mean_token_accuracy": 0.19442098438739777,
      "num_tokens": 37240763.0,
      "step": 14690
    },
    {
      "entropy": 6.214226913452149,
      "epoch": 1.6959030582804386,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047220874825183387,
      "loss": 5.7332,
      "mean_token_accuracy": 0.1984926700592041,
      "num_tokens": 37252238.0,
      "step": 14695
    },
    {
      "entropy": 6.232885646820068,
      "epoch": 1.696480092325447,
      "grad_norm": 0.875,
      "learning_rate": 0.00047218887894702656,
      "loss": 5.8502,
      "mean_token_accuracy": 0.18963888734579087,
      "num_tokens": 37264475.0,
      "step": 14700
    },
    {
      "entropy": 6.24628963470459,
      "epoch": 1.697057126370456,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004721690030098693,
      "loss": 5.7872,
      "mean_token_accuracy": 0.1950768917798996,
      "num_tokens": 37277338.0,
      "step": 14705
    },
    {
      "entropy": 6.296155500411987,
      "epoch": 1.6976341604154646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004721491204410305,
      "loss": 5.8333,
      "mean_token_accuracy": 0.1903053790330887,
      "num_tokens": 37289985.0,
      "step": 14710
    },
    {
      "entropy": 6.302835416793823,
      "epoch": 1.6982111944604732,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047212923124117915,
      "loss": 5.8458,
      "mean_token_accuracy": 0.1939708486199379,
      "num_tokens": 37303177.0,
      "step": 14715
    },
    {
      "entropy": 6.256782960891724,
      "epoch": 1.6987882285054818,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004721093354109839,
      "loss": 5.7969,
      "mean_token_accuracy": 0.19007373452186585,
      "num_tokens": 37316662.0,
      "step": 14720
    },
    {
      "entropy": 6.218498659133911,
      "epoch": 1.6993652625504905,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047208943295111406,
      "loss": 5.8112,
      "mean_token_accuracy": 0.19404733330011367,
      "num_tokens": 37327782.0,
      "step": 14725
    },
    {
      "entropy": 6.195527029037476,
      "epoch": 1.6999422965954991,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047206952386223905,
      "loss": 5.7178,
      "mean_token_accuracy": 0.19844307899475097,
      "num_tokens": 37340650.0,
      "step": 14730
    },
    {
      "entropy": 6.222253227233887,
      "epoch": 1.7005193306405078,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004720496081450285,
      "loss": 5.8588,
      "mean_token_accuracy": 0.18523570001125336,
      "num_tokens": 37354330.0,
      "step": 14735
    },
    {
      "entropy": 6.240153646469116,
      "epoch": 1.7010963646855164,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047202968580015224,
      "loss": 5.8174,
      "mean_token_accuracy": 0.18898373246192932,
      "num_tokens": 37367447.0,
      "step": 14740
    },
    {
      "entropy": 6.219916915893554,
      "epoch": 1.701673398730525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047200975682828045,
      "loss": 5.7293,
      "mean_token_accuracy": 0.19999866485595702,
      "num_tokens": 37380171.0,
      "step": 14745
    },
    {
      "entropy": 6.246268177032471,
      "epoch": 1.702250432775534,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004719898212300832,
      "loss": 5.7811,
      "mean_token_accuracy": 0.19559714049100876,
      "num_tokens": 37393042.0,
      "step": 14750
    },
    {
      "entropy": 6.276305484771728,
      "epoch": 1.7028274668205423,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047196987900623134,
      "loss": 5.8472,
      "mean_token_accuracy": 0.19532237350940704,
      "num_tokens": 37406854.0,
      "step": 14755
    },
    {
      "entropy": 6.256501531600952,
      "epoch": 1.7034045008655512,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047194993015739527,
      "loss": 5.8441,
      "mean_token_accuracy": 0.1859874665737152,
      "num_tokens": 37419765.0,
      "step": 14760
    },
    {
      "entropy": 6.272218561172485,
      "epoch": 1.7039815349105596,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047192997468424624,
      "loss": 5.7792,
      "mean_token_accuracy": 0.19645111858844758,
      "num_tokens": 37432280.0,
      "step": 14765
    },
    {
      "entropy": 6.261264514923096,
      "epoch": 1.7045585689555685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719100125874553,
      "loss": 5.8773,
      "mean_token_accuracy": 0.18530086278915406,
      "num_tokens": 37444315.0,
      "step": 14770
    },
    {
      "entropy": 6.280199289321899,
      "epoch": 1.7051356030005769,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004718900438676939,
      "loss": 5.8648,
      "mean_token_accuracy": 0.18270065039396285,
      "num_tokens": 37457588.0,
      "step": 14775
    },
    {
      "entropy": 6.25987024307251,
      "epoch": 1.7057126370455857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004718700685256337,
      "loss": 5.7932,
      "mean_token_accuracy": 0.1880245476961136,
      "num_tokens": 37471740.0,
      "step": 14780
    },
    {
      "entropy": 6.27062954902649,
      "epoch": 1.7062896710905944,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004718500865619465,
      "loss": 5.7803,
      "mean_token_accuracy": 0.18788072764873504,
      "num_tokens": 37484835.0,
      "step": 14785
    },
    {
      "entropy": 6.214168357849121,
      "epoch": 1.706866705135603,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004718300979773044,
      "loss": 5.7985,
      "mean_token_accuracy": 0.1931379333138466,
      "num_tokens": 37497987.0,
      "step": 14790
    },
    {
      "entropy": 6.246798276901245,
      "epoch": 1.7074437391806117,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047181010277237977,
      "loss": 5.8126,
      "mean_token_accuracy": 0.1932763010263443,
      "num_tokens": 37510454.0,
      "step": 14795
    },
    {
      "entropy": 6.24987268447876,
      "epoch": 1.7080207732256203,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004717901009478451,
      "loss": 5.7475,
      "mean_token_accuracy": 0.19508094787597657,
      "num_tokens": 37523959.0,
      "step": 14800
    },
    {
      "entropy": 6.2551196098327635,
      "epoch": 1.708597807270629,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047177009250437313,
      "loss": 5.8734,
      "mean_token_accuracy": 0.18973658382892608,
      "num_tokens": 37538116.0,
      "step": 14805
    },
    {
      "entropy": 6.2460369110107425,
      "epoch": 1.7091748413156376,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047175007744263683,
      "loss": 5.7928,
      "mean_token_accuracy": 0.18800595551729202,
      "num_tokens": 37549940.0,
      "step": 14810
    },
    {
      "entropy": 6.232281970977783,
      "epoch": 1.7097518753606464,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047173005576330935,
      "loss": 5.8149,
      "mean_token_accuracy": 0.1916723757982254,
      "num_tokens": 37562854.0,
      "step": 14815
    },
    {
      "entropy": 6.1734044551849365,
      "epoch": 1.7103289094056549,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047171002746706424,
      "loss": 5.7026,
      "mean_token_accuracy": 0.19862491339445115,
      "num_tokens": 37575639.0,
      "step": 14820
    },
    {
      "entropy": 6.176781940460205,
      "epoch": 1.7109059434506637,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00047168999255457506,
      "loss": 5.7267,
      "mean_token_accuracy": 0.19745132178068162,
      "num_tokens": 37588555.0,
      "step": 14825
    },
    {
      "entropy": 6.273313760757446,
      "epoch": 1.7114829774956721,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047166995102651565,
      "loss": 5.8136,
      "mean_token_accuracy": 0.18935156017541885,
      "num_tokens": 37602636.0,
      "step": 14830
    },
    {
      "entropy": 6.172989082336426,
      "epoch": 1.712060011540681,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004716499028835601,
      "loss": 5.7633,
      "mean_token_accuracy": 0.1885058254003525,
      "num_tokens": 37616769.0,
      "step": 14835
    },
    {
      "entropy": 6.289084243774414,
      "epoch": 1.7126370455856894,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004716298481263828,
      "loss": 5.8143,
      "mean_token_accuracy": 0.1968541383743286,
      "num_tokens": 37628173.0,
      "step": 14840
    },
    {
      "entropy": 6.255649375915527,
      "epoch": 1.7132140796306983,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004716097867556583,
      "loss": 5.7463,
      "mean_token_accuracy": 0.19388801753520965,
      "num_tokens": 37640953.0,
      "step": 14845
    },
    {
      "entropy": 6.244218873977661,
      "epoch": 1.713791113675707,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047158971877206114,
      "loss": 5.8214,
      "mean_token_accuracy": 0.19193712770938873,
      "num_tokens": 37655091.0,
      "step": 14850
    },
    {
      "entropy": 6.258729600906372,
      "epoch": 1.7143681477207156,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004715696441762665,
      "loss": 5.8274,
      "mean_token_accuracy": 0.18968533873558044,
      "num_tokens": 37668120.0,
      "step": 14855
    },
    {
      "entropy": 6.2443231582641605,
      "epoch": 1.7149451817657242,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047154956296894954,
      "loss": 5.6906,
      "mean_token_accuracy": 0.1998714342713356,
      "num_tokens": 37679844.0,
      "step": 14860
    },
    {
      "entropy": 6.165535116195679,
      "epoch": 1.7155222158107328,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004715294751507856,
      "loss": 5.7947,
      "mean_token_accuracy": 0.19430916160345077,
      "num_tokens": 37692756.0,
      "step": 14865
    },
    {
      "entropy": 6.2394819259643555,
      "epoch": 1.7160992498557415,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004715093807224505,
      "loss": 5.8517,
      "mean_token_accuracy": 0.1919565126299858,
      "num_tokens": 37704525.0,
      "step": 14870
    },
    {
      "entropy": 6.305459356307983,
      "epoch": 1.71667628390075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047148927968462,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1915173888206482,
      "num_tokens": 37715980.0,
      "step": 14875
    },
    {
      "entropy": 6.1830058097839355,
      "epoch": 1.7172533179457588,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047146917203797,
      "loss": 5.7012,
      "mean_token_accuracy": 0.19431441873311997,
      "num_tokens": 37728448.0,
      "step": 14880
    },
    {
      "entropy": 6.2103590965271,
      "epoch": 1.7178303519907674,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004714490577831771,
      "loss": 5.756,
      "mean_token_accuracy": 0.19255241304636,
      "num_tokens": 37741798.0,
      "step": 14885
    },
    {
      "entropy": 6.179524230957031,
      "epoch": 1.7184073860357763,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004714289369209177,
      "loss": 5.7087,
      "mean_token_accuracy": 0.20174277275800706,
      "num_tokens": 37755056.0,
      "step": 14890
    },
    {
      "entropy": 6.289288568496704,
      "epoch": 1.7189844200807847,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047140880945186863,
      "loss": 5.8454,
      "mean_token_accuracy": 0.18469300866127014,
      "num_tokens": 37767554.0,
      "step": 14895
    },
    {
      "entropy": 6.242574977874756,
      "epoch": 1.7195614541257935,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047138867537670676,
      "loss": 5.8285,
      "mean_token_accuracy": 0.19243978559970856,
      "num_tokens": 37781430.0,
      "step": 14900
    },
    {
      "entropy": 6.148932027816772,
      "epoch": 1.720138488170802,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047136853469610933,
      "loss": 5.6881,
      "mean_token_accuracy": 0.19345901310443878,
      "num_tokens": 37793138.0,
      "step": 14905
    },
    {
      "entropy": 6.286306285858155,
      "epoch": 1.7207155222158108,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047134838741075383,
      "loss": 5.8592,
      "mean_token_accuracy": 0.18784959018230438,
      "num_tokens": 37806685.0,
      "step": 14910
    },
    {
      "entropy": 6.1418256759643555,
      "epoch": 1.7212925562608192,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047132823352131787,
      "loss": 5.7046,
      "mean_token_accuracy": 0.20198537707328795,
      "num_tokens": 37818571.0,
      "step": 14915
    },
    {
      "entropy": 6.231262302398681,
      "epoch": 1.721869590305828,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004713080730284793,
      "loss": 5.8265,
      "mean_token_accuracy": 0.18657076060771943,
      "num_tokens": 37831245.0,
      "step": 14920
    },
    {
      "entropy": 6.298978233337403,
      "epoch": 1.7224466243508367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047128790593291617,
      "loss": 5.8572,
      "mean_token_accuracy": 0.19251249134540557,
      "num_tokens": 37844140.0,
      "step": 14925
    },
    {
      "entropy": 6.259516429901123,
      "epoch": 1.7230236583958454,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047126773223530677,
      "loss": 5.8567,
      "mean_token_accuracy": 0.18552057445049286,
      "num_tokens": 37856783.0,
      "step": 14930
    },
    {
      "entropy": 6.253098678588867,
      "epoch": 1.723600692440854,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047124755193632975,
      "loss": 5.8347,
      "mean_token_accuracy": 0.19024232029914856,
      "num_tokens": 37871219.0,
      "step": 14935
    },
    {
      "entropy": 6.252159690856933,
      "epoch": 1.7241777264858626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047122736503666377,
      "loss": 5.8112,
      "mean_token_accuracy": 0.1987973064184189,
      "num_tokens": 37882510.0,
      "step": 14940
    },
    {
      "entropy": 6.252618741989136,
      "epoch": 1.7247547605308713,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004712071715369878,
      "loss": 5.8315,
      "mean_token_accuracy": 0.190506511926651,
      "num_tokens": 37894785.0,
      "step": 14945
    },
    {
      "entropy": 6.202209568023681,
      "epoch": 1.72533179457588,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004711869714379812,
      "loss": 5.7663,
      "mean_token_accuracy": 0.18978879004716873,
      "num_tokens": 37907598.0,
      "step": 14950
    },
    {
      "entropy": 6.219098663330078,
      "epoch": 1.7259088286208888,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004711667647403232,
      "loss": 5.7466,
      "mean_token_accuracy": 0.19176314026117325,
      "num_tokens": 37920399.0,
      "step": 14955
    },
    {
      "entropy": 6.254694271087646,
      "epoch": 1.7264858626658972,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047114655144469354,
      "loss": 5.8665,
      "mean_token_accuracy": 0.1902274712920189,
      "num_tokens": 37934157.0,
      "step": 14960
    },
    {
      "entropy": 6.231907939910888,
      "epoch": 1.727062896710906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047112633155177203,
      "loss": 5.7293,
      "mean_token_accuracy": 0.1917924240231514,
      "num_tokens": 37946058.0,
      "step": 14965
    },
    {
      "entropy": 6.261799097061157,
      "epoch": 1.7276399307559145,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004711061050622388,
      "loss": 5.8141,
      "mean_token_accuracy": 0.19253401607275009,
      "num_tokens": 37958549.0,
      "step": 14970
    },
    {
      "entropy": 6.2651549816131595,
      "epoch": 1.7282169648009233,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047108587197677404,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1908634677529335,
      "num_tokens": 37971633.0,
      "step": 14975
    },
    {
      "entropy": 6.200336694717407,
      "epoch": 1.7287939988459318,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047106563229605845,
      "loss": 5.7156,
      "mean_token_accuracy": 0.1946705773472786,
      "num_tokens": 37984048.0,
      "step": 14980
    },
    {
      "entropy": 6.253131628036499,
      "epoch": 1.7293710328909406,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047104538602077276,
      "loss": 5.8168,
      "mean_token_accuracy": 0.19120276272296904,
      "num_tokens": 37996092.0,
      "step": 14985
    },
    {
      "entropy": 6.171030569076538,
      "epoch": 1.7299480669359493,
      "grad_norm": 0.875,
      "learning_rate": 0.0004710251331515978,
      "loss": 5.7099,
      "mean_token_accuracy": 0.19812670946121216,
      "num_tokens": 38008769.0,
      "step": 14990
    },
    {
      "entropy": 6.20517406463623,
      "epoch": 1.730525100980958,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047100487368921485,
      "loss": 5.7088,
      "mean_token_accuracy": 0.19842482954263688,
      "num_tokens": 38021344.0,
      "step": 14995
    },
    {
      "entropy": 6.254121541976929,
      "epoch": 1.7311021350259665,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004709846076343055,
      "loss": 5.7906,
      "mean_token_accuracy": 0.19390783309936524,
      "num_tokens": 38033839.0,
      "step": 15000
    },
    {
      "epoch": 1.7311021350259665,
      "eval_entropy": 6.039237286782177,
      "eval_loss": 5.8590874671936035,
      "eval_mean_token_accuracy": 0.19710241953842605,
      "eval_num_tokens": 38033839.0,
      "eval_runtime": 17.6912,
      "eval_samples_per_second": 1590.398,
      "eval_steps_per_second": 198.8,
      "step": 15000
    },
    {
      "entropy": 6.274616241455078,
      "epoch": 1.7316791690709752,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00047096433498755103,
      "loss": 5.8728,
      "mean_token_accuracy": 0.18765036165714263,
      "num_tokens": 38047026.0,
      "step": 15005
    },
    {
      "entropy": 6.195619583129883,
      "epoch": 1.7322562031159838,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047094405574963364,
      "loss": 5.7561,
      "mean_token_accuracy": 0.19641678035259247,
      "num_tokens": 38059089.0,
      "step": 15010
    },
    {
      "entropy": 6.213481521606445,
      "epoch": 1.7328332371609925,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047092376992123516,
      "loss": 5.7759,
      "mean_token_accuracy": 0.18926770091056824,
      "num_tokens": 38072925.0,
      "step": 15015
    },
    {
      "entropy": 6.2407022476196286,
      "epoch": 1.733410271206001,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047090347750303803,
      "loss": 5.7848,
      "mean_token_accuracy": 0.19753599911928177,
      "num_tokens": 38084532.0,
      "step": 15020
    },
    {
      "entropy": 6.2293273448944095,
      "epoch": 1.7339873052510097,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004708831784957247,
      "loss": 5.7995,
      "mean_token_accuracy": 0.19780379980802537,
      "num_tokens": 38098024.0,
      "step": 15025
    },
    {
      "entropy": 6.234379434585572,
      "epoch": 1.7345643392960186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004708628728999781,
      "loss": 5.7841,
      "mean_token_accuracy": 0.19226982146501542,
      "num_tokens": 38110718.0,
      "step": 15030
    },
    {
      "entropy": 6.238504791259766,
      "epoch": 1.735141373341027,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004708425607164809,
      "loss": 5.8251,
      "mean_token_accuracy": 0.19152746200561524,
      "num_tokens": 38124807.0,
      "step": 15035
    },
    {
      "entropy": 6.253223180770874,
      "epoch": 1.7357184073860359,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004708222419459165,
      "loss": 5.7897,
      "mean_token_accuracy": 0.18876226842403412,
      "num_tokens": 38136660.0,
      "step": 15040
    },
    {
      "entropy": 6.255292129516602,
      "epoch": 1.7362954414310443,
      "grad_norm": 3.09375,
      "learning_rate": 0.0004708019165889683,
      "loss": 5.8053,
      "mean_token_accuracy": 0.19430315494537354,
      "num_tokens": 38148154.0,
      "step": 15045
    },
    {
      "entropy": 6.263941049575806,
      "epoch": 1.7368724754760532,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004707815846463199,
      "loss": 5.7217,
      "mean_token_accuracy": 0.19700276404619216,
      "num_tokens": 38161092.0,
      "step": 15050
    },
    {
      "entropy": 6.267111158370971,
      "epoch": 1.7374495095210616,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004707612461186552,
      "loss": 5.8275,
      "mean_token_accuracy": 0.195013527572155,
      "num_tokens": 38172349.0,
      "step": 15055
    },
    {
      "entropy": 6.21812334060669,
      "epoch": 1.7380265435660704,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047074090100665805,
      "loss": 5.8348,
      "mean_token_accuracy": 0.18741801530122756,
      "num_tokens": 38184266.0,
      "step": 15060
    },
    {
      "entropy": 6.12758092880249,
      "epoch": 1.738603577611079,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047072054931101306,
      "loss": 5.6588,
      "mean_token_accuracy": 0.207034695148468,
      "num_tokens": 38196865.0,
      "step": 15065
    },
    {
      "entropy": 6.222312688827515,
      "epoch": 1.7391806116560877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004707001910324046,
      "loss": 5.7574,
      "mean_token_accuracy": 0.1984534204006195,
      "num_tokens": 38208416.0,
      "step": 15070
    },
    {
      "entropy": 6.196582698822022,
      "epoch": 1.7397576457010964,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047067982617151737,
      "loss": 5.8039,
      "mean_token_accuracy": 0.18924115598201752,
      "num_tokens": 38220649.0,
      "step": 15075
    },
    {
      "entropy": 6.250541543960571,
      "epoch": 1.740334679746105,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004706594547290365,
      "loss": 5.8062,
      "mean_token_accuracy": 0.1893243283033371,
      "num_tokens": 38233538.0,
      "step": 15080
    },
    {
      "entropy": 6.167980098724366,
      "epoch": 1.7409117137911136,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047063907670564695,
      "loss": 5.6572,
      "mean_token_accuracy": 0.20257883369922638,
      "num_tokens": 38245603.0,
      "step": 15085
    },
    {
      "entropy": 6.261222982406617,
      "epoch": 1.7414887478361223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004706186921020342,
      "loss": 5.8654,
      "mean_token_accuracy": 0.1891934394836426,
      "num_tokens": 38258201.0,
      "step": 15090
    },
    {
      "entropy": 6.220455360412598,
      "epoch": 1.7420657818811311,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047059830091888407,
      "loss": 5.7186,
      "mean_token_accuracy": 0.20864225924015045,
      "num_tokens": 38270659.0,
      "step": 15095
    },
    {
      "entropy": 6.264389562606811,
      "epoch": 1.7426428159261396,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004705779031568821,
      "loss": 5.8002,
      "mean_token_accuracy": 0.18801535069942474,
      "num_tokens": 38283101.0,
      "step": 15100
    },
    {
      "entropy": 6.248580503463745,
      "epoch": 1.7432198499711484,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047055749881671463,
      "loss": 5.7024,
      "mean_token_accuracy": 0.2041386753320694,
      "num_tokens": 38296855.0,
      "step": 15105
    },
    {
      "entropy": 6.158930778503418,
      "epoch": 1.7437968840161568,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004705370878990677,
      "loss": 5.6676,
      "mean_token_accuracy": 0.19917973279953002,
      "num_tokens": 38308354.0,
      "step": 15110
    },
    {
      "entropy": 6.2030439376831055,
      "epoch": 1.7443739180611657,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047051667040462806,
      "loss": 5.8266,
      "mean_token_accuracy": 0.1888536125421524,
      "num_tokens": 38321239.0,
      "step": 15115
    },
    {
      "entropy": 6.2983709335327145,
      "epoch": 1.7449509521061741,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047049624633408224,
      "loss": 5.8494,
      "mean_token_accuracy": 0.18843162506818772,
      "num_tokens": 38333672.0,
      "step": 15120
    },
    {
      "entropy": 6.204380941390991,
      "epoch": 1.745527986151183,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047047581568811724,
      "loss": 5.7757,
      "mean_token_accuracy": 0.19435797184705733,
      "num_tokens": 38346809.0,
      "step": 15125
    },
    {
      "entropy": 6.238096714019775,
      "epoch": 1.7461050201961916,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047045537846742043,
      "loss": 5.7805,
      "mean_token_accuracy": 0.19013507068157195,
      "num_tokens": 38359129.0,
      "step": 15130
    },
    {
      "entropy": 6.327469682693481,
      "epoch": 1.7466820542412003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000470434934672679,
      "loss": 5.8997,
      "mean_token_accuracy": 0.18664243966341018,
      "num_tokens": 38371621.0,
      "step": 15135
    },
    {
      "entropy": 6.275482797622681,
      "epoch": 1.747259088286209,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047041448430458054,
      "loss": 5.8153,
      "mean_token_accuracy": 0.19149385243654252,
      "num_tokens": 38384579.0,
      "step": 15140
    },
    {
      "entropy": 6.336668300628662,
      "epoch": 1.7478361223312175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047039402736381305,
      "loss": 5.8803,
      "mean_token_accuracy": 0.184342160820961,
      "num_tokens": 38396765.0,
      "step": 15145
    },
    {
      "entropy": 6.228005743026733,
      "epoch": 1.7484131563762262,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004703735638510645,
      "loss": 5.7754,
      "mean_token_accuracy": 0.19603645503520967,
      "num_tokens": 38409778.0,
      "step": 15150
    },
    {
      "entropy": 6.248737001419068,
      "epoch": 1.7489901904212348,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004703530937670232,
      "loss": 5.7813,
      "mean_token_accuracy": 0.19069131314754487,
      "num_tokens": 38421729.0,
      "step": 15155
    },
    {
      "entropy": 6.283941459655762,
      "epoch": 1.7495672244662435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004703326171123776,
      "loss": 5.761,
      "mean_token_accuracy": 0.19319654405117034,
      "num_tokens": 38435136.0,
      "step": 15160
    },
    {
      "entropy": 6.26770167350769,
      "epoch": 1.750144258511252,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004703121338878164,
      "loss": 5.7946,
      "mean_token_accuracy": 0.1905953109264374,
      "num_tokens": 38447362.0,
      "step": 15165
    },
    {
      "entropy": 6.167471790313721,
      "epoch": 1.750721292556261,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004702916440940286,
      "loss": 5.6738,
      "mean_token_accuracy": 0.203914874792099,
      "num_tokens": 38459796.0,
      "step": 15170
    },
    {
      "entropy": 6.3102460384368895,
      "epoch": 1.7512983266012694,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004702711477317034,
      "loss": 5.9317,
      "mean_token_accuracy": 0.18021377474069594,
      "num_tokens": 38472536.0,
      "step": 15175
    },
    {
      "entropy": 6.29654369354248,
      "epoch": 1.7518753606462782,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004702506448015301,
      "loss": 5.8063,
      "mean_token_accuracy": 0.1903869166970253,
      "num_tokens": 38485107.0,
      "step": 15180
    },
    {
      "entropy": 6.249254560470581,
      "epoch": 1.7524523946912867,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047023013530419843,
      "loss": 5.8788,
      "mean_token_accuracy": 0.18370821624994277,
      "num_tokens": 38498560.0,
      "step": 15185
    },
    {
      "entropy": 6.193416690826416,
      "epoch": 1.7530294287362955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004702096192403981,
      "loss": 5.5768,
      "mean_token_accuracy": 0.20579371899366378,
      "num_tokens": 38510553.0,
      "step": 15190
    },
    {
      "entropy": 6.201163864135742,
      "epoch": 1.753606462781304,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004701890966108192,
      "loss": 5.7757,
      "mean_token_accuracy": 0.19423052072525024,
      "num_tokens": 38524191.0,
      "step": 15195
    },
    {
      "entropy": 6.211992692947388,
      "epoch": 1.7541834968263128,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000470168567416152,
      "loss": 5.8189,
      "mean_token_accuracy": 0.19442755430936814,
      "num_tokens": 38535654.0,
      "step": 15200
    },
    {
      "entropy": 6.2909040451049805,
      "epoch": 1.7547605308713214,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004701480316570869,
      "loss": 5.8674,
      "mean_token_accuracy": 0.18599323630332948,
      "num_tokens": 38548596.0,
      "step": 15205
    },
    {
      "entropy": 6.2079911708831785,
      "epoch": 1.75533756491633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004701274893343147,
      "loss": 5.7701,
      "mean_token_accuracy": 0.19283776432275773,
      "num_tokens": 38559852.0,
      "step": 15210
    },
    {
      "entropy": 6.124769067764282,
      "epoch": 1.7559145989613387,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047010694044852643,
      "loss": 5.7256,
      "mean_token_accuracy": 0.19432504922151567,
      "num_tokens": 38571773.0,
      "step": 15215
    },
    {
      "entropy": 6.263028335571289,
      "epoch": 1.7564916330063474,
      "grad_norm": 0.9375,
      "learning_rate": 0.000470086385000413,
      "loss": 5.7555,
      "mean_token_accuracy": 0.20026387870311738,
      "num_tokens": 38583898.0,
      "step": 15220
    },
    {
      "entropy": 6.229983615875244,
      "epoch": 1.757068667051356,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004700658229906661,
      "loss": 5.8219,
      "mean_token_accuracy": 0.18950065523386,
      "num_tokens": 38596573.0,
      "step": 15225
    },
    {
      "entropy": 6.241439723968506,
      "epoch": 1.7576457010963646,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047004525441997694,
      "loss": 5.8242,
      "mean_token_accuracy": 0.18651655912399293,
      "num_tokens": 38610083.0,
      "step": 15230
    },
    {
      "entropy": 6.2788787364959715,
      "epoch": 1.7582227351413735,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004700246792890376,
      "loss": 5.81,
      "mean_token_accuracy": 0.1898006170988083,
      "num_tokens": 38622710.0,
      "step": 15235
    },
    {
      "entropy": 6.208138942718506,
      "epoch": 1.758799769186382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004700040975985401,
      "loss": 5.7199,
      "mean_token_accuracy": 0.198635470867157,
      "num_tokens": 38634633.0,
      "step": 15240
    },
    {
      "entropy": 6.2374348640441895,
      "epoch": 1.7593768032313908,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046998350934917654,
      "loss": 5.8233,
      "mean_token_accuracy": 0.19056568741798402,
      "num_tokens": 38646836.0,
      "step": 15245
    },
    {
      "entropy": 6.278511476516724,
      "epoch": 1.7599538372763992,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046996291454163956,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1969834163784981,
      "num_tokens": 38659251.0,
      "step": 15250
    },
    {
      "entropy": 6.280281734466553,
      "epoch": 1.760530871321408,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004699423131766218,
      "loss": 5.8024,
      "mean_token_accuracy": 0.19038399308919907,
      "num_tokens": 38672464.0,
      "step": 15255
    },
    {
      "entropy": 6.199853897094727,
      "epoch": 1.7611079053664165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004699217052548161,
      "loss": 5.8207,
      "mean_token_accuracy": 0.19051229506731032,
      "num_tokens": 38684989.0,
      "step": 15260
    },
    {
      "entropy": 6.288009595870972,
      "epoch": 1.7616849394114253,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046990109077691577,
      "loss": 5.856,
      "mean_token_accuracy": 0.1899728834629059,
      "num_tokens": 38698528.0,
      "step": 15265
    },
    {
      "entropy": 6.227785587310791,
      "epoch": 1.762261973456434,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046988046974361406,
      "loss": 5.7521,
      "mean_token_accuracy": 0.18951063156127929,
      "num_tokens": 38710525.0,
      "step": 15270
    },
    {
      "entropy": 6.201137542724609,
      "epoch": 1.7628390075014426,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004698598421556045,
      "loss": 5.7837,
      "mean_token_accuracy": 0.19043705761432647,
      "num_tokens": 38723668.0,
      "step": 15275
    },
    {
      "entropy": 6.278661584854126,
      "epoch": 1.7634160415464513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004698392080135809,
      "loss": 5.8131,
      "mean_token_accuracy": 0.18998748660087586,
      "num_tokens": 38735136.0,
      "step": 15280
    },
    {
      "entropy": 6.247955226898194,
      "epoch": 1.76399307559146,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004698185673182374,
      "loss": 5.7959,
      "mean_token_accuracy": 0.19235802590847015,
      "num_tokens": 38748118.0,
      "step": 15285
    },
    {
      "entropy": 6.16922402381897,
      "epoch": 1.7645701096364685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046979792007026817,
      "loss": 5.7343,
      "mean_token_accuracy": 0.19652674049139024,
      "num_tokens": 38759685.0,
      "step": 15290
    },
    {
      "entropy": 6.181498336791992,
      "epoch": 1.7651471436814772,
      "grad_norm": 0.875,
      "learning_rate": 0.0004697772662703676,
      "loss": 5.686,
      "mean_token_accuracy": 0.20263478755950928,
      "num_tokens": 38772541.0,
      "step": 15295
    },
    {
      "entropy": 6.283728694915771,
      "epoch": 1.7657241777264858,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046975660591923047,
      "loss": 5.8371,
      "mean_token_accuracy": 0.18907580971717836,
      "num_tokens": 38785319.0,
      "step": 15300
    },
    {
      "entropy": 6.213206720352173,
      "epoch": 1.7663012117714945,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004697359390175516,
      "loss": 5.8046,
      "mean_token_accuracy": 0.18562636375427247,
      "num_tokens": 38797979.0,
      "step": 15305
    },
    {
      "entropy": 6.1605034351348875,
      "epoch": 1.7668782458165033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046971526556602625,
      "loss": 5.6757,
      "mean_token_accuracy": 0.20190499871969222,
      "num_tokens": 38810797.0,
      "step": 15310
    },
    {
      "entropy": 6.205953788757324,
      "epoch": 1.7674552798615117,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004696945855653495,
      "loss": 5.7606,
      "mean_token_accuracy": 0.18970787823200225,
      "num_tokens": 38822409.0,
      "step": 15315
    },
    {
      "entropy": 6.284510850906372,
      "epoch": 1.7680323139065206,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004696738990162172,
      "loss": 5.8394,
      "mean_token_accuracy": 0.1881306752562523,
      "num_tokens": 38835952.0,
      "step": 15320
    },
    {
      "entropy": 6.249301147460938,
      "epoch": 1.768609347951529,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000469653205919325,
      "loss": 5.7863,
      "mean_token_accuracy": 0.19565696120262147,
      "num_tokens": 38848778.0,
      "step": 15325
    },
    {
      "entropy": 6.201503849029541,
      "epoch": 1.7691863819965379,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046963250627536884,
      "loss": 5.782,
      "mean_token_accuracy": 0.1950277715921402,
      "num_tokens": 38860885.0,
      "step": 15330
    },
    {
      "entropy": 6.2414515018463135,
      "epoch": 1.7697634160415463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004696118000850451,
      "loss": 5.7262,
      "mean_token_accuracy": 0.19871688932180404,
      "num_tokens": 38873939.0,
      "step": 15335
    },
    {
      "entropy": 6.318985319137573,
      "epoch": 1.7703404500865552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046959108734905003,
      "loss": 5.8338,
      "mean_token_accuracy": 0.18968331664800644,
      "num_tokens": 38887648.0,
      "step": 15340
    },
    {
      "entropy": 6.253871011734009,
      "epoch": 1.7709174841315638,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046957036806808045,
      "loss": 5.8554,
      "mean_token_accuracy": 0.18659120351076125,
      "num_tokens": 38899908.0,
      "step": 15345
    },
    {
      "entropy": 6.307688665390015,
      "epoch": 1.7714945181765724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004695496422428332,
      "loss": 5.8519,
      "mean_token_accuracy": 0.1870699480175972,
      "num_tokens": 38910918.0,
      "step": 15350
    },
    {
      "entropy": 6.26733365058899,
      "epoch": 1.772071552221581,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004695289098740054,
      "loss": 5.8279,
      "mean_token_accuracy": 0.18899553567171096,
      "num_tokens": 38923427.0,
      "step": 15355
    },
    {
      "entropy": 6.149141693115235,
      "epoch": 1.7726485862665897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004695081709622943,
      "loss": 5.739,
      "mean_token_accuracy": 0.19880751222372056,
      "num_tokens": 38935308.0,
      "step": 15360
    },
    {
      "entropy": 6.3128434181213375,
      "epoch": 1.7732256203115984,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046948742550839744,
      "loss": 5.8984,
      "mean_token_accuracy": 0.18861982226371765,
      "num_tokens": 38948444.0,
      "step": 15365
    },
    {
      "entropy": 6.257227563858033,
      "epoch": 1.773802654356607,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004694666735130127,
      "loss": 5.7723,
      "mean_token_accuracy": 0.18938975781202316,
      "num_tokens": 38959888.0,
      "step": 15370
    },
    {
      "entropy": 6.1226355075836185,
      "epoch": 1.7743796884016156,
      "grad_norm": 1.09375,
      "learning_rate": 0.000469445914976838,
      "loss": 5.6955,
      "mean_token_accuracy": 0.1981079176068306,
      "num_tokens": 38971769.0,
      "step": 15375
    },
    {
      "entropy": 6.249244260787964,
      "epoch": 1.7749567224466243,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004694251499005715,
      "loss": 5.7407,
      "mean_token_accuracy": 0.19736608266830444,
      "num_tokens": 38984718.0,
      "step": 15380
    },
    {
      "entropy": 6.276199769973755,
      "epoch": 1.7755337564916331,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004694043782849116,
      "loss": 5.7811,
      "mean_token_accuracy": 0.19389694929122925,
      "num_tokens": 38996670.0,
      "step": 15385
    },
    {
      "entropy": 6.172970199584961,
      "epoch": 1.7761107905366416,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046938360013055715,
      "loss": 5.7571,
      "mean_token_accuracy": 0.19956991225481033,
      "num_tokens": 39009270.0,
      "step": 15390
    },
    {
      "entropy": 6.259806966781616,
      "epoch": 1.7766878245816504,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046936281543820673,
      "loss": 5.8233,
      "mean_token_accuracy": 0.18570415675640106,
      "num_tokens": 39022332.0,
      "step": 15395
    },
    {
      "entropy": 6.262522220611572,
      "epoch": 1.7772648586266588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046934202420855967,
      "loss": 5.7347,
      "mean_token_accuracy": 0.1933535799384117,
      "num_tokens": 39034964.0,
      "step": 15400
    },
    {
      "entropy": 6.2366499423980715,
      "epoch": 1.7778418926716677,
      "grad_norm": 3.046875,
      "learning_rate": 0.00046932122644231507,
      "loss": 5.7869,
      "mean_token_accuracy": 0.19790620654821395,
      "num_tokens": 39048705.0,
      "step": 15405
    },
    {
      "entropy": 6.224163055419922,
      "epoch": 1.7784189267166761,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046930042214017256,
      "loss": 5.8219,
      "mean_token_accuracy": 0.1893312320113182,
      "num_tokens": 39061994.0,
      "step": 15410
    },
    {
      "entropy": 6.268382787704468,
      "epoch": 1.778995960761685,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004692796113028319,
      "loss": 5.7858,
      "mean_token_accuracy": 0.19634834676980972,
      "num_tokens": 39075279.0,
      "step": 15415
    },
    {
      "entropy": 6.2748369693756105,
      "epoch": 1.7795729948066936,
      "grad_norm": 0.9375,
      "learning_rate": 0.000469258793930993,
      "loss": 5.8036,
      "mean_token_accuracy": 0.19051386564970016,
      "num_tokens": 39087914.0,
      "step": 15420
    },
    {
      "entropy": 6.239663934707641,
      "epoch": 1.7801500288517023,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046923797002535605,
      "loss": 5.7727,
      "mean_token_accuracy": 0.19798492640256882,
      "num_tokens": 39100302.0,
      "step": 15425
    },
    {
      "entropy": 6.197549295425415,
      "epoch": 1.780727062896711,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004692171395866214,
      "loss": 5.8043,
      "mean_token_accuracy": 0.19416270405054092,
      "num_tokens": 39114073.0,
      "step": 15430
    },
    {
      "entropy": 6.224683809280395,
      "epoch": 1.7813040969417195,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046919630261548986,
      "loss": 5.7919,
      "mean_token_accuracy": 0.19371367543935775,
      "num_tokens": 39126213.0,
      "step": 15435
    },
    {
      "entropy": 6.269141149520874,
      "epoch": 1.7818811309867282,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046917545911266207,
      "loss": 5.7893,
      "mean_token_accuracy": 0.18980225920677185,
      "num_tokens": 39138600.0,
      "step": 15440
    },
    {
      "entropy": 6.2688933372497555,
      "epoch": 1.7824581650317368,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046915460907883923,
      "loss": 5.7776,
      "mean_token_accuracy": 0.1912994474172592,
      "num_tokens": 39149356.0,
      "step": 15445
    },
    {
      "entropy": 6.196773815155029,
      "epoch": 1.7830351990767457,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046913375251472246,
      "loss": 5.7428,
      "mean_token_accuracy": 0.19297046810388566,
      "num_tokens": 39162141.0,
      "step": 15450
    },
    {
      "entropy": 6.21298131942749,
      "epoch": 1.783612233121754,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046911288942101345,
      "loss": 5.8053,
      "mean_token_accuracy": 0.1919914796948433,
      "num_tokens": 39173714.0,
      "step": 15455
    },
    {
      "entropy": 6.205303573608399,
      "epoch": 1.784189267166763,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004690920197984138,
      "loss": 5.7003,
      "mean_token_accuracy": 0.19532560110092162,
      "num_tokens": 39186498.0,
      "step": 15460
    },
    {
      "entropy": 6.211568069458008,
      "epoch": 1.7847663012117714,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004690711436476254,
      "loss": 5.7846,
      "mean_token_accuracy": 0.19685201942920685,
      "num_tokens": 39199762.0,
      "step": 15465
    },
    {
      "entropy": 6.140454053878784,
      "epoch": 1.7853433352567802,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046905026096935056,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19158038645982742,
      "num_tokens": 39213236.0,
      "step": 15470
    },
    {
      "entropy": 6.29591326713562,
      "epoch": 1.7859203693017887,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004690293717642916,
      "loss": 5.8148,
      "mean_token_accuracy": 0.18855440616607666,
      "num_tokens": 39225538.0,
      "step": 15475
    },
    {
      "entropy": 6.303308534622192,
      "epoch": 1.7864974033467975,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000469008476033151,
      "loss": 5.8306,
      "mean_token_accuracy": 0.19007943719625472,
      "num_tokens": 39237889.0,
      "step": 15480
    },
    {
      "entropy": 6.2144042491912845,
      "epoch": 1.7870744373918062,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046898757377663176,
      "loss": 5.8133,
      "mean_token_accuracy": 0.1894120529294014,
      "num_tokens": 39250686.0,
      "step": 15485
    },
    {
      "entropy": 6.219473171234131,
      "epoch": 1.7876514714368148,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046896666499543683,
      "loss": 5.7278,
      "mean_token_accuracy": 0.1920175477862358,
      "num_tokens": 39262708.0,
      "step": 15490
    },
    {
      "entropy": 6.091581058502197,
      "epoch": 1.7882285054818234,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046894574969026946,
      "loss": 5.6015,
      "mean_token_accuracy": 0.20872094929218293,
      "num_tokens": 39274778.0,
      "step": 15495
    },
    {
      "entropy": 6.2646256446838375,
      "epoch": 1.788805539526832,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046892482786183313,
      "loss": 5.8207,
      "mean_token_accuracy": 0.18871899396181108,
      "num_tokens": 39288088.0,
      "step": 15500
    },
    {
      "entropy": 6.185035753250122,
      "epoch": 1.7893825735718407,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046890389951083155,
      "loss": 5.6939,
      "mean_token_accuracy": 0.20895328968763352,
      "num_tokens": 39300828.0,
      "step": 15505
    },
    {
      "entropy": 6.232370281219483,
      "epoch": 1.7899596076168494,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046888296463796857,
      "loss": 5.723,
      "mean_token_accuracy": 0.1999172881245613,
      "num_tokens": 39313839.0,
      "step": 15510
    },
    {
      "entropy": 6.143598318099976,
      "epoch": 1.790536641661858,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004688620232439485,
      "loss": 5.7998,
      "mean_token_accuracy": 0.19292398542165756,
      "num_tokens": 39326115.0,
      "step": 15515
    },
    {
      "entropy": 6.349455976486206,
      "epoch": 1.7911136757068666,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046884107532947547,
      "loss": 5.7806,
      "mean_token_accuracy": 0.19286584556102754,
      "num_tokens": 39338756.0,
      "step": 15520
    },
    {
      "entropy": 6.335796260833741,
      "epoch": 1.7916907097518755,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046882012089525415,
      "loss": 5.8365,
      "mean_token_accuracy": 0.1807466298341751,
      "num_tokens": 39352406.0,
      "step": 15525
    },
    {
      "entropy": 6.233527994155883,
      "epoch": 1.792267743796884,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004687991599419895,
      "loss": 5.8199,
      "mean_token_accuracy": 0.1852112054824829,
      "num_tokens": 39366532.0,
      "step": 15530
    },
    {
      "entropy": 6.246149587631225,
      "epoch": 1.7928447778418928,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046877819247038624,
      "loss": 5.8138,
      "mean_token_accuracy": 0.19160676896572112,
      "num_tokens": 39378730.0,
      "step": 15535
    },
    {
      "entropy": 6.287089872360229,
      "epoch": 1.7934218118869012,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004687572184811497,
      "loss": 5.8392,
      "mean_token_accuracy": 0.18856973052024842,
      "num_tokens": 39391448.0,
      "step": 15540
    },
    {
      "entropy": 6.259625005722046,
      "epoch": 1.79399884593191,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046873623797498545,
      "loss": 5.7664,
      "mean_token_accuracy": 0.19872234016656876,
      "num_tokens": 39405120.0,
      "step": 15545
    },
    {
      "entropy": 6.262493562698364,
      "epoch": 1.7945758799769185,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000468715250952599,
      "loss": 5.8111,
      "mean_token_accuracy": 0.1871326148509979,
      "num_tokens": 39417508.0,
      "step": 15550
    },
    {
      "entropy": 6.263840770721435,
      "epoch": 1.7951529140219273,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046869425741469635,
      "loss": 5.7805,
      "mean_token_accuracy": 0.19236947745084762,
      "num_tokens": 39430185.0,
      "step": 15555
    },
    {
      "entropy": 6.272488355636597,
      "epoch": 1.795729948066936,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004686732573619835,
      "loss": 5.8552,
      "mean_token_accuracy": 0.19748634696006775,
      "num_tokens": 39442855.0,
      "step": 15560
    },
    {
      "entropy": 6.1940394878387455,
      "epoch": 1.7963069821119446,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004686522507951669,
      "loss": 5.7366,
      "mean_token_accuracy": 0.2001826748251915,
      "num_tokens": 39454855.0,
      "step": 15565
    },
    {
      "entropy": 6.26869535446167,
      "epoch": 1.7968840161569533,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004686312377149531,
      "loss": 5.7833,
      "mean_token_accuracy": 0.18621684610843658,
      "num_tokens": 39466905.0,
      "step": 15570
    },
    {
      "entropy": 6.259698343276978,
      "epoch": 1.797461050201962,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046861021812204874,
      "loss": 5.7695,
      "mean_token_accuracy": 0.19225990921258926,
      "num_tokens": 39479907.0,
      "step": 15575
    },
    {
      "entropy": 6.162016296386719,
      "epoch": 1.7980380842469705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046858919201716085,
      "loss": 5.7274,
      "mean_token_accuracy": 0.19403126537799836,
      "num_tokens": 39491934.0,
      "step": 15580
    },
    {
      "entropy": 6.25756950378418,
      "epoch": 1.7986151182919792,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004685681594009966,
      "loss": 5.8335,
      "mean_token_accuracy": 0.1953853487968445,
      "num_tokens": 39503669.0,
      "step": 15585
    },
    {
      "entropy": 6.336859130859375,
      "epoch": 1.799192152336988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046854712027426357,
      "loss": 5.8108,
      "mean_token_accuracy": 0.19571201205253602,
      "num_tokens": 39515555.0,
      "step": 15590
    },
    {
      "entropy": 6.161793518066406,
      "epoch": 1.7997691863819965,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046852607463766923,
      "loss": 5.7413,
      "mean_token_accuracy": 0.19812346249818802,
      "num_tokens": 39528867.0,
      "step": 15595
    },
    {
      "entropy": 6.278372001647949,
      "epoch": 1.8003462204270053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004685050224919215,
      "loss": 5.8186,
      "mean_token_accuracy": 0.19383434802293778,
      "num_tokens": 39541768.0,
      "step": 15600
    },
    {
      "entropy": 6.3132466793060305,
      "epoch": 1.8009232544720137,
      "grad_norm": 0.84375,
      "learning_rate": 0.00046848396383772844,
      "loss": 5.7951,
      "mean_token_accuracy": 0.1934720277786255,
      "num_tokens": 39553578.0,
      "step": 15605
    },
    {
      "entropy": 6.2875152111053465,
      "epoch": 1.8015002885170226,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004684628986757984,
      "loss": 5.7332,
      "mean_token_accuracy": 0.19695787876844406,
      "num_tokens": 39565846.0,
      "step": 15610
    },
    {
      "entropy": 6.2905107021331785,
      "epoch": 1.802077322562031,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004684418270068398,
      "loss": 5.796,
      "mean_token_accuracy": 0.19284765422344208,
      "num_tokens": 39578536.0,
      "step": 15615
    },
    {
      "entropy": 6.221612644195557,
      "epoch": 1.8026543566070399,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004684207488315615,
      "loss": 5.7621,
      "mean_token_accuracy": 0.19373471140861512,
      "num_tokens": 39591233.0,
      "step": 15620
    },
    {
      "entropy": 6.253147935867309,
      "epoch": 1.8032313906520485,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004683996641506724,
      "loss": 5.7423,
      "mean_token_accuracy": 0.1892036408185959,
      "num_tokens": 39602335.0,
      "step": 15625
    },
    {
      "entropy": 6.1253626346588135,
      "epoch": 1.8038084246970572,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046837857296488163,
      "loss": 5.6736,
      "mean_token_accuracy": 0.19969331324100495,
      "num_tokens": 39614278.0,
      "step": 15630
    },
    {
      "entropy": 6.2135913372039795,
      "epoch": 1.8043854587420658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046835747527489857,
      "loss": 5.7335,
      "mean_token_accuracy": 0.20158665478229523,
      "num_tokens": 39627218.0,
      "step": 15635
    },
    {
      "entropy": 6.191605138778686,
      "epoch": 1.8049624927870744,
      "grad_norm": 0.96875,
      "learning_rate": 0.000468336371081433,
      "loss": 5.8075,
      "mean_token_accuracy": 0.191551274061203,
      "num_tokens": 39639607.0,
      "step": 15640
    },
    {
      "entropy": 6.23504114151001,
      "epoch": 1.805539526832083,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046831526038519444,
      "loss": 5.7639,
      "mean_token_accuracy": 0.18933095484972,
      "num_tokens": 39652748.0,
      "step": 15645
    },
    {
      "entropy": 6.2898756980896,
      "epoch": 1.8061165608770917,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004682941431868933,
      "loss": 5.7432,
      "mean_token_accuracy": 0.19507113546133042,
      "num_tokens": 39665984.0,
      "step": 15650
    },
    {
      "entropy": 6.235638189315796,
      "epoch": 1.8066935949221004,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00046827301948723963,
      "loss": 5.734,
      "mean_token_accuracy": 0.19768848717212678,
      "num_tokens": 39680017.0,
      "step": 15655
    },
    {
      "entropy": 6.197491979598999,
      "epoch": 1.807270628967109,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046825188928694393,
      "loss": 5.7532,
      "mean_token_accuracy": 0.1928873047232628,
      "num_tokens": 39694070.0,
      "step": 15660
    },
    {
      "entropy": 6.257678270339966,
      "epoch": 1.8078476630121179,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004682307525867169,
      "loss": 5.7712,
      "mean_token_accuracy": 0.18583786785602568,
      "num_tokens": 39706341.0,
      "step": 15665
    },
    {
      "entropy": 6.230684852600097,
      "epoch": 1.8084246970571263,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004682096093872695,
      "loss": 5.8789,
      "mean_token_accuracy": 0.1941887155175209,
      "num_tokens": 39719552.0,
      "step": 15670
    },
    {
      "entropy": 6.338205146789551,
      "epoch": 1.8090017311021351,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046818845968931284,
      "loss": 5.8897,
      "mean_token_accuracy": 0.18326305747032165,
      "num_tokens": 39732424.0,
      "step": 15675
    },
    {
      "entropy": 6.240459632873535,
      "epoch": 1.8095787651471436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046816730349355843,
      "loss": 5.7829,
      "mean_token_accuracy": 0.1926908016204834,
      "num_tokens": 39745088.0,
      "step": 15680
    },
    {
      "entropy": 6.280318403244019,
      "epoch": 1.8101557991921524,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046814614080071756,
      "loss": 5.7616,
      "mean_token_accuracy": 0.1979309543967247,
      "num_tokens": 39756912.0,
      "step": 15685
    },
    {
      "entropy": 6.230122041702271,
      "epoch": 1.8107328332371608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046812497161150224,
      "loss": 5.7511,
      "mean_token_accuracy": 0.19394026398658754,
      "num_tokens": 39768224.0,
      "step": 15690
    },
    {
      "entropy": 6.188601922988892,
      "epoch": 1.8113098672821697,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046810379592662445,
      "loss": 5.7513,
      "mean_token_accuracy": 0.1924677312374115,
      "num_tokens": 39780486.0,
      "step": 15695
    },
    {
      "entropy": 6.336585855484008,
      "epoch": 1.8118869013271783,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046808261374679637,
      "loss": 5.7496,
      "mean_token_accuracy": 0.19565217792987824,
      "num_tokens": 39793572.0,
      "step": 15700
    },
    {
      "entropy": 6.2124206066131595,
      "epoch": 1.812463935372187,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004680614250727305,
      "loss": 5.7469,
      "mean_token_accuracy": 0.1988078847527504,
      "num_tokens": 39806064.0,
      "step": 15705
    },
    {
      "entropy": 6.156337785720825,
      "epoch": 1.8130409694171956,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004680402299051395,
      "loss": 5.735,
      "mean_token_accuracy": 0.19639781713485718,
      "num_tokens": 39818638.0,
      "step": 15710
    },
    {
      "entropy": 6.287319326400757,
      "epoch": 1.8136180034622043,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004680190282447363,
      "loss": 5.819,
      "mean_token_accuracy": 0.1878993481397629,
      "num_tokens": 39831788.0,
      "step": 15715
    },
    {
      "entropy": 6.27786111831665,
      "epoch": 1.814195037507213,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004679978200922339,
      "loss": 5.7609,
      "mean_token_accuracy": 0.19526689350605012,
      "num_tokens": 39844043.0,
      "step": 15720
    },
    {
      "entropy": 6.215310716629029,
      "epoch": 1.8147720715522215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004679766054483457,
      "loss": 5.839,
      "mean_token_accuracy": 0.19321362525224686,
      "num_tokens": 39857351.0,
      "step": 15725
    },
    {
      "entropy": 6.187804365158081,
      "epoch": 1.8153491055972304,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004679553843137852,
      "loss": 5.7405,
      "mean_token_accuracy": 0.1965793937444687,
      "num_tokens": 39868975.0,
      "step": 15730
    },
    {
      "entropy": 6.186919403076172,
      "epoch": 1.8159261396422388,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046793415668926625,
      "loss": 5.7014,
      "mean_token_accuracy": 0.20223922729492189,
      "num_tokens": 39880995.0,
      "step": 15735
    },
    {
      "entropy": 6.233175754547119,
      "epoch": 1.8165031736872477,
      "grad_norm": 0.875,
      "learning_rate": 0.0004679129225755028,
      "loss": 5.7193,
      "mean_token_accuracy": 0.19336370825767518,
      "num_tokens": 39893821.0,
      "step": 15740
    },
    {
      "entropy": 6.224366140365601,
      "epoch": 1.817080207732256,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000467891681973209,
      "loss": 5.78,
      "mean_token_accuracy": 0.19717964828014373,
      "num_tokens": 39906938.0,
      "step": 15745
    },
    {
      "entropy": 6.248828458786011,
      "epoch": 1.817657241777265,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046787043488309926,
      "loss": 5.7265,
      "mean_token_accuracy": 0.19956559389829637,
      "num_tokens": 39920647.0,
      "step": 15750
    },
    {
      "entropy": 6.278346681594849,
      "epoch": 1.8182342758222734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004678491813058882,
      "loss": 5.7633,
      "mean_token_accuracy": 0.19119132459163665,
      "num_tokens": 39932925.0,
      "step": 15755
    },
    {
      "entropy": 6.23821177482605,
      "epoch": 1.8188113098672822,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004678279212422908,
      "loss": 5.7758,
      "mean_token_accuracy": 0.1951742097735405,
      "num_tokens": 39945950.0,
      "step": 15760
    },
    {
      "entropy": 6.202953624725342,
      "epoch": 1.8193883439122909,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004678066546930221,
      "loss": 5.8113,
      "mean_token_accuracy": 0.1936275541782379,
      "num_tokens": 39957231.0,
      "step": 15765
    },
    {
      "entropy": 6.237277364730835,
      "epoch": 1.8199653779572995,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046778538165879725,
      "loss": 5.7155,
      "mean_token_accuracy": 0.20207455158233642,
      "num_tokens": 39969390.0,
      "step": 15770
    },
    {
      "entropy": 6.2362157821655275,
      "epoch": 1.8205424120023082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046776410214033185,
      "loss": 5.8139,
      "mean_token_accuracy": 0.18982597887516023,
      "num_tokens": 39982893.0,
      "step": 15775
    },
    {
      "entropy": 6.234827518463135,
      "epoch": 1.8211194460473168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004677428161383417,
      "loss": 5.8394,
      "mean_token_accuracy": 0.19033005833625793,
      "num_tokens": 39996413.0,
      "step": 15780
    },
    {
      "entropy": 6.285609388351441,
      "epoch": 1.8216964800923254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004677215236535426,
      "loss": 5.8296,
      "mean_token_accuracy": 0.1923249825835228,
      "num_tokens": 40008662.0,
      "step": 15785
    },
    {
      "entropy": 6.242558097839355,
      "epoch": 1.822273514137334,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004677002246866508,
      "loss": 5.7617,
      "mean_token_accuracy": 0.19329970479011535,
      "num_tokens": 40021511.0,
      "step": 15790
    },
    {
      "entropy": 6.20531964302063,
      "epoch": 1.8228505481823427,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046767891923838264,
      "loss": 5.7345,
      "mean_token_accuracy": 0.19284595549106598,
      "num_tokens": 40034205.0,
      "step": 15795
    },
    {
      "entropy": 6.242962837219238,
      "epoch": 1.8234275822273514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004676576073094548,
      "loss": 5.8129,
      "mean_token_accuracy": 0.1903470501303673,
      "num_tokens": 40046507.0,
      "step": 15800
    },
    {
      "entropy": 6.186423873901367,
      "epoch": 1.8240046162723602,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046763628890058396,
      "loss": 5.6621,
      "mean_token_accuracy": 0.20135427862405778,
      "num_tokens": 40058561.0,
      "step": 15805
    },
    {
      "entropy": 6.2149138927459715,
      "epoch": 1.8245816503173686,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046761496401248734,
      "loss": 5.7684,
      "mean_token_accuracy": 0.20394706428050996,
      "num_tokens": 40070351.0,
      "step": 15810
    },
    {
      "entropy": 6.125611209869385,
      "epoch": 1.8251586843623775,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000467593632645882,
      "loss": 5.6349,
      "mean_token_accuracy": 0.21305559426546097,
      "num_tokens": 40084057.0,
      "step": 15815
    },
    {
      "entropy": 6.300311183929443,
      "epoch": 1.825735718407386,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004675722948014855,
      "loss": 5.875,
      "mean_token_accuracy": 0.18812395483255387,
      "num_tokens": 40097584.0,
      "step": 15820
    },
    {
      "entropy": 6.093109655380249,
      "epoch": 1.8263127524523948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046755095048001556,
      "loss": 5.7656,
      "mean_token_accuracy": 0.200607630610466,
      "num_tokens": 40110303.0,
      "step": 15825
    },
    {
      "entropy": 6.279723882675171,
      "epoch": 1.8268897864974032,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004675295996821899,
      "loss": 5.8619,
      "mean_token_accuracy": 0.19158370047807693,
      "num_tokens": 40122594.0,
      "step": 15830
    },
    {
      "entropy": 6.224762010574341,
      "epoch": 1.827466820542412,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000467508242408727,
      "loss": 5.737,
      "mean_token_accuracy": 0.1933351293206215,
      "num_tokens": 40133737.0,
      "step": 15835
    },
    {
      "entropy": 6.281069231033325,
      "epoch": 1.8280438545874207,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004674868786603448,
      "loss": 5.8068,
      "mean_token_accuracy": 0.1905568554997444,
      "num_tokens": 40144891.0,
      "step": 15840
    },
    {
      "entropy": 6.265624380111694,
      "epoch": 1.8286208886324293,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004674655084377622,
      "loss": 5.8727,
      "mean_token_accuracy": 0.19042191058397293,
      "num_tokens": 40158217.0,
      "step": 15845
    },
    {
      "entropy": 6.253718900680542,
      "epoch": 1.829197922677438,
      "grad_norm": 0.875,
      "learning_rate": 0.0004674441317416978,
      "loss": 5.7655,
      "mean_token_accuracy": 0.19195002168416977,
      "num_tokens": 40170733.0,
      "step": 15850
    },
    {
      "entropy": 6.216894245147705,
      "epoch": 1.8297749567224466,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046742274857287057,
      "loss": 5.8037,
      "mean_token_accuracy": 0.19177088588476182,
      "num_tokens": 40183204.0,
      "step": 15855
    },
    {
      "entropy": 6.283664131164551,
      "epoch": 1.8303519907674553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004674013589319998,
      "loss": 5.8406,
      "mean_token_accuracy": 0.18640261590480806,
      "num_tokens": 40194816.0,
      "step": 15860
    },
    {
      "entropy": 6.26936707496643,
      "epoch": 1.830929024812464,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004673799628198049,
      "loss": 5.7825,
      "mean_token_accuracy": 0.19137618988752364,
      "num_tokens": 40205796.0,
      "step": 15865
    },
    {
      "entropy": 6.223792457580567,
      "epoch": 1.8315060588574728,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046735856023700546,
      "loss": 5.7403,
      "mean_token_accuracy": 0.1915697932243347,
      "num_tokens": 40218232.0,
      "step": 15870
    },
    {
      "entropy": 6.28590841293335,
      "epoch": 1.8320830929024812,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004673371511843215,
      "loss": 5.8333,
      "mean_token_accuracy": 0.19246375262737275,
      "num_tokens": 40230441.0,
      "step": 15875
    },
    {
      "entropy": 6.232549095153809,
      "epoch": 1.83266012694749,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004673157356624729,
      "loss": 5.8275,
      "mean_token_accuracy": 0.1854191690683365,
      "num_tokens": 40241858.0,
      "step": 15880
    },
    {
      "entropy": 6.283795022964478,
      "epoch": 1.8332371609924984,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004672943136721801,
      "loss": 5.7287,
      "mean_token_accuracy": 0.1905465230345726,
      "num_tokens": 40254310.0,
      "step": 15885
    },
    {
      "entropy": 6.236601543426514,
      "epoch": 1.8338141950375073,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004672728852141636,
      "loss": 5.7768,
      "mean_token_accuracy": 0.1922473281621933,
      "num_tokens": 40267851.0,
      "step": 15890
    },
    {
      "entropy": 6.30166335105896,
      "epoch": 1.8343912290825157,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046725145028914404,
      "loss": 5.7959,
      "mean_token_accuracy": 0.19823629558086395,
      "num_tokens": 40279899.0,
      "step": 15895
    },
    {
      "entropy": 6.22863359451294,
      "epoch": 1.8349682631275246,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004672300088978425,
      "loss": 5.8148,
      "mean_token_accuracy": 0.18159475475549697,
      "num_tokens": 40291996.0,
      "step": 15900
    },
    {
      "entropy": 6.217133569717407,
      "epoch": 1.8355452971725332,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004672085610409801,
      "loss": 5.6342,
      "mean_token_accuracy": 0.20693485140800477,
      "num_tokens": 40304426.0,
      "step": 15905
    },
    {
      "entropy": 6.260714769363403,
      "epoch": 1.8361223312175419,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046718710671927815,
      "loss": 5.7911,
      "mean_token_accuracy": 0.18990426510572433,
      "num_tokens": 40318287.0,
      "step": 15910
    },
    {
      "entropy": 6.19984130859375,
      "epoch": 1.8366993652625505,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046716564593345843,
      "loss": 5.7073,
      "mean_token_accuracy": 0.20030577778816222,
      "num_tokens": 40330906.0,
      "step": 15915
    },
    {
      "entropy": 6.180579662322998,
      "epoch": 1.8372763993075591,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046714417868424265,
      "loss": 5.7613,
      "mean_token_accuracy": 0.1926427274942398,
      "num_tokens": 40342310.0,
      "step": 15920
    },
    {
      "entropy": 6.292046546936035,
      "epoch": 1.8378534333525678,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046712270497235284,
      "loss": 5.8328,
      "mean_token_accuracy": 0.1924153372645378,
      "num_tokens": 40355596.0,
      "step": 15925
    },
    {
      "entropy": 6.3186859607696535,
      "epoch": 1.8384304673975764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004671012247985112,
      "loss": 5.8142,
      "mean_token_accuracy": 0.1861497163772583,
      "num_tokens": 40368841.0,
      "step": 15930
    },
    {
      "entropy": 6.243971681594848,
      "epoch": 1.839007501442585,
      "grad_norm": 1.0,
      "learning_rate": 0.00046707973816344044,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1947872132062912,
      "num_tokens": 40381496.0,
      "step": 15935
    },
    {
      "entropy": 6.308496189117432,
      "epoch": 1.8395845354875937,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046705824506786304,
      "loss": 5.8244,
      "mean_token_accuracy": 0.19166578948497773,
      "num_tokens": 40393750.0,
      "step": 15940
    },
    {
      "entropy": 6.109599971771241,
      "epoch": 1.8401615695326026,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046703674551250193,
      "loss": 5.6082,
      "mean_token_accuracy": 0.20835887640714645,
      "num_tokens": 40406262.0,
      "step": 15945
    },
    {
      "entropy": 6.123774766921997,
      "epoch": 1.840738603577611,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004670152394980803,
      "loss": 5.6297,
      "mean_token_accuracy": 0.20178558230400084,
      "num_tokens": 40418776.0,
      "step": 15950
    },
    {
      "entropy": 6.223650074005127,
      "epoch": 1.8413156376226198,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004669937270253214,
      "loss": 5.8269,
      "mean_token_accuracy": 0.18247616291046143,
      "num_tokens": 40432300.0,
      "step": 15955
    },
    {
      "entropy": 6.225249862670898,
      "epoch": 1.8418926716676283,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000466972208094949,
      "loss": 5.8556,
      "mean_token_accuracy": 0.19137105196714402,
      "num_tokens": 40446215.0,
      "step": 15960
    },
    {
      "entropy": 6.319041013717651,
      "epoch": 1.8424697057126371,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046695068270768665,
      "loss": 5.8122,
      "mean_token_accuracy": 0.19217948615550995,
      "num_tokens": 40457932.0,
      "step": 15965
    },
    {
      "entropy": 6.238263654708862,
      "epoch": 1.8430467397576455,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046692915086425846,
      "loss": 5.7428,
      "mean_token_accuracy": 0.19858844429254532,
      "num_tokens": 40469897.0,
      "step": 15970
    },
    {
      "entropy": 6.215254878997802,
      "epoch": 1.8436237738026544,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046690761256538857,
      "loss": 5.8046,
      "mean_token_accuracy": 0.193985615670681,
      "num_tokens": 40483104.0,
      "step": 15975
    },
    {
      "entropy": 6.197273111343383,
      "epoch": 1.844200807847663,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004668860678118015,
      "loss": 5.712,
      "mean_token_accuracy": 0.201288865506649,
      "num_tokens": 40494844.0,
      "step": 15980
    },
    {
      "entropy": 6.259690141677856,
      "epoch": 1.8447778418926717,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046686451660422185,
      "loss": 5.8311,
      "mean_token_accuracy": 0.19278805404901506,
      "num_tokens": 40506972.0,
      "step": 15985
    },
    {
      "entropy": 6.226708889007568,
      "epoch": 1.8453548759376803,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046684295894337455,
      "loss": 5.7763,
      "mean_token_accuracy": 0.19328029602766036,
      "num_tokens": 40519575.0,
      "step": 15990
    },
    {
      "entropy": 6.215507888793946,
      "epoch": 1.845931909982689,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004668213948299845,
      "loss": 5.7204,
      "mean_token_accuracy": 0.19949699938297272,
      "num_tokens": 40532895.0,
      "step": 15995
    },
    {
      "entropy": 6.2020505428314205,
      "epoch": 1.8465089440276976,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004667998242647772,
      "loss": 5.7425,
      "mean_token_accuracy": 0.19412947297096253,
      "num_tokens": 40544315.0,
      "step": 16000
    },
    {
      "entropy": 6.214215898513794,
      "epoch": 1.8470859780727062,
      "grad_norm": 0.90625,
      "learning_rate": 0.000466778247248478,
      "loss": 5.775,
      "mean_token_accuracy": 0.1943423017859459,
      "num_tokens": 40557942.0,
      "step": 16005
    },
    {
      "entropy": 6.310935211181641,
      "epoch": 1.847663012117715,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046675666378181275,
      "loss": 5.8563,
      "mean_token_accuracy": 0.1852554738521576,
      "num_tokens": 40570364.0,
      "step": 16010
    },
    {
      "entropy": 6.296020126342773,
      "epoch": 1.8482400461627235,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004667350738655074,
      "loss": 5.8741,
      "mean_token_accuracy": 0.1843133807182312,
      "num_tokens": 40583259.0,
      "step": 16015
    },
    {
      "entropy": 6.189713668823242,
      "epoch": 1.8488170802077324,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046671347750028806,
      "loss": 5.7117,
      "mean_token_accuracy": 0.19806393086910248,
      "num_tokens": 40595810.0,
      "step": 16020
    },
    {
      "entropy": 6.260494184494019,
      "epoch": 1.8493941142527408,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004666918746868811,
      "loss": 5.8132,
      "mean_token_accuracy": 0.192134889960289,
      "num_tokens": 40606856.0,
      "step": 16025
    },
    {
      "entropy": 6.2589414596557615,
      "epoch": 1.8499711482977497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004666702654260133,
      "loss": 5.8342,
      "mean_token_accuracy": 0.18665049821138383,
      "num_tokens": 40618764.0,
      "step": 16030
    },
    {
      "entropy": 6.272533416748047,
      "epoch": 1.850548182342758,
      "grad_norm": 0.875,
      "learning_rate": 0.00046664864971841113,
      "loss": 5.8135,
      "mean_token_accuracy": 0.19175857454538345,
      "num_tokens": 40631833.0,
      "step": 16035
    },
    {
      "entropy": 6.308272647857666,
      "epoch": 1.851125216387767,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046662702756480195,
      "loss": 5.8312,
      "mean_token_accuracy": 0.18984233736991882,
      "num_tokens": 40644215.0,
      "step": 16040
    },
    {
      "entropy": 6.246163558959961,
      "epoch": 1.8517022504327756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046660539896591286,
      "loss": 5.7207,
      "mean_token_accuracy": 0.20192432254552842,
      "num_tokens": 40656473.0,
      "step": 16045
    },
    {
      "entropy": 6.272618150711059,
      "epoch": 1.8522792844777842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004665837639224713,
      "loss": 5.7405,
      "mean_token_accuracy": 0.20001796931028365,
      "num_tokens": 40668949.0,
      "step": 16050
    },
    {
      "entropy": 6.259154176712036,
      "epoch": 1.8528563185227929,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046656212243520505,
      "loss": 5.8119,
      "mean_token_accuracy": 0.1883505180478096,
      "num_tokens": 40681550.0,
      "step": 16055
    },
    {
      "entropy": 6.175982856750489,
      "epoch": 1.8534333525678015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046654047450484193,
      "loss": 5.6919,
      "mean_token_accuracy": 0.19975466430187225,
      "num_tokens": 40695197.0,
      "step": 16060
    },
    {
      "entropy": 6.237799072265625,
      "epoch": 1.8540103866128101,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004665188201321102,
      "loss": 5.7457,
      "mean_token_accuracy": 0.19468836784362792,
      "num_tokens": 40707091.0,
      "step": 16065
    },
    {
      "entropy": 6.228470182418823,
      "epoch": 1.8545874206578188,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046649715931773795,
      "loss": 5.6495,
      "mean_token_accuracy": 0.2074417546391487,
      "num_tokens": 40719929.0,
      "step": 16070
    },
    {
      "entropy": 6.182590961456299,
      "epoch": 1.8551644547028274,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000466475492062454,
      "loss": 5.7471,
      "mean_token_accuracy": 0.20465220212936402,
      "num_tokens": 40732897.0,
      "step": 16075
    },
    {
      "entropy": 6.154891490936279,
      "epoch": 1.855741488747836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046645381836698684,
      "loss": 5.6957,
      "mean_token_accuracy": 0.19983115047216415,
      "num_tokens": 40746427.0,
      "step": 16080
    },
    {
      "entropy": 6.261778974533081,
      "epoch": 1.856318522792845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004664321382320657,
      "loss": 5.7661,
      "mean_token_accuracy": 0.19696154147386552,
      "num_tokens": 40758538.0,
      "step": 16085
    },
    {
      "entropy": 6.281080341339111,
      "epoch": 1.8568955568378533,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046641045165841965,
      "loss": 5.8502,
      "mean_token_accuracy": 0.19018301963806153,
      "num_tokens": 40772008.0,
      "step": 16090
    },
    {
      "entropy": 6.269111776351929,
      "epoch": 1.8574725908828622,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046638875864677814,
      "loss": 5.8218,
      "mean_token_accuracy": 0.18543903082609176,
      "num_tokens": 40785654.0,
      "step": 16095
    },
    {
      "entropy": 6.367569541931152,
      "epoch": 1.8580496249278706,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004663670591978708,
      "loss": 5.8564,
      "mean_token_accuracy": 0.19085211902856827,
      "num_tokens": 40799394.0,
      "step": 16100
    },
    {
      "entropy": 6.320497131347656,
      "epoch": 1.8586266589728795,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004663453533124275,
      "loss": 5.8665,
      "mean_token_accuracy": 0.18640171587467194,
      "num_tokens": 40812054.0,
      "step": 16105
    },
    {
      "entropy": 6.260502195358276,
      "epoch": 1.859203693017888,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004663236409911783,
      "loss": 5.9093,
      "mean_token_accuracy": 0.18281945139169692,
      "num_tokens": 40825138.0,
      "step": 16110
    },
    {
      "entropy": 6.279899072647095,
      "epoch": 1.8597807270628968,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046630192223485345,
      "loss": 5.8465,
      "mean_token_accuracy": 0.18805207908153534,
      "num_tokens": 40837766.0,
      "step": 16115
    },
    {
      "entropy": 6.273029613494873,
      "epoch": 1.8603577611079054,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046628019704418345,
      "loss": 5.7803,
      "mean_token_accuracy": 0.19459346532821656,
      "num_tokens": 40851042.0,
      "step": 16120
    },
    {
      "entropy": 6.2181549072265625,
      "epoch": 1.860934795152914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466258465419899,
      "loss": 5.6763,
      "mean_token_accuracy": 0.20825317800045012,
      "num_tokens": 40864074.0,
      "step": 16125
    },
    {
      "entropy": 6.261647891998291,
      "epoch": 1.8615118291979227,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004662367273627312,
      "loss": 5.7657,
      "mean_token_accuracy": 0.19991529881954193,
      "num_tokens": 40876404.0,
      "step": 16130
    },
    {
      "entropy": 6.286422443389893,
      "epoch": 1.8620888632429313,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000466214982873411,
      "loss": 5.7417,
      "mean_token_accuracy": 0.19869434684515,
      "num_tokens": 40889684.0,
      "step": 16135
    },
    {
      "entropy": 6.215024423599243,
      "epoch": 1.86266589728794,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046619323195266975,
      "loss": 5.693,
      "mean_token_accuracy": 0.2024300992488861,
      "num_tokens": 40902223.0,
      "step": 16140
    },
    {
      "entropy": 6.280419063568115,
      "epoch": 1.8632429313329486,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004661714746012392,
      "loss": 5.8356,
      "mean_token_accuracy": 0.1964442402124405,
      "num_tokens": 40916251.0,
      "step": 16145
    },
    {
      "entropy": 6.228949499130249,
      "epoch": 1.8638199653779572,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000466149710819851,
      "loss": 5.8282,
      "mean_token_accuracy": 0.1901947185397148,
      "num_tokens": 40929248.0,
      "step": 16150
    },
    {
      "entropy": 6.263371324539184,
      "epoch": 1.8643969994229659,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004661279406092373,
      "loss": 5.7476,
      "mean_token_accuracy": 0.19684217870235443,
      "num_tokens": 40943575.0,
      "step": 16155
    },
    {
      "entropy": 6.322947835922241,
      "epoch": 1.8649740334679747,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046610616397013027,
      "loss": 5.7939,
      "mean_token_accuracy": 0.18917343020439148,
      "num_tokens": 40956256.0,
      "step": 16160
    },
    {
      "entropy": 6.242354249954223,
      "epoch": 1.8655510675129832,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004660843809032623,
      "loss": 5.8341,
      "mean_token_accuracy": 0.18999543339014052,
      "num_tokens": 40968786.0,
      "step": 16165
    },
    {
      "entropy": 6.22568712234497,
      "epoch": 1.866128101557992,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004660625914093661,
      "loss": 5.8444,
      "mean_token_accuracy": 0.1929718717932701,
      "num_tokens": 40982221.0,
      "step": 16170
    },
    {
      "entropy": 6.236858177185058,
      "epoch": 1.8667051356030004,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004660407954891745,
      "loss": 5.8025,
      "mean_token_accuracy": 0.1900842919945717,
      "num_tokens": 40995226.0,
      "step": 16175
    },
    {
      "entropy": 6.2696362972259525,
      "epoch": 1.8672821696480093,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004660189931434207,
      "loss": 5.7294,
      "mean_token_accuracy": 0.20068929940462113,
      "num_tokens": 41007251.0,
      "step": 16180
    },
    {
      "entropy": 6.290879821777343,
      "epoch": 1.8678592036930177,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004659971843728379,
      "loss": 5.7846,
      "mean_token_accuracy": 0.19502439498901367,
      "num_tokens": 41020218.0,
      "step": 16185
    },
    {
      "entropy": 6.263503122329712,
      "epoch": 1.8684362377380266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004659753691781597,
      "loss": 5.7812,
      "mean_token_accuracy": 0.191061170399189,
      "num_tokens": 41032862.0,
      "step": 16190
    },
    {
      "entropy": 6.194137334823608,
      "epoch": 1.8690132717830352,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004659535475601198,
      "loss": 5.6722,
      "mean_token_accuracy": 0.19756209552288057,
      "num_tokens": 41045992.0,
      "step": 16195
    },
    {
      "entropy": 6.166383886337281,
      "epoch": 1.8695903058280439,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046593171951945226,
      "loss": 5.7185,
      "mean_token_accuracy": 0.19709556102752684,
      "num_tokens": 41057902.0,
      "step": 16200
    },
    {
      "entropy": 6.177010583877563,
      "epoch": 1.8701673398730525,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046590988505689114,
      "loss": 5.7098,
      "mean_token_accuracy": 0.20126421749591827,
      "num_tokens": 41070318.0,
      "step": 16205
    },
    {
      "entropy": 6.227213048934937,
      "epoch": 1.8707443739180611,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046588804417317084,
      "loss": 5.7946,
      "mean_token_accuracy": 0.19370009154081344,
      "num_tokens": 41083918.0,
      "step": 16210
    },
    {
      "entropy": 6.162137794494629,
      "epoch": 1.8713214079630698,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046586619686902597,
      "loss": 5.7365,
      "mean_token_accuracy": 0.2002287432551384,
      "num_tokens": 41097730.0,
      "step": 16215
    },
    {
      "entropy": 6.285425853729248,
      "epoch": 1.8718984420080784,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046584434314519144,
      "loss": 5.839,
      "mean_token_accuracy": 0.18983967304229737,
      "num_tokens": 41110795.0,
      "step": 16220
    },
    {
      "entropy": 6.262578296661377,
      "epoch": 1.8724754760530873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004658224830024022,
      "loss": 5.7438,
      "mean_token_accuracy": 0.19550170004367828,
      "num_tokens": 41123216.0,
      "step": 16225
    },
    {
      "entropy": 6.269934177398682,
      "epoch": 1.8730525100980957,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004658006164413935,
      "loss": 5.7982,
      "mean_token_accuracy": 0.19631521105766297,
      "num_tokens": 41135293.0,
      "step": 16230
    },
    {
      "entropy": 6.162675619125366,
      "epoch": 1.8736295441431046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004657787434629009,
      "loss": 5.7839,
      "mean_token_accuracy": 0.20136774480342864,
      "num_tokens": 41149475.0,
      "step": 16235
    },
    {
      "entropy": 6.302384519577027,
      "epoch": 1.874206578188113,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046575686406765993,
      "loss": 5.8428,
      "mean_token_accuracy": 0.18724915385246277,
      "num_tokens": 41162380.0,
      "step": 16240
    },
    {
      "entropy": 6.26046929359436,
      "epoch": 1.8747836122331218,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046573497825640664,
      "loss": 5.6791,
      "mean_token_accuracy": 0.19724078774452208,
      "num_tokens": 41175483.0,
      "step": 16245
    },
    {
      "entropy": 6.102692317962647,
      "epoch": 1.8753606462781303,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004657130860298771,
      "loss": 5.6469,
      "mean_token_accuracy": 0.20460374653339386,
      "num_tokens": 41189342.0,
      "step": 16250
    },
    {
      "entropy": 6.17583532333374,
      "epoch": 1.8759376803231391,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004656911873888077,
      "loss": 5.7486,
      "mean_token_accuracy": 0.20105878561735152,
      "num_tokens": 41202342.0,
      "step": 16255
    },
    {
      "entropy": 6.178520536422729,
      "epoch": 1.8765147143681478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004656692823339349,
      "loss": 5.6504,
      "mean_token_accuracy": 0.19981459975242616,
      "num_tokens": 41215006.0,
      "step": 16260
    },
    {
      "entropy": 6.299689483642578,
      "epoch": 1.8770917484131564,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004656473708659955,
      "loss": 5.771,
      "mean_token_accuracy": 0.19751262068748474,
      "num_tokens": 41227735.0,
      "step": 16265
    },
    {
      "entropy": 6.24557580947876,
      "epoch": 1.877668782458165,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00046562545298572646,
      "loss": 5.7408,
      "mean_token_accuracy": 0.19738951325416565,
      "num_tokens": 41240731.0,
      "step": 16270
    },
    {
      "entropy": 6.278197479248047,
      "epoch": 1.8782458165031737,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000465603528693865,
      "loss": 5.8005,
      "mean_token_accuracy": 0.19058478772640228,
      "num_tokens": 41253190.0,
      "step": 16275
    },
    {
      "entropy": 6.248212718963623,
      "epoch": 1.8788228505481823,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046558159799114853,
      "loss": 5.8229,
      "mean_token_accuracy": 0.18659763187170028,
      "num_tokens": 41265224.0,
      "step": 16280
    },
    {
      "entropy": 6.157110261917114,
      "epoch": 1.879399884593191,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004655596608783147,
      "loss": 5.6945,
      "mean_token_accuracy": 0.20232847481966018,
      "num_tokens": 41277865.0,
      "step": 16285
    },
    {
      "entropy": 6.1927509784698485,
      "epoch": 1.8799769186381996,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046553771735610136,
      "loss": 5.7115,
      "mean_token_accuracy": 0.19568862318992614,
      "num_tokens": 41291660.0,
      "step": 16290
    },
    {
      "entropy": 6.303184938430786,
      "epoch": 1.8805539526832082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004655157674252465,
      "loss": 5.8545,
      "mean_token_accuracy": 0.1846228465437889,
      "num_tokens": 41304284.0,
      "step": 16295
    },
    {
      "entropy": 6.247860717773437,
      "epoch": 1.881130986728217,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046549381108648843,
      "loss": 5.7267,
      "mean_token_accuracy": 0.20209852308034898,
      "num_tokens": 41316541.0,
      "step": 16300
    },
    {
      "entropy": 6.260425329208374,
      "epoch": 1.8817080207732255,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004654718483405656,
      "loss": 5.7612,
      "mean_token_accuracy": 0.1985420897603035,
      "num_tokens": 41329809.0,
      "step": 16305
    },
    {
      "entropy": 6.234941005706787,
      "epoch": 1.8822850548182344,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046544987918821685,
      "loss": 5.7298,
      "mean_token_accuracy": 0.19663198441267013,
      "num_tokens": 41341883.0,
      "step": 16310
    },
    {
      "entropy": 6.190593242645264,
      "epoch": 1.8828620888632428,
      "grad_norm": 0.9375,
      "learning_rate": 0.000465427903630181,
      "loss": 5.6803,
      "mean_token_accuracy": 0.197895085811615,
      "num_tokens": 41354161.0,
      "step": 16315
    },
    {
      "entropy": 6.1717743396759035,
      "epoch": 1.8834391229082517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004654059216671972,
      "loss": 5.6172,
      "mean_token_accuracy": 0.2065956860780716,
      "num_tokens": 41366737.0,
      "step": 16320
    },
    {
      "entropy": 6.239705228805542,
      "epoch": 1.88401615695326,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004653839333000048,
      "loss": 5.7776,
      "mean_token_accuracy": 0.19727020859718322,
      "num_tokens": 41380525.0,
      "step": 16325
    },
    {
      "entropy": 6.22348427772522,
      "epoch": 1.884593190998269,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046536193852934334,
      "loss": 5.7082,
      "mean_token_accuracy": 0.19650211185216904,
      "num_tokens": 41393361.0,
      "step": 16330
    },
    {
      "entropy": 6.27392611503601,
      "epoch": 1.8851702250432776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046533993735595273,
      "loss": 5.7752,
      "mean_token_accuracy": 0.19842275083065034,
      "num_tokens": 41405919.0,
      "step": 16335
    },
    {
      "entropy": 6.295265340805054,
      "epoch": 1.8857472590882862,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046531792978057277,
      "loss": 5.7742,
      "mean_token_accuracy": 0.1942655473947525,
      "num_tokens": 41417870.0,
      "step": 16340
    },
    {
      "entropy": 6.177969598770142,
      "epoch": 1.8863242931332949,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004652959158039438,
      "loss": 5.7402,
      "mean_token_accuracy": 0.20612961053848267,
      "num_tokens": 41429242.0,
      "step": 16345
    },
    {
      "entropy": 6.280581998825073,
      "epoch": 1.8869013271783035,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004652738954268062,
      "loss": 5.8342,
      "mean_token_accuracy": 0.19157402515411376,
      "num_tokens": 41442073.0,
      "step": 16350
    },
    {
      "entropy": 6.254329824447632,
      "epoch": 1.8874783612233121,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046525186864990073,
      "loss": 5.8357,
      "mean_token_accuracy": 0.19065157920122147,
      "num_tokens": 41453429.0,
      "step": 16355
    },
    {
      "entropy": 6.229830694198609,
      "epoch": 1.8880553952683208,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046522983547396806,
      "loss": 5.7819,
      "mean_token_accuracy": 0.1957307830452919,
      "num_tokens": 41464940.0,
      "step": 16360
    },
    {
      "entropy": 6.264857912063599,
      "epoch": 1.8886324293133296,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004652077958997494,
      "loss": 5.7643,
      "mean_token_accuracy": 0.19267844259738923,
      "num_tokens": 41477562.0,
      "step": 16365
    },
    {
      "entropy": 6.24571213722229,
      "epoch": 1.889209463358338,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046518574992798604,
      "loss": 5.7439,
      "mean_token_accuracy": 0.20241641104221345,
      "num_tokens": 41490299.0,
      "step": 16370
    },
    {
      "entropy": 6.234987401962281,
      "epoch": 1.889786497403347,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046516369755941945,
      "loss": 5.69,
      "mean_token_accuracy": 0.1947416752576828,
      "num_tokens": 41501305.0,
      "step": 16375
    },
    {
      "entropy": 6.273256301879883,
      "epoch": 1.8903635314483553,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004651416387947914,
      "loss": 5.824,
      "mean_token_accuracy": 0.19561228156089783,
      "num_tokens": 41515390.0,
      "step": 16380
    },
    {
      "entropy": 6.286541700363159,
      "epoch": 1.8909405654933642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651195736348437,
      "loss": 5.8439,
      "mean_token_accuracy": 0.19594662338495256,
      "num_tokens": 41528247.0,
      "step": 16385
    },
    {
      "entropy": 6.228002452850342,
      "epoch": 1.8915175995383726,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004650975020803186,
      "loss": 5.7192,
      "mean_token_accuracy": 0.19499242156744004,
      "num_tokens": 41542249.0,
      "step": 16390
    },
    {
      "entropy": 6.2081914901733395,
      "epoch": 1.8920946335833815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004650754241319584,
      "loss": 5.7348,
      "mean_token_accuracy": 0.19614273607730864,
      "num_tokens": 41554835.0,
      "step": 16395
    },
    {
      "entropy": 6.309216165542603,
      "epoch": 1.8926716676283901,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046505333979050573,
      "loss": 5.8135,
      "mean_token_accuracy": 0.18881808817386628,
      "num_tokens": 41566427.0,
      "step": 16400
    },
    {
      "entropy": 6.263038921356201,
      "epoch": 1.8932487016733988,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004650312490567035,
      "loss": 5.8265,
      "mean_token_accuracy": 0.19272570759058,
      "num_tokens": 41580193.0,
      "step": 16405
    },
    {
      "entropy": 6.171067905426026,
      "epoch": 1.8938257357184074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004650091519312945,
      "loss": 5.662,
      "mean_token_accuracy": 0.20800953060388566,
      "num_tokens": 41594048.0,
      "step": 16410
    },
    {
      "entropy": 6.294504165649414,
      "epoch": 1.894402769763416,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046498704841502203,
      "loss": 5.7929,
      "mean_token_accuracy": 0.1920799732208252,
      "num_tokens": 41607283.0,
      "step": 16415
    },
    {
      "entropy": 6.198110342025757,
      "epoch": 1.8949798038084247,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004649649385086296,
      "loss": 5.68,
      "mean_token_accuracy": 0.19927254915237427,
      "num_tokens": 41619844.0,
      "step": 16420
    },
    {
      "entropy": 6.078066635131836,
      "epoch": 1.8955568378534333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004649428222128608,
      "loss": 5.6555,
      "mean_token_accuracy": 0.2020894929766655,
      "num_tokens": 41631764.0,
      "step": 16425
    },
    {
      "entropy": 6.2903937816619875,
      "epoch": 1.896133871898442,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046492069952845953,
      "loss": 5.8111,
      "mean_token_accuracy": 0.18808402121067047,
      "num_tokens": 41643703.0,
      "step": 16430
    },
    {
      "entropy": 6.226485872268677,
      "epoch": 1.8967109059434506,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046489857045617,
      "loss": 5.7518,
      "mean_token_accuracy": 0.20191602855920793,
      "num_tokens": 41657057.0,
      "step": 16435
    },
    {
      "entropy": 6.204953050613403,
      "epoch": 1.8972879399884595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046487643499673625,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1923403263092041,
      "num_tokens": 41669870.0,
      "step": 16440
    },
    {
      "entropy": 6.299987649917602,
      "epoch": 1.8978649740334679,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004648542931509029,
      "loss": 5.7476,
      "mean_token_accuracy": 0.19092549234628678,
      "num_tokens": 41681015.0,
      "step": 16445
    },
    {
      "entropy": 6.226082611083984,
      "epoch": 1.8984420080784767,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004648321449194148,
      "loss": 5.7502,
      "mean_token_accuracy": 0.2024933725595474,
      "num_tokens": 41694116.0,
      "step": 16450
    },
    {
      "entropy": 6.212370538711548,
      "epoch": 1.8990190421234852,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046480999030301673,
      "loss": 5.7133,
      "mean_token_accuracy": 0.19792062789201736,
      "num_tokens": 41707294.0,
      "step": 16455
    },
    {
      "entropy": 6.321447801589966,
      "epoch": 1.899596076168494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046478782930245395,
      "loss": 5.861,
      "mean_token_accuracy": 0.19055529236793517,
      "num_tokens": 41719837.0,
      "step": 16460
    },
    {
      "entropy": 6.301760578155518,
      "epoch": 1.9001731102135024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046476566191847176,
      "loss": 5.7445,
      "mean_token_accuracy": 0.1955765664577484,
      "num_tokens": 41732308.0,
      "step": 16465
    },
    {
      "entropy": 6.180375719070435,
      "epoch": 1.9007501442585113,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004647434881518159,
      "loss": 5.7104,
      "mean_token_accuracy": 0.1964403882622719,
      "num_tokens": 41745051.0,
      "step": 16470
    },
    {
      "entropy": 6.2650550365447994,
      "epoch": 1.90132717830352,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046472130800323194,
      "loss": 5.7651,
      "mean_token_accuracy": 0.19826030135154724,
      "num_tokens": 41757545.0,
      "step": 16475
    },
    {
      "entropy": 6.193189859390259,
      "epoch": 1.9019042123485286,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004646991214734661,
      "loss": 5.7655,
      "mean_token_accuracy": 0.19609576016664504,
      "num_tokens": 41772157.0,
      "step": 16480
    },
    {
      "entropy": 6.225481843948364,
      "epoch": 1.9024812463935372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004646769285632645,
      "loss": 5.774,
      "mean_token_accuracy": 0.19209871888160707,
      "num_tokens": 41785404.0,
      "step": 16485
    },
    {
      "entropy": 6.235558080673218,
      "epoch": 1.9030582804385459,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004646547292733737,
      "loss": 5.7175,
      "mean_token_accuracy": 0.20398244708776475,
      "num_tokens": 41797758.0,
      "step": 16490
    },
    {
      "entropy": 6.1991030216217045,
      "epoch": 1.9036353144835545,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004646325236045402,
      "loss": 5.7981,
      "mean_token_accuracy": 0.19480407238006592,
      "num_tokens": 41811162.0,
      "step": 16495
    },
    {
      "entropy": 6.196804141998291,
      "epoch": 1.9042123485285631,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000464610311557511,
      "loss": 5.6879,
      "mean_token_accuracy": 0.20205622017383576,
      "num_tokens": 41823304.0,
      "step": 16500
    },
    {
      "entropy": 6.132744455337525,
      "epoch": 1.904789382573572,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004645880931330331,
      "loss": 5.6684,
      "mean_token_accuracy": 0.19789520055055618,
      "num_tokens": 41837027.0,
      "step": 16505
    },
    {
      "entropy": 6.261451053619385,
      "epoch": 1.9053664166185804,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004645658683318539,
      "loss": 5.7695,
      "mean_token_accuracy": 0.19238380789756776,
      "num_tokens": 41851887.0,
      "step": 16510
    },
    {
      "entropy": 6.2274068832397464,
      "epoch": 1.9059434506635893,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004645436371547209,
      "loss": 5.7265,
      "mean_token_accuracy": 0.20197607278823854,
      "num_tokens": 41864908.0,
      "step": 16515
    },
    {
      "entropy": 6.234651565551758,
      "epoch": 1.9065204847085977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046452139960238186,
      "loss": 5.7191,
      "mean_token_accuracy": 0.20455272048711776,
      "num_tokens": 41877604.0,
      "step": 16520
    },
    {
      "entropy": 6.260664796829223,
      "epoch": 1.9070975187536066,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046449915567558467,
      "loss": 5.7806,
      "mean_token_accuracy": 0.18605006039142608,
      "num_tokens": 41890580.0,
      "step": 16525
    },
    {
      "entropy": 6.215595293045044,
      "epoch": 1.907674552798615,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004644769053750775,
      "loss": 5.7604,
      "mean_token_accuracy": 0.19228918403387069,
      "num_tokens": 41903833.0,
      "step": 16530
    },
    {
      "entropy": 6.248987483978271,
      "epoch": 1.9082515868436238,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004644546487016087,
      "loss": 5.77,
      "mean_token_accuracy": 0.19168411940336227,
      "num_tokens": 41916094.0,
      "step": 16535
    },
    {
      "entropy": 6.212969446182251,
      "epoch": 1.9088286208886325,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046443238565592687,
      "loss": 5.7061,
      "mean_token_accuracy": 0.20160195827484131,
      "num_tokens": 41929199.0,
      "step": 16540
    },
    {
      "entropy": 6.234220933914185,
      "epoch": 1.9094056549336411,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046441011623878087,
      "loss": 5.7572,
      "mean_token_accuracy": 0.1962108626961708,
      "num_tokens": 41942254.0,
      "step": 16545
    },
    {
      "entropy": 6.253196907043457,
      "epoch": 1.9099826889786498,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004643878404509197,
      "loss": 5.7808,
      "mean_token_accuracy": 0.19185812175273895,
      "num_tokens": 41954129.0,
      "step": 16550
    },
    {
      "entropy": 6.34280481338501,
      "epoch": 1.9105597230236584,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046436555829309264,
      "loss": 5.8103,
      "mean_token_accuracy": 0.18848645836114883,
      "num_tokens": 41966489.0,
      "step": 16555
    },
    {
      "entropy": 6.180977773666382,
      "epoch": 1.911136757068667,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000464343269766049,
      "loss": 5.7232,
      "mean_token_accuracy": 0.2049938827753067,
      "num_tokens": 41980235.0,
      "step": 16560
    },
    {
      "entropy": 6.219506120681762,
      "epoch": 1.9117137911136757,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046432097487053857,
      "loss": 5.7202,
      "mean_token_accuracy": 0.2015002191066742,
      "num_tokens": 41992531.0,
      "step": 16565
    },
    {
      "entropy": 6.237913608551025,
      "epoch": 1.9122908251586843,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046429867360731124,
      "loss": 5.7469,
      "mean_token_accuracy": 0.2003808006644249,
      "num_tokens": 42004825.0,
      "step": 16570
    },
    {
      "entropy": 6.259665155410767,
      "epoch": 1.912867859203693,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046427636597711695,
      "loss": 5.672,
      "mean_token_accuracy": 0.20354353934526442,
      "num_tokens": 42017275.0,
      "step": 16575
    },
    {
      "entropy": 6.199267816543579,
      "epoch": 1.9134448932487018,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046425405198070624,
      "loss": 5.7257,
      "mean_token_accuracy": 0.20392066538333892,
      "num_tokens": 42031102.0,
      "step": 16580
    },
    {
      "entropy": 6.244847774505615,
      "epoch": 1.9140219272937102,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046423173161882944,
      "loss": 5.8002,
      "mean_token_accuracy": 0.18953103721141815,
      "num_tokens": 42043276.0,
      "step": 16585
    },
    {
      "entropy": 6.275360059738159,
      "epoch": 1.914598961338719,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046420940489223735,
      "loss": 5.7602,
      "mean_token_accuracy": 0.19737593382596968,
      "num_tokens": 42055634.0,
      "step": 16590
    },
    {
      "entropy": 6.202316093444824,
      "epoch": 1.9151759953837275,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004641870718016809,
      "loss": 5.696,
      "mean_token_accuracy": 0.19758296608924866,
      "num_tokens": 42068910.0,
      "step": 16595
    },
    {
      "entropy": 6.187279415130615,
      "epoch": 1.9157530294287364,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004641647323479113,
      "loss": 5.6677,
      "mean_token_accuracy": 0.19819653928279876,
      "num_tokens": 42081008.0,
      "step": 16600
    },
    {
      "entropy": 6.204299116134644,
      "epoch": 1.9163300634737448,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004641423865316798,
      "loss": 5.7518,
      "mean_token_accuracy": 0.20135241448879243,
      "num_tokens": 42093982.0,
      "step": 16605
    },
    {
      "entropy": 6.249295234680176,
      "epoch": 1.9169070975187537,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004641200343537381,
      "loss": 5.7641,
      "mean_token_accuracy": 0.19256772249937057,
      "num_tokens": 42105441.0,
      "step": 16610
    },
    {
      "entropy": 6.304444456100464,
      "epoch": 1.9174841315637623,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000464097675814838,
      "loss": 5.7642,
      "mean_token_accuracy": 0.19669710844755173,
      "num_tokens": 42117636.0,
      "step": 16615
    },
    {
      "entropy": 6.250064277648926,
      "epoch": 1.918061165608771,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004640753109157316,
      "loss": 5.7889,
      "mean_token_accuracy": 0.18935710638761521,
      "num_tokens": 42130009.0,
      "step": 16620
    },
    {
      "entropy": 6.240985155105591,
      "epoch": 1.9186381996537796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046405293965717093,
      "loss": 5.7658,
      "mean_token_accuracy": 0.19223652482032777,
      "num_tokens": 42142181.0,
      "step": 16625
    },
    {
      "entropy": 6.235103893280029,
      "epoch": 1.9192152336987882,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004640305620399086,
      "loss": 5.7696,
      "mean_token_accuracy": 0.19943803995847703,
      "num_tokens": 42155626.0,
      "step": 16630
    },
    {
      "entropy": 6.200764322280884,
      "epoch": 1.9197922677437969,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004640081780646971,
      "loss": 5.7537,
      "mean_token_accuracy": 0.1985251471400261,
      "num_tokens": 42169309.0,
      "step": 16635
    },
    {
      "entropy": 6.329618740081787,
      "epoch": 1.9203693017888055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046398578773228954,
      "loss": 5.7982,
      "mean_token_accuracy": 0.19926647394895552,
      "num_tokens": 42182372.0,
      "step": 16640
    },
    {
      "entropy": 6.210278415679932,
      "epoch": 1.9209463358338144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046396339104343886,
      "loss": 5.6366,
      "mean_token_accuracy": 0.2051170289516449,
      "num_tokens": 42194792.0,
      "step": 16645
    },
    {
      "entropy": 6.23080997467041,
      "epoch": 1.9215233698788228,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004639409879988984,
      "loss": 5.8163,
      "mean_token_accuracy": 0.19260417371988298,
      "num_tokens": 42207194.0,
      "step": 16650
    },
    {
      "entropy": 6.202491521835327,
      "epoch": 1.9221004039238316,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004639185785994216,
      "loss": 5.7384,
      "mean_token_accuracy": 0.20144012719392776,
      "num_tokens": 42219058.0,
      "step": 16655
    },
    {
      "entropy": 6.200684976577759,
      "epoch": 1.92267743796884,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046389616284576226,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1990598440170288,
      "num_tokens": 42230726.0,
      "step": 16660
    },
    {
      "entropy": 6.097742080688477,
      "epoch": 1.923254472013849,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046387374073867435,
      "loss": 5.5988,
      "mean_token_accuracy": 0.208533875644207,
      "num_tokens": 42243859.0,
      "step": 16665
    },
    {
      "entropy": 6.2442773342132565,
      "epoch": 1.9238315060588573,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046385131227891197,
      "loss": 5.8068,
      "mean_token_accuracy": 0.1915905997157097,
      "num_tokens": 42257253.0,
      "step": 16670
    },
    {
      "entropy": 6.262240171432495,
      "epoch": 1.9244085401038662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004638288774672295,
      "loss": 5.7979,
      "mean_token_accuracy": 0.19101605862379073,
      "num_tokens": 42269202.0,
      "step": 16675
    },
    {
      "entropy": 6.258184003829956,
      "epoch": 1.9249855741488748,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004638064363043816,
      "loss": 5.7783,
      "mean_token_accuracy": 0.19510626792907715,
      "num_tokens": 42282714.0,
      "step": 16680
    },
    {
      "entropy": 6.293429946899414,
      "epoch": 1.9255626081938835,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046378398879112293,
      "loss": 5.7719,
      "mean_token_accuracy": 0.19166188985109328,
      "num_tokens": 42297272.0,
      "step": 16685
    },
    {
      "entropy": 6.2582902908325195,
      "epoch": 1.9261396422388921,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004637615349282086,
      "loss": 5.818,
      "mean_token_accuracy": 0.19498054981231688,
      "num_tokens": 42308938.0,
      "step": 16690
    },
    {
      "entropy": 6.230677461624145,
      "epoch": 1.9267166762839008,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004637390747163938,
      "loss": 5.7492,
      "mean_token_accuracy": 0.19492332339286805,
      "num_tokens": 42321310.0,
      "step": 16695
    },
    {
      "entropy": 6.151617956161499,
      "epoch": 1.9272937103289094,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004637166081564339,
      "loss": 5.6937,
      "mean_token_accuracy": 0.20505770295858383,
      "num_tokens": 42332760.0,
      "step": 16700
    },
    {
      "entropy": 6.262108087539673,
      "epoch": 1.927870744373918,
      "grad_norm": 0.875,
      "learning_rate": 0.00046369413524908473,
      "loss": 5.7449,
      "mean_token_accuracy": 0.19855707436800002,
      "num_tokens": 42346993.0,
      "step": 16705
    },
    {
      "entropy": 6.253094148635864,
      "epoch": 1.9284477784189267,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046367165599510207,
      "loss": 5.7913,
      "mean_token_accuracy": 0.193456669151783,
      "num_tokens": 42358701.0,
      "step": 16710
    },
    {
      "entropy": 6.232552194595337,
      "epoch": 1.9290248124639353,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004636491703952419,
      "loss": 5.849,
      "mean_token_accuracy": 0.19578557908535005,
      "num_tokens": 42371429.0,
      "step": 16715
    },
    {
      "entropy": 6.25922269821167,
      "epoch": 1.9296018465089442,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046362667845026057,
      "loss": 5.7093,
      "mean_token_accuracy": 0.19812185168266297,
      "num_tokens": 42383346.0,
      "step": 16720
    },
    {
      "entropy": 6.241939973831177,
      "epoch": 1.9301788805539526,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046360418016091467,
      "loss": 5.7579,
      "mean_token_accuracy": 0.20104775279760362,
      "num_tokens": 42395247.0,
      "step": 16725
    },
    {
      "entropy": 6.257398796081543,
      "epoch": 1.9307559145989615,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046358167552796085,
      "loss": 5.7426,
      "mean_token_accuracy": 0.19337452501058577,
      "num_tokens": 42408009.0,
      "step": 16730
    },
    {
      "entropy": 6.2471648216247555,
      "epoch": 1.9313329486439699,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046355916455215597,
      "loss": 5.814,
      "mean_token_accuracy": 0.1892106607556343,
      "num_tokens": 42420367.0,
      "step": 16735
    },
    {
      "entropy": 6.268014526367187,
      "epoch": 1.9319099826889787,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004635366472342573,
      "loss": 5.8349,
      "mean_token_accuracy": 0.19084270149469376,
      "num_tokens": 42433298.0,
      "step": 16740
    },
    {
      "entropy": 6.269929218292236,
      "epoch": 1.9324870167339872,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004635141235750222,
      "loss": 5.7813,
      "mean_token_accuracy": 0.19304104149341583,
      "num_tokens": 42446422.0,
      "step": 16745
    },
    {
      "entropy": 6.287223863601684,
      "epoch": 1.933064050778996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046349159357520815,
      "loss": 5.7736,
      "mean_token_accuracy": 0.19219256192445755,
      "num_tokens": 42458874.0,
      "step": 16750
    },
    {
      "entropy": 6.249391555786133,
      "epoch": 1.9336410848240047,
      "grad_norm": 0.90625,
      "learning_rate": 0.000463469057235573,
      "loss": 5.7184,
      "mean_token_accuracy": 0.19843102395534515,
      "num_tokens": 42472093.0,
      "step": 16755
    },
    {
      "entropy": 6.1164631843566895,
      "epoch": 1.9342181188690133,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046344651455687476,
      "loss": 5.6645,
      "mean_token_accuracy": 0.20644594728946686,
      "num_tokens": 42485396.0,
      "step": 16760
    },
    {
      "entropy": 6.254391622543335,
      "epoch": 1.934795152914022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046342396553987155,
      "loss": 5.7551,
      "mean_token_accuracy": 0.19790984094142913,
      "num_tokens": 42498379.0,
      "step": 16765
    },
    {
      "entropy": 6.246284627914429,
      "epoch": 1.9353721869590306,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046340141018532186,
      "loss": 5.7715,
      "mean_token_accuracy": 0.19197465777397155,
      "num_tokens": 42512125.0,
      "step": 16770
    },
    {
      "entropy": 6.196970462799072,
      "epoch": 1.9359492210040392,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004633788484939843,
      "loss": 5.7078,
      "mean_token_accuracy": 0.2009172633290291,
      "num_tokens": 42525539.0,
      "step": 16775
    },
    {
      "entropy": 6.198785209655762,
      "epoch": 1.9365262550490479,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046335628046661776,
      "loss": 5.6947,
      "mean_token_accuracy": 0.19902394264936446,
      "num_tokens": 42538399.0,
      "step": 16780
    },
    {
      "entropy": 6.256881380081177,
      "epoch": 1.9371032890940567,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046333370610398135,
      "loss": 5.8322,
      "mean_token_accuracy": 0.18850149363279342,
      "num_tokens": 42550607.0,
      "step": 16785
    },
    {
      "entropy": 6.3799365043640135,
      "epoch": 1.9376803231390651,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004633111254068342,
      "loss": 5.8724,
      "mean_token_accuracy": 0.18440057337284088,
      "num_tokens": 42563113.0,
      "step": 16790
    },
    {
      "entropy": 6.215098428726196,
      "epoch": 1.938257357184074,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004632885383759359,
      "loss": 5.7434,
      "mean_token_accuracy": 0.1965686023235321,
      "num_tokens": 42574792.0,
      "step": 16795
    },
    {
      "entropy": 6.16642541885376,
      "epoch": 1.9388343912290824,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046326594501204624,
      "loss": 5.6999,
      "mean_token_accuracy": 0.19677437096834183,
      "num_tokens": 42587360.0,
      "step": 16800
    },
    {
      "entropy": 6.253411340713501,
      "epoch": 1.9394114252740913,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000463243345315925,
      "loss": 5.7212,
      "mean_token_accuracy": 0.19436445236206054,
      "num_tokens": 42599901.0,
      "step": 16805
    },
    {
      "entropy": 6.27481427192688,
      "epoch": 1.9399884593190997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046322073928833224,
      "loss": 5.7673,
      "mean_token_accuracy": 0.19093959778547287,
      "num_tokens": 42612127.0,
      "step": 16810
    },
    {
      "entropy": 6.288098526000977,
      "epoch": 1.9405654933641086,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004631981269300285,
      "loss": 5.8748,
      "mean_token_accuracy": 0.18903179168701173,
      "num_tokens": 42624086.0,
      "step": 16815
    },
    {
      "entropy": 6.276782369613647,
      "epoch": 1.9411425274091172,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004631755082417742,
      "loss": 5.8001,
      "mean_token_accuracy": 0.19266050606966018,
      "num_tokens": 42637402.0,
      "step": 16820
    },
    {
      "entropy": 6.276293992996216,
      "epoch": 1.9417195614541258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004631528832243302,
      "loss": 5.8097,
      "mean_token_accuracy": 0.1864730715751648,
      "num_tokens": 42648162.0,
      "step": 16825
    },
    {
      "entropy": 6.21612663269043,
      "epoch": 1.9422965954991345,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046313025187845735,
      "loss": 5.7666,
      "mean_token_accuracy": 0.1960858479142189,
      "num_tokens": 42658787.0,
      "step": 16830
    },
    {
      "entropy": 6.250492525100708,
      "epoch": 1.942873629544143,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046310761420491705,
      "loss": 5.7561,
      "mean_token_accuracy": 0.1873353973031044,
      "num_tokens": 42672036.0,
      "step": 16835
    },
    {
      "entropy": 6.280843687057495,
      "epoch": 1.9434506635891518,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004630849702044705,
      "loss": 5.761,
      "mean_token_accuracy": 0.1913541316986084,
      "num_tokens": 42683691.0,
      "step": 16840
    },
    {
      "entropy": 6.204617929458618,
      "epoch": 1.9440276976341604,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046306231987787937,
      "loss": 5.7412,
      "mean_token_accuracy": 0.19753404557704926,
      "num_tokens": 42697469.0,
      "step": 16845
    },
    {
      "entropy": 6.237096166610717,
      "epoch": 1.944604731679169,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046303966322590556,
      "loss": 5.7657,
      "mean_token_accuracy": 0.19414567053318024,
      "num_tokens": 42710792.0,
      "step": 16850
    },
    {
      "entropy": 6.215739774703979,
      "epoch": 1.9451817657241777,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004630170002493111,
      "loss": 5.7152,
      "mean_token_accuracy": 0.19930611103773116,
      "num_tokens": 42722549.0,
      "step": 16855
    },
    {
      "entropy": 6.201346778869629,
      "epoch": 1.9457587997691865,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046299433094885826,
      "loss": 5.7226,
      "mean_token_accuracy": 0.19663690626621247,
      "num_tokens": 42736113.0,
      "step": 16860
    },
    {
      "entropy": 6.265635204315186,
      "epoch": 1.946335833814195,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046297165532530944,
      "loss": 5.7355,
      "mean_token_accuracy": 0.19952215105295182,
      "num_tokens": 42748801.0,
      "step": 16865
    },
    {
      "entropy": 6.258964967727661,
      "epoch": 1.9469128678592038,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004629489733794274,
      "loss": 5.7938,
      "mean_token_accuracy": 0.19447925239801406,
      "num_tokens": 42760738.0,
      "step": 16870
    },
    {
      "entropy": 6.259508514404297,
      "epoch": 1.9474899019042122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000462926285111975,
      "loss": 5.7896,
      "mean_token_accuracy": 0.18786740899086,
      "num_tokens": 42772470.0,
      "step": 16875
    },
    {
      "entropy": 6.261531639099121,
      "epoch": 1.948066935949221,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046290359052371535,
      "loss": 5.8194,
      "mean_token_accuracy": 0.19798532277345657,
      "num_tokens": 42784919.0,
      "step": 16880
    },
    {
      "entropy": 6.215500164031982,
      "epoch": 1.9486439699942295,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004628808896154117,
      "loss": 5.7047,
      "mean_token_accuracy": 0.1991393968462944,
      "num_tokens": 42797722.0,
      "step": 16885
    },
    {
      "entropy": 6.307239580154419,
      "epoch": 1.9492210040392384,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004628581823878277,
      "loss": 5.8117,
      "mean_token_accuracy": 0.18834513127803804,
      "num_tokens": 42809539.0,
      "step": 16890
    },
    {
      "entropy": 6.176803779602051,
      "epoch": 1.949798038084247,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000462835468841727,
      "loss": 5.5756,
      "mean_token_accuracy": 0.20189131498336793,
      "num_tokens": 42821519.0,
      "step": 16895
    },
    {
      "entropy": 6.215475225448609,
      "epoch": 1.9503750721292556,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004628127489778737,
      "loss": 5.6761,
      "mean_token_accuracy": 0.2006756141781807,
      "num_tokens": 42833966.0,
      "step": 16900
    },
    {
      "entropy": 6.210838174819946,
      "epoch": 1.9509521061742643,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004627900227970318,
      "loss": 5.7442,
      "mean_token_accuracy": 0.2009410247206688,
      "num_tokens": 42846162.0,
      "step": 16905
    },
    {
      "entropy": 6.247046756744385,
      "epoch": 1.951529140219273,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046276729029996576,
      "loss": 5.7712,
      "mean_token_accuracy": 0.2012566164135933,
      "num_tokens": 42860019.0,
      "step": 16910
    },
    {
      "entropy": 6.268760013580322,
      "epoch": 1.9521061742642816,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046274455148744025,
      "loss": 5.8269,
      "mean_token_accuracy": 0.18770326375961305,
      "num_tokens": 42872699.0,
      "step": 16915
    },
    {
      "entropy": 6.254588794708252,
      "epoch": 1.9526832083092902,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046272180636022,
      "loss": 5.7778,
      "mean_token_accuracy": 0.19753188490867615,
      "num_tokens": 42884907.0,
      "step": 16920
    },
    {
      "entropy": 6.267702579498291,
      "epoch": 1.953260242354299,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046269905491907,
      "loss": 5.7756,
      "mean_token_accuracy": 0.19919765442609788,
      "num_tokens": 42898387.0,
      "step": 16925
    },
    {
      "entropy": 6.30309271812439,
      "epoch": 1.9538372763993075,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004626762971647555,
      "loss": 5.8208,
      "mean_token_accuracy": 0.18956138491630553,
      "num_tokens": 42910845.0,
      "step": 16930
    },
    {
      "entropy": 6.252713060379028,
      "epoch": 1.9544143104443163,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046265353309804205,
      "loss": 5.7943,
      "mean_token_accuracy": 0.18836448788642884,
      "num_tokens": 42922494.0,
      "step": 16935
    },
    {
      "entropy": 6.271551084518433,
      "epoch": 1.9549913444893248,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004626307627196952,
      "loss": 5.7593,
      "mean_token_accuracy": 0.19097371846437455,
      "num_tokens": 42934618.0,
      "step": 16940
    },
    {
      "entropy": 6.270970726013184,
      "epoch": 1.9555683785343336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004626079860304808,
      "loss": 5.7123,
      "mean_token_accuracy": 0.19799288511276245,
      "num_tokens": 42947518.0,
      "step": 16945
    },
    {
      "entropy": 6.151194667816162,
      "epoch": 1.956145412579342,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046258520303116496,
      "loss": 5.7494,
      "mean_token_accuracy": 0.19470454454421998,
      "num_tokens": 42960068.0,
      "step": 16950
    },
    {
      "entropy": 6.258401679992676,
      "epoch": 1.956722446624351,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004625624137225141,
      "loss": 5.7457,
      "mean_token_accuracy": 0.18979074209928512,
      "num_tokens": 42972982.0,
      "step": 16955
    },
    {
      "entropy": 6.168237495422363,
      "epoch": 1.9572994806693593,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004625396181052945,
      "loss": 5.6874,
      "mean_token_accuracy": 0.2066708728671074,
      "num_tokens": 42985040.0,
      "step": 16960
    },
    {
      "entropy": 6.34441294670105,
      "epoch": 1.9578765147143682,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046251681618027316,
      "loss": 5.8094,
      "mean_token_accuracy": 0.1825690597295761,
      "num_tokens": 42998503.0,
      "step": 16965
    },
    {
      "entropy": 6.199335098266602,
      "epoch": 1.9584535487593768,
      "grad_norm": 0.875,
      "learning_rate": 0.0004624940079482167,
      "loss": 5.6776,
      "mean_token_accuracy": 0.2036006435751915,
      "num_tokens": 43011334.0,
      "step": 16970
    },
    {
      "entropy": 6.139124584197998,
      "epoch": 1.9590305828043855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046247119340989254,
      "loss": 5.6827,
      "mean_token_accuracy": 0.20829924792051316,
      "num_tokens": 43024772.0,
      "step": 16975
    },
    {
      "entropy": 6.279188776016236,
      "epoch": 1.959607616849394,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004624483725660678,
      "loss": 5.7346,
      "mean_token_accuracy": 0.2017065092921257,
      "num_tokens": 43036733.0,
      "step": 16980
    },
    {
      "entropy": 6.220383977890014,
      "epoch": 1.9601846508944027,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004624255454175102,
      "loss": 5.762,
      "mean_token_accuracy": 0.19977388232946397,
      "num_tokens": 43050482.0,
      "step": 16985
    },
    {
      "entropy": 6.326394701004029,
      "epoch": 1.9607616849394114,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004624027119649875,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18740826100111008,
      "num_tokens": 43062826.0,
      "step": 16990
    },
    {
      "entropy": 6.238157176971436,
      "epoch": 1.96133871898442,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046237987220926766,
      "loss": 5.6961,
      "mean_token_accuracy": 0.20505666881799697,
      "num_tokens": 43077134.0,
      "step": 16995
    },
    {
      "entropy": 6.154120922088623,
      "epoch": 1.9619157530294289,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046235702615111886,
      "loss": 5.6921,
      "mean_token_accuracy": 0.20490272492170333,
      "num_tokens": 43089871.0,
      "step": 17000
    },
    {
      "entropy": 6.276572799682617,
      "epoch": 1.9624927870744373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004623341737913096,
      "loss": 5.7894,
      "mean_token_accuracy": 0.19642854928970338,
      "num_tokens": 43102977.0,
      "step": 17005
    },
    {
      "entropy": 6.222786378860474,
      "epoch": 1.9630698211194462,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004623113151306085,
      "loss": 5.7515,
      "mean_token_accuracy": 0.19243604838848113,
      "num_tokens": 43114685.0,
      "step": 17010
    },
    {
      "entropy": 6.180130910873413,
      "epoch": 1.9636468551644546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046228845016978425,
      "loss": 5.6986,
      "mean_token_accuracy": 0.20110821574926377,
      "num_tokens": 43127208.0,
      "step": 17015
    },
    {
      "entropy": 6.266080331802368,
      "epoch": 1.9642238892094634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004622655789096061,
      "loss": 5.6871,
      "mean_token_accuracy": 0.20514248311519623,
      "num_tokens": 43140732.0,
      "step": 17020
    },
    {
      "entropy": 6.244066762924194,
      "epoch": 1.9648009232544719,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046224270135084315,
      "loss": 5.7647,
      "mean_token_accuracy": 0.19214110374450682,
      "num_tokens": 43152746.0,
      "step": 17025
    },
    {
      "entropy": 6.2248434066772464,
      "epoch": 1.9653779572994807,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046221981749426503,
      "loss": 5.714,
      "mean_token_accuracy": 0.19311913698911667,
      "num_tokens": 43165608.0,
      "step": 17030
    },
    {
      "entropy": 6.244058084487915,
      "epoch": 1.9659549913444894,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046219692734064124,
      "loss": 5.741,
      "mean_token_accuracy": 0.19818030893802643,
      "num_tokens": 43178805.0,
      "step": 17035
    },
    {
      "entropy": 6.293460655212402,
      "epoch": 1.966532025389498,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004621740308907419,
      "loss": 5.8293,
      "mean_token_accuracy": 0.18797580897808075,
      "num_tokens": 43191366.0,
      "step": 17040
    },
    {
      "entropy": 6.277713394165039,
      "epoch": 1.9671090594345066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004621511281453369,
      "loss": 5.79,
      "mean_token_accuracy": 0.19037945866584777,
      "num_tokens": 43203748.0,
      "step": 17045
    },
    {
      "entropy": 6.213826036453247,
      "epoch": 1.9676860934795153,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004621282191051967,
      "loss": 5.728,
      "mean_token_accuracy": 0.20011843591928483,
      "num_tokens": 43215661.0,
      "step": 17050
    },
    {
      "entropy": 6.289625978469848,
      "epoch": 1.968263127524524,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004621053037710918,
      "loss": 5.7874,
      "mean_token_accuracy": 0.19696867614984512,
      "num_tokens": 43227869.0,
      "step": 17055
    },
    {
      "entropy": 6.247723054885864,
      "epoch": 1.9688401615695326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000462082382143793,
      "loss": 5.7628,
      "mean_token_accuracy": 0.193458254635334,
      "num_tokens": 43240620.0,
      "step": 17060
    },
    {
      "entropy": 6.259679460525513,
      "epoch": 1.9694171956145412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046205945422407113,
      "loss": 5.7184,
      "mean_token_accuracy": 0.19636217951774598,
      "num_tokens": 43252523.0,
      "step": 17065
    },
    {
      "entropy": 6.26147232055664,
      "epoch": 1.9699942296595498,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046203652001269754,
      "loss": 5.7699,
      "mean_token_accuracy": 0.20067997127771378,
      "num_tokens": 43265731.0,
      "step": 17070
    },
    {
      "entropy": 6.1158212184906,
      "epoch": 1.9705712637045587,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046201357951044337,
      "loss": 5.6433,
      "mean_token_accuracy": 0.20104490518569945,
      "num_tokens": 43277929.0,
      "step": 17075
    },
    {
      "entropy": 6.281884956359863,
      "epoch": 1.9711482977495671,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046199063271808047,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1912989377975464,
      "num_tokens": 43290011.0,
      "step": 17080
    },
    {
      "entropy": 6.265655136108398,
      "epoch": 1.971725331794576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004619676796363804,
      "loss": 5.7726,
      "mean_token_accuracy": 0.19231055527925492,
      "num_tokens": 43303205.0,
      "step": 17085
    },
    {
      "entropy": 6.083212232589721,
      "epoch": 1.9723023658395844,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046194472026611546,
      "loss": 5.612,
      "mean_token_accuracy": 0.21285548955202102,
      "num_tokens": 43316214.0,
      "step": 17090
    },
    {
      "entropy": 6.186386346817017,
      "epoch": 1.9728793998845933,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004619217546080576,
      "loss": 5.7101,
      "mean_token_accuracy": 0.19639154374599457,
      "num_tokens": 43328497.0,
      "step": 17095
    },
    {
      "entropy": 6.289447689056397,
      "epoch": 1.9734564339296017,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004618987826629794,
      "loss": 5.8013,
      "mean_token_accuracy": 0.19299401640892028,
      "num_tokens": 43341081.0,
      "step": 17100
    },
    {
      "entropy": 6.304375314712525,
      "epoch": 1.9740334679746105,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046187580443165344,
      "loss": 5.834,
      "mean_token_accuracy": 0.19179582446813584,
      "num_tokens": 43354001.0,
      "step": 17105
    },
    {
      "entropy": 6.105527925491333,
      "epoch": 1.9746105020196192,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004618528199148527,
      "loss": 5.6094,
      "mean_token_accuracy": 0.20828552693128585,
      "num_tokens": 43368925.0,
      "step": 17110
    },
    {
      "entropy": 6.167400646209717,
      "epoch": 1.9751875360646278,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046182982911335016,
      "loss": 5.6738,
      "mean_token_accuracy": 0.20098673850297927,
      "num_tokens": 43382063.0,
      "step": 17115
    },
    {
      "entropy": 6.294412660598755,
      "epoch": 1.9757645701096365,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046180683202791905,
      "loss": 5.7374,
      "mean_token_accuracy": 0.19542604982852935,
      "num_tokens": 43393810.0,
      "step": 17120
    },
    {
      "entropy": 6.206226015090943,
      "epoch": 1.976341604154645,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000461783828659333,
      "loss": 5.7705,
      "mean_token_accuracy": 0.19850302040576934,
      "num_tokens": 43406500.0,
      "step": 17125
    },
    {
      "entropy": 6.225682783126831,
      "epoch": 1.9769186381996537,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046176081900836565,
      "loss": 5.7769,
      "mean_token_accuracy": 0.19113655984401703,
      "num_tokens": 43418284.0,
      "step": 17130
    },
    {
      "entropy": 6.243600559234619,
      "epoch": 1.9774956722446624,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046173780307579086,
      "loss": 5.7442,
      "mean_token_accuracy": 0.19873351007699966,
      "num_tokens": 43430913.0,
      "step": 17135
    },
    {
      "entropy": 6.228173828125,
      "epoch": 1.9780727062896712,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004617147808623829,
      "loss": 5.7081,
      "mean_token_accuracy": 0.20005712509155274,
      "num_tokens": 43444637.0,
      "step": 17140
    },
    {
      "entropy": 6.224865436553955,
      "epoch": 1.9786497403346797,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046169175236891605,
      "loss": 5.7577,
      "mean_token_accuracy": 0.20288255661725998,
      "num_tokens": 43457725.0,
      "step": 17145
    },
    {
      "entropy": 6.288400602340698,
      "epoch": 1.9792267743796885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004616687175961648,
      "loss": 5.7884,
      "mean_token_accuracy": 0.1922070115804672,
      "num_tokens": 43469902.0,
      "step": 17150
    },
    {
      "entropy": 6.289521169662476,
      "epoch": 1.979803808424697,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004616456765449039,
      "loss": 5.7655,
      "mean_token_accuracy": 0.19170056879520417,
      "num_tokens": 43482959.0,
      "step": 17155
    },
    {
      "entropy": 6.2213935375213625,
      "epoch": 1.9803808424697058,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046162262921590845,
      "loss": 5.7862,
      "mean_token_accuracy": 0.19721000641584396,
      "num_tokens": 43496051.0,
      "step": 17160
    },
    {
      "entropy": 6.276457834243774,
      "epoch": 1.9809578765147142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004615995756099535,
      "loss": 5.7695,
      "mean_token_accuracy": 0.19776639342308044,
      "num_tokens": 43509031.0,
      "step": 17165
    },
    {
      "entropy": 6.239616012573242,
      "epoch": 1.981534910559723,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004615765157278146,
      "loss": 5.6733,
      "mean_token_accuracy": 0.1991044834256172,
      "num_tokens": 43521577.0,
      "step": 17170
    },
    {
      "entropy": 6.274320507049561,
      "epoch": 1.9821119446047317,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046155344957026726,
      "loss": 5.7903,
      "mean_token_accuracy": 0.19143225848674775,
      "num_tokens": 43536029.0,
      "step": 17175
    },
    {
      "entropy": 6.209638595581055,
      "epoch": 1.9826889786497404,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004615303771380873,
      "loss": 5.7269,
      "mean_token_accuracy": 0.19245321601629256,
      "num_tokens": 43548663.0,
      "step": 17180
    },
    {
      "entropy": 6.302344274520874,
      "epoch": 1.983266012694749,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046150729843205077,
      "loss": 5.7943,
      "mean_token_accuracy": 0.1933743551373482,
      "num_tokens": 43560302.0,
      "step": 17185
    },
    {
      "entropy": 6.252721643447876,
      "epoch": 1.9838430467397576,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046148421345293384,
      "loss": 5.7499,
      "mean_token_accuracy": 0.1988838478922844,
      "num_tokens": 43572657.0,
      "step": 17190
    },
    {
      "entropy": 6.233315372467041,
      "epoch": 1.9844200807847663,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004614611222015131,
      "loss": 5.8142,
      "mean_token_accuracy": 0.19474863708019258,
      "num_tokens": 43584973.0,
      "step": 17195
    },
    {
      "entropy": 6.2853758335113525,
      "epoch": 1.984997114829775,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046143802467856507,
      "loss": 5.8065,
      "mean_token_accuracy": 0.19342428594827651,
      "num_tokens": 43597457.0,
      "step": 17200
    },
    {
      "entropy": 6.284518527984619,
      "epoch": 1.9855741488747836,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046141492088486673,
      "loss": 5.7607,
      "mean_token_accuracy": 0.19222481846809386,
      "num_tokens": 43610256.0,
      "step": 17205
    },
    {
      "entropy": 6.311424303054809,
      "epoch": 1.9861511829197922,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004613918108211951,
      "loss": 5.7505,
      "mean_token_accuracy": 0.1954013228416443,
      "num_tokens": 43622765.0,
      "step": 17210
    },
    {
      "entropy": 6.221784257888794,
      "epoch": 1.986728216964801,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004613686944883275,
      "loss": 5.7651,
      "mean_token_accuracy": 0.19342921823263168,
      "num_tokens": 43634353.0,
      "step": 17215
    },
    {
      "entropy": 6.281559896469116,
      "epoch": 1.9873052510098095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046134557188704146,
      "loss": 5.7791,
      "mean_token_accuracy": 0.1949659049510956,
      "num_tokens": 43647778.0,
      "step": 17220
    },
    {
      "entropy": 6.252798557281494,
      "epoch": 1.9878822850548183,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046132244301811466,
      "loss": 5.76,
      "mean_token_accuracy": 0.19927013665437698,
      "num_tokens": 43659759.0,
      "step": 17225
    },
    {
      "entropy": 6.205256986618042,
      "epoch": 1.9884593190998268,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00046129930788232497,
      "loss": 5.6927,
      "mean_token_accuracy": 0.19489791840314866,
      "num_tokens": 43673223.0,
      "step": 17230
    },
    {
      "entropy": 6.265499210357666,
      "epoch": 1.9890363531448356,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046127616648045073,
      "loss": 5.7504,
      "mean_token_accuracy": 0.1965523138642311,
      "num_tokens": 43684657.0,
      "step": 17235
    },
    {
      "entropy": 6.2796422958374025,
      "epoch": 1.989613387189844,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046125301881327007,
      "loss": 5.8439,
      "mean_token_accuracy": 0.1907721847295761,
      "num_tokens": 43697583.0,
      "step": 17240
    },
    {
      "entropy": 6.239670705795288,
      "epoch": 1.990190421234853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046122986488156167,
      "loss": 5.6987,
      "mean_token_accuracy": 0.2057715579867363,
      "num_tokens": 43711175.0,
      "step": 17245
    },
    {
      "entropy": 6.192219495773315,
      "epoch": 1.9907674552798615,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046120670468610426,
      "loss": 5.7126,
      "mean_token_accuracy": 0.20740281492471696,
      "num_tokens": 43724595.0,
      "step": 17250
    },
    {
      "entropy": 6.231134080886841,
      "epoch": 1.9913444893248702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046118353822767683,
      "loss": 5.695,
      "mean_token_accuracy": 0.20649369955062866,
      "num_tokens": 43737192.0,
      "step": 17255
    },
    {
      "entropy": 6.23353009223938,
      "epoch": 1.9919215233698788,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004611603655070586,
      "loss": 5.7352,
      "mean_token_accuracy": 0.20064079910516738,
      "num_tokens": 43749543.0,
      "step": 17260
    },
    {
      "entropy": 6.216631507873535,
      "epoch": 1.9924985574148875,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046113718652502896,
      "loss": 5.7822,
      "mean_token_accuracy": 0.19239158034324647,
      "num_tokens": 43762754.0,
      "step": 17265
    },
    {
      "entropy": 6.325669574737549,
      "epoch": 1.993075591459896,
      "grad_norm": 1.0,
      "learning_rate": 0.0004611140012823675,
      "loss": 5.7684,
      "mean_token_accuracy": 0.19782496094703675,
      "num_tokens": 43774489.0,
      "step": 17270
    },
    {
      "entropy": 6.281518840789795,
      "epoch": 1.9936526255049047,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046109080977985395,
      "loss": 5.7357,
      "mean_token_accuracy": 0.1979639410972595,
      "num_tokens": 43786519.0,
      "step": 17275
    },
    {
      "entropy": 6.096533107757568,
      "epoch": 1.9942296595499136,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046106761201826854,
      "loss": 5.7182,
      "mean_token_accuracy": 0.20551601350307463,
      "num_tokens": 43800359.0,
      "step": 17280
    },
    {
      "entropy": 6.135875940322876,
      "epoch": 1.994806693594922,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046104440799839145,
      "loss": 5.6637,
      "mean_token_accuracy": 0.20402491688728333,
      "num_tokens": 43812130.0,
      "step": 17285
    },
    {
      "entropy": 6.239100313186645,
      "epoch": 1.9953837276399309,
      "grad_norm": 0.96875,
      "learning_rate": 0.000461021197721003,
      "loss": 5.6995,
      "mean_token_accuracy": 0.20816293358802795,
      "num_tokens": 43823544.0,
      "step": 17290
    },
    {
      "entropy": 6.334615898132324,
      "epoch": 1.9959607616849393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046099798118688407,
      "loss": 5.8108,
      "mean_token_accuracy": 0.19505728483200074,
      "num_tokens": 43836283.0,
      "step": 17295
    },
    {
      "entropy": 6.2541275978088375,
      "epoch": 1.9965377957299482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004609747583968154,
      "loss": 5.7786,
      "mean_token_accuracy": 0.1912152051925659,
      "num_tokens": 43849077.0,
      "step": 17300
    },
    {
      "entropy": 6.270588827133179,
      "epoch": 1.9971148297749566,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004609515293515781,
      "loss": 5.7913,
      "mean_token_accuracy": 0.1918274775147438,
      "num_tokens": 43862805.0,
      "step": 17305
    },
    {
      "entropy": 6.312590503692627,
      "epoch": 1.9976918638199654,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004609282940519535,
      "loss": 5.7857,
      "mean_token_accuracy": 0.1940328821539879,
      "num_tokens": 43875420.0,
      "step": 17310
    },
    {
      "entropy": 6.277432775497436,
      "epoch": 1.998268897864974,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004609050524987231,
      "loss": 5.8083,
      "mean_token_accuracy": 0.18863283842802048,
      "num_tokens": 43888445.0,
      "step": 17315
    },
    {
      "entropy": 6.249401187896728,
      "epoch": 1.9988459319099827,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004608818046926686,
      "loss": 5.8356,
      "mean_token_accuracy": 0.19374081939458848,
      "num_tokens": 43900943.0,
      "step": 17320
    },
    {
      "entropy": 6.219607973098755,
      "epoch": 1.9994229659549914,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004608585506345719,
      "loss": 5.777,
      "mean_token_accuracy": 0.20056314915418624,
      "num_tokens": 43913951.0,
      "step": 17325
    },
    {
      "entropy": 6.273096895217895,
      "epoch": 2.0,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004608352903252152,
      "loss": 5.7256,
      "mean_token_accuracy": 0.20015713274478913,
      "num_tokens": 43926234.0,
      "step": 17330
    },
    {
      "entropy": 6.194000291824341,
      "epoch": 2.000577034045009,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046081202376538084,
      "loss": 5.6258,
      "mean_token_accuracy": 0.21186948120594024,
      "num_tokens": 43939519.0,
      "step": 17335
    },
    {
      "entropy": 6.211329126358033,
      "epoch": 2.0011540680900173,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004607887509558513,
      "loss": 5.6464,
      "mean_token_accuracy": 0.20543332099914552,
      "num_tokens": 43953209.0,
      "step": 17340
    },
    {
      "entropy": 6.226348829269409,
      "epoch": 2.001731102135026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004607654718974094,
      "loss": 5.6429,
      "mean_token_accuracy": 0.2060275450348854,
      "num_tokens": 43965578.0,
      "step": 17345
    },
    {
      "entropy": 6.215400075912475,
      "epoch": 2.0023081361800346,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004607421865908382,
      "loss": 5.6372,
      "mean_token_accuracy": 0.20082560032606125,
      "num_tokens": 43978508.0,
      "step": 17350
    },
    {
      "entropy": 6.2793763160705565,
      "epoch": 2.0028851702250434,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004607188950369207,
      "loss": 5.7484,
      "mean_token_accuracy": 0.19841404408216476,
      "num_tokens": 43992266.0,
      "step": 17355
    },
    {
      "entropy": 6.276701736450195,
      "epoch": 2.003462204270052,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004606955972364405,
      "loss": 5.6919,
      "mean_token_accuracy": 0.1979399397969246,
      "num_tokens": 44004603.0,
      "step": 17360
    },
    {
      "entropy": 6.246250057220459,
      "epoch": 2.0040392383150607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004606722931901812,
      "loss": 5.669,
      "mean_token_accuracy": 0.19785837978124618,
      "num_tokens": 44017578.0,
      "step": 17365
    },
    {
      "entropy": 6.246729755401612,
      "epoch": 2.004616272360069,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004606489828989264,
      "loss": 5.686,
      "mean_token_accuracy": 0.20525402277708055,
      "num_tokens": 44031745.0,
      "step": 17370
    },
    {
      "entropy": 6.271999931335449,
      "epoch": 2.005193306405078,
      "grad_norm": 1.0,
      "learning_rate": 0.0004606256663634604,
      "loss": 5.7244,
      "mean_token_accuracy": 0.19691572338342667,
      "num_tokens": 44043027.0,
      "step": 17375
    },
    {
      "entropy": 6.1975549221038815,
      "epoch": 2.0057703404500864,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004606023435845672,
      "loss": 5.6577,
      "mean_token_accuracy": 0.20978390723466872,
      "num_tokens": 44056211.0,
      "step": 17380
    },
    {
      "entropy": 6.161208963394165,
      "epoch": 2.0063473744950953,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004605790145630314,
      "loss": 5.5525,
      "mean_token_accuracy": 0.2148720934987068,
      "num_tokens": 44069861.0,
      "step": 17385
    },
    {
      "entropy": 6.285813713073731,
      "epoch": 2.0069244085401037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004605556792996377,
      "loss": 5.7107,
      "mean_token_accuracy": 0.19712819010019303,
      "num_tokens": 44083373.0,
      "step": 17390
    },
    {
      "entropy": 6.166922235488892,
      "epoch": 2.0075014425851125,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004605323377951709,
      "loss": 5.6031,
      "mean_token_accuracy": 0.20643806755542754,
      "num_tokens": 44096138.0,
      "step": 17395
    },
    {
      "entropy": 6.20800552368164,
      "epoch": 2.008078476630121,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000460508990050416,
      "loss": 5.6558,
      "mean_token_accuracy": 0.19798593372106552,
      "num_tokens": 44107591.0,
      "step": 17400
    },
    {
      "entropy": 6.256885957717896,
      "epoch": 2.00865551067513,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004604856360661584,
      "loss": 5.6504,
      "mean_token_accuracy": 0.20174759328365327,
      "num_tokens": 44119733.0,
      "step": 17405
    },
    {
      "entropy": 6.327699327468872,
      "epoch": 2.0092325447201387,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004604622758431835,
      "loss": 5.7737,
      "mean_token_accuracy": 0.19579524099826812,
      "num_tokens": 44131967.0,
      "step": 17410
    },
    {
      "entropy": 6.2416300773620605,
      "epoch": 2.009809578765147,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046043890938227724,
      "loss": 5.711,
      "mean_token_accuracy": 0.20428868532180786,
      "num_tokens": 44144426.0,
      "step": 17415
    },
    {
      "entropy": 6.263639831542969,
      "epoch": 2.010386612810156,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046041553668422526,
      "loss": 5.7362,
      "mean_token_accuracy": 0.19565992206335067,
      "num_tokens": 44156298.0,
      "step": 17420
    },
    {
      "entropy": 6.235705995559693,
      "epoch": 2.0109636468551644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046039215774981376,
      "loss": 5.6275,
      "mean_token_accuracy": 0.19949375540018083,
      "num_tokens": 44167955.0,
      "step": 17425
    },
    {
      "entropy": 6.252126789093017,
      "epoch": 2.0115406809001732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046036877257982917,
      "loss": 5.6608,
      "mean_token_accuracy": 0.20291411578655244,
      "num_tokens": 44180555.0,
      "step": 17430
    },
    {
      "entropy": 6.172300386428833,
      "epoch": 2.0121177149451817,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000460345381175058,
      "loss": 5.6699,
      "mean_token_accuracy": 0.20165491849184036,
      "num_tokens": 44194335.0,
      "step": 17435
    },
    {
      "entropy": 6.285435152053833,
      "epoch": 2.0126947489901905,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000460321983536287,
      "loss": 5.6644,
      "mean_token_accuracy": 0.20328864604234695,
      "num_tokens": 44207258.0,
      "step": 17440
    },
    {
      "entropy": 6.2977532863616945,
      "epoch": 2.013271783035199,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046029857966430315,
      "loss": 5.6902,
      "mean_token_accuracy": 0.20235728174448014,
      "num_tokens": 44220424.0,
      "step": 17445
    },
    {
      "entropy": 6.150838851928711,
      "epoch": 2.013848817080208,
      "grad_norm": 0.875,
      "learning_rate": 0.00046027516955989364,
      "loss": 5.6117,
      "mean_token_accuracy": 0.2029151812195778,
      "num_tokens": 44233394.0,
      "step": 17450
    },
    {
      "entropy": 6.307342910766602,
      "epoch": 2.014425851125216,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046025175322384577,
      "loss": 5.6771,
      "mean_token_accuracy": 0.1908987805247307,
      "num_tokens": 44246255.0,
      "step": 17455
    },
    {
      "entropy": 6.277203512191773,
      "epoch": 2.015002885170225,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046022833065694727,
      "loss": 5.6565,
      "mean_token_accuracy": 0.19885320514440535,
      "num_tokens": 44258315.0,
      "step": 17460
    },
    {
      "entropy": 6.25507345199585,
      "epoch": 2.0155799192152335,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046020490185998575,
      "loss": 5.7608,
      "mean_token_accuracy": 0.19464564472436904,
      "num_tokens": 44271667.0,
      "step": 17465
    },
    {
      "entropy": 6.319039726257325,
      "epoch": 2.0161569532602424,
      "grad_norm": 0.875,
      "learning_rate": 0.0004601814668337495,
      "loss": 5.6817,
      "mean_token_accuracy": 0.20105798095464705,
      "num_tokens": 44284831.0,
      "step": 17470
    },
    {
      "entropy": 6.233797597885132,
      "epoch": 2.016733987305251,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046015802557902654,
      "loss": 5.6861,
      "mean_token_accuracy": 0.20210664421319963,
      "num_tokens": 44297161.0,
      "step": 17475
    },
    {
      "entropy": 6.2880847454071045,
      "epoch": 2.0173110213502596,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046013457809660537,
      "loss": 5.6484,
      "mean_token_accuracy": 0.20195046365261077,
      "num_tokens": 44309537.0,
      "step": 17480
    },
    {
      "entropy": 6.20090389251709,
      "epoch": 2.0178880553952685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046011112438727454,
      "loss": 5.65,
      "mean_token_accuracy": 0.20998493134975432,
      "num_tokens": 44320761.0,
      "step": 17485
    },
    {
      "entropy": 6.187840604782105,
      "epoch": 2.018465089440277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004600876644518231,
      "loss": 5.5584,
      "mean_token_accuracy": 0.20834412872791291,
      "num_tokens": 44333323.0,
      "step": 17490
    },
    {
      "entropy": 6.106359767913818,
      "epoch": 2.0190421234852858,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046006419829104,
      "loss": 5.5094,
      "mean_token_accuracy": 0.2134860172867775,
      "num_tokens": 44345919.0,
      "step": 17495
    },
    {
      "entropy": 6.260765838623047,
      "epoch": 2.019619157530294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004600407259057145,
      "loss": 5.7092,
      "mean_token_accuracy": 0.19886721521615983,
      "num_tokens": 44359702.0,
      "step": 17500
    },
    {
      "entropy": 6.198544263839722,
      "epoch": 2.020196191575303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004600172472966361,
      "loss": 5.6236,
      "mean_token_accuracy": 0.204592664539814,
      "num_tokens": 44374186.0,
      "step": 17505
    },
    {
      "entropy": 6.279191780090332,
      "epoch": 2.0207732256203115,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045999376246459446,
      "loss": 5.7695,
      "mean_token_accuracy": 0.19268970042467118,
      "num_tokens": 44387870.0,
      "step": 17510
    },
    {
      "entropy": 6.244718027114868,
      "epoch": 2.0213502596653203,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004599702714103795,
      "loss": 5.6962,
      "mean_token_accuracy": 0.19641393125057222,
      "num_tokens": 44400706.0,
      "step": 17515
    },
    {
      "entropy": 6.3039475440979,
      "epoch": 2.0219272937103288,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004599467741347814,
      "loss": 5.6999,
      "mean_token_accuracy": 0.19415407180786132,
      "num_tokens": 44413166.0,
      "step": 17520
    },
    {
      "entropy": 6.28196439743042,
      "epoch": 2.0225043277553376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004599232706385904,
      "loss": 5.6916,
      "mean_token_accuracy": 0.2029717057943344,
      "num_tokens": 44425265.0,
      "step": 17525
    },
    {
      "entropy": 6.220837926864624,
      "epoch": 2.023081361800346,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000459899760922597,
      "loss": 5.6311,
      "mean_token_accuracy": 0.20344843715429306,
      "num_tokens": 44436990.0,
      "step": 17530
    },
    {
      "entropy": 6.257685852050781,
      "epoch": 2.023658395845355,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004598762449875921,
      "loss": 5.6681,
      "mean_token_accuracy": 0.20381868928670882,
      "num_tokens": 44450503.0,
      "step": 17535
    },
    {
      "entropy": 6.298973178863525,
      "epoch": 2.0242354298903633,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004598527228343665,
      "loss": 5.7201,
      "mean_token_accuracy": 0.19569796919822693,
      "num_tokens": 44463559.0,
      "step": 17540
    },
    {
      "entropy": 6.265443897247314,
      "epoch": 2.024812463935372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004598291944637112,
      "loss": 5.7503,
      "mean_token_accuracy": 0.19749541580677032,
      "num_tokens": 44475064.0,
      "step": 17545
    },
    {
      "entropy": 6.27156662940979,
      "epoch": 2.025389497980381,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045980565987641797,
      "loss": 5.6601,
      "mean_token_accuracy": 0.20068962275981903,
      "num_tokens": 44487094.0,
      "step": 17550
    },
    {
      "entropy": 6.276722097396851,
      "epoch": 2.0259665320253895,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045978211907327804,
      "loss": 5.7003,
      "mean_token_accuracy": 0.19610402584075928,
      "num_tokens": 44498950.0,
      "step": 17555
    },
    {
      "entropy": 6.25366997718811,
      "epoch": 2.0265435660703983,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004597585720550834,
      "loss": 5.7011,
      "mean_token_accuracy": 0.19720872789621352,
      "num_tokens": 44511277.0,
      "step": 17560
    },
    {
      "entropy": 6.1721179485321045,
      "epoch": 2.0271206001154067,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000459735018822626,
      "loss": 5.6405,
      "mean_token_accuracy": 0.20919449925422667,
      "num_tokens": 44525595.0,
      "step": 17565
    },
    {
      "entropy": 6.266058397293091,
      "epoch": 2.0276976341604156,
      "grad_norm": 1.0,
      "learning_rate": 0.00045971145937669794,
      "loss": 5.6985,
      "mean_token_accuracy": 0.19872716963291168,
      "num_tokens": 44537470.0,
      "step": 17570
    },
    {
      "entropy": 6.30381965637207,
      "epoch": 2.028274668205424,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004596878937180917,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19233401566743852,
      "num_tokens": 44550594.0,
      "step": 17575
    },
    {
      "entropy": 6.275610399246216,
      "epoch": 2.028851702250433,
      "grad_norm": 1.0,
      "learning_rate": 0.0004596643218475999,
      "loss": 5.6859,
      "mean_token_accuracy": 0.19564604312181472,
      "num_tokens": 44563204.0,
      "step": 17580
    },
    {
      "entropy": 6.259721040725708,
      "epoch": 2.0294287362954413,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045964074376601534,
      "loss": 5.7196,
      "mean_token_accuracy": 0.19368371069431306,
      "num_tokens": 44577377.0,
      "step": 17585
    },
    {
      "entropy": 6.275659561157227,
      "epoch": 2.03000577034045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045961715947413106,
      "loss": 5.7133,
      "mean_token_accuracy": 0.1982722818851471,
      "num_tokens": 44589658.0,
      "step": 17590
    },
    {
      "entropy": 6.184750699996949,
      "epoch": 2.0305828043854586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004595935689727404,
      "loss": 5.6933,
      "mean_token_accuracy": 0.2018525332212448,
      "num_tokens": 44602477.0,
      "step": 17595
    },
    {
      "entropy": 6.266600847244263,
      "epoch": 2.0311598384304674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004595699722626367,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19513145387172698,
      "num_tokens": 44613734.0,
      "step": 17600
    },
    {
      "entropy": 6.337140893936157,
      "epoch": 2.031736872475476,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045954636934461367,
      "loss": 5.8,
      "mean_token_accuracy": 0.18939160257577897,
      "num_tokens": 44626608.0,
      "step": 17605
    },
    {
      "entropy": 6.17408709526062,
      "epoch": 2.0323139065204847,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004595227602194652,
      "loss": 5.5615,
      "mean_token_accuracy": 0.21548786461353303,
      "num_tokens": 44640982.0,
      "step": 17610
    },
    {
      "entropy": 6.203925275802613,
      "epoch": 2.0328909405654936,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004594991448879853,
      "loss": 5.694,
      "mean_token_accuracy": 0.1996866136789322,
      "num_tokens": 44653546.0,
      "step": 17615
    },
    {
      "entropy": 6.231396579742432,
      "epoch": 2.033467974610502,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004594755233509683,
      "loss": 5.6275,
      "mean_token_accuracy": 0.2021331802010536,
      "num_tokens": 44667555.0,
      "step": 17620
    },
    {
      "entropy": 6.174169206619263,
      "epoch": 2.034045008655511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045945189560920875,
      "loss": 5.5612,
      "mean_token_accuracy": 0.20570155531167983,
      "num_tokens": 44680205.0,
      "step": 17625
    },
    {
      "entropy": 6.2767740249633786,
      "epoch": 2.0346220427005193,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004594282616635013,
      "loss": 5.6921,
      "mean_token_accuracy": 0.19772822111845018,
      "num_tokens": 44692838.0,
      "step": 17630
    },
    {
      "entropy": 6.235586833953858,
      "epoch": 2.035199076745528,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004594046215146409,
      "loss": 5.6408,
      "mean_token_accuracy": 0.20706749558448792,
      "num_tokens": 44706010.0,
      "step": 17635
    },
    {
      "entropy": 6.186456727981567,
      "epoch": 2.0357761107905366,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045938097516342257,
      "loss": 5.6582,
      "mean_token_accuracy": 0.2096608519554138,
      "num_tokens": 44718857.0,
      "step": 17640
    },
    {
      "entropy": 6.281586647033691,
      "epoch": 2.0363531448355454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045935732261064184,
      "loss": 5.6706,
      "mean_token_accuracy": 0.2015170305967331,
      "num_tokens": 44730876.0,
      "step": 17645
    },
    {
      "entropy": 6.1743262767791744,
      "epoch": 2.036930178880554,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045933366385709405,
      "loss": 5.6291,
      "mean_token_accuracy": 0.2040094554424286,
      "num_tokens": 44743236.0,
      "step": 17650
    },
    {
      "entropy": 6.133583211898804,
      "epoch": 2.0375072129255627,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004593099989035751,
      "loss": 5.5866,
      "mean_token_accuracy": 0.2103258788585663,
      "num_tokens": 44755762.0,
      "step": 17655
    },
    {
      "entropy": 6.316069507598877,
      "epoch": 2.038084246970571,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004592863277508809,
      "loss": 5.732,
      "mean_token_accuracy": 0.19772288650274278,
      "num_tokens": 44768439.0,
      "step": 17660
    },
    {
      "entropy": 6.250504970550537,
      "epoch": 2.03866128101558,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004592626503998076,
      "loss": 5.6985,
      "mean_token_accuracy": 0.1941557466983795,
      "num_tokens": 44781751.0,
      "step": 17665
    },
    {
      "entropy": 6.159578418731689,
      "epoch": 2.0392383150605884,
      "grad_norm": 1.0,
      "learning_rate": 0.0004592389668511515,
      "loss": 5.5477,
      "mean_token_accuracy": 0.20575915426015853,
      "num_tokens": 44794221.0,
      "step": 17670
    },
    {
      "entropy": 6.246522855758667,
      "epoch": 2.0398153491055973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004592152771057093,
      "loss": 5.6944,
      "mean_token_accuracy": 0.1989157497882843,
      "num_tokens": 44807027.0,
      "step": 17675
    },
    {
      "entropy": 6.218291854858398,
      "epoch": 2.0403923831506057,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045919158116427785,
      "loss": 5.6983,
      "mean_token_accuracy": 0.2016799196600914,
      "num_tokens": 44820399.0,
      "step": 17680
    },
    {
      "entropy": 6.332528734207154,
      "epoch": 2.0409694171956145,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045916787902765396,
      "loss": 5.762,
      "mean_token_accuracy": 0.1897583410143852,
      "num_tokens": 44833179.0,
      "step": 17685
    },
    {
      "entropy": 6.201383447647094,
      "epoch": 2.0415464512406234,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004591441706966349,
      "loss": 5.5738,
      "mean_token_accuracy": 0.21960131227970123,
      "num_tokens": 44846169.0,
      "step": 17690
    },
    {
      "entropy": 6.125986051559448,
      "epoch": 2.042123485285632,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004591204561720183,
      "loss": 5.5391,
      "mean_token_accuracy": 0.2042707234621048,
      "num_tokens": 44858400.0,
      "step": 17695
    },
    {
      "entropy": 6.233135461807251,
      "epoch": 2.0427005193306407,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004590967354546015,
      "loss": 5.7517,
      "mean_token_accuracy": 0.19689769744873048,
      "num_tokens": 44870291.0,
      "step": 17700
    },
    {
      "entropy": 6.3327044486999515,
      "epoch": 2.043277553375649,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004590730085451824,
      "loss": 5.717,
      "mean_token_accuracy": 0.2040923520922661,
      "num_tokens": 44883893.0,
      "step": 17705
    },
    {
      "entropy": 6.245716619491577,
      "epoch": 2.043854587420658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045904927544455914,
      "loss": 5.6198,
      "mean_token_accuracy": 0.200164096057415,
      "num_tokens": 44896075.0,
      "step": 17710
    },
    {
      "entropy": 6.1517706394195555,
      "epoch": 2.0444316214656664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045902553615353005,
      "loss": 5.6011,
      "mean_token_accuracy": 0.2013479948043823,
      "num_tokens": 44908752.0,
      "step": 17715
    },
    {
      "entropy": 6.2874915599823,
      "epoch": 2.0450086555106752,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004590017906728933,
      "loss": 5.6796,
      "mean_token_accuracy": 0.20257628262042998,
      "num_tokens": 44920782.0,
      "step": 17720
    },
    {
      "entropy": 6.2006621837615965,
      "epoch": 2.0455856895556837,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045897803900344774,
      "loss": 5.592,
      "mean_token_accuracy": 0.20246542245149612,
      "num_tokens": 44934565.0,
      "step": 17725
    },
    {
      "entropy": 6.186515617370605,
      "epoch": 2.0461627236006925,
      "grad_norm": 1.0,
      "learning_rate": 0.0004589542811459923,
      "loss": 5.6192,
      "mean_token_accuracy": 0.20540280342102052,
      "num_tokens": 44948232.0,
      "step": 17730
    },
    {
      "entropy": 6.254387235641479,
      "epoch": 2.046739757645701,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004589305171013259,
      "loss": 5.6347,
      "mean_token_accuracy": 0.20235151946544647,
      "num_tokens": 44960092.0,
      "step": 17735
    },
    {
      "entropy": 6.3298375606536865,
      "epoch": 2.04731679169071,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000458906746870248,
      "loss": 5.7512,
      "mean_token_accuracy": 0.19219062775373458,
      "num_tokens": 44972065.0,
      "step": 17740
    },
    {
      "entropy": 6.261089420318603,
      "epoch": 2.047893825735718,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000458882970453558,
      "loss": 5.7042,
      "mean_token_accuracy": 0.20852959007024766,
      "num_tokens": 44985142.0,
      "step": 17745
    },
    {
      "entropy": 6.279339122772217,
      "epoch": 2.048470859780727,
      "grad_norm": 1.0,
      "learning_rate": 0.0004588591878520556,
      "loss": 5.7578,
      "mean_token_accuracy": 0.19072488993406295,
      "num_tokens": 44998131.0,
      "step": 17750
    },
    {
      "entropy": 6.220921468734741,
      "epoch": 2.049047893825736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004588353990665407,
      "loss": 5.5889,
      "mean_token_accuracy": 0.20640793144702912,
      "num_tokens": 45012715.0,
      "step": 17755
    },
    {
      "entropy": 6.250972270965576,
      "epoch": 2.0496249278707444,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004588116040978136,
      "loss": 5.7555,
      "mean_token_accuracy": 0.1953182712197304,
      "num_tokens": 45025214.0,
      "step": 17760
    },
    {
      "entropy": 6.2365130424499515,
      "epoch": 2.050201961915753,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045878780294667437,
      "loss": 5.5738,
      "mean_token_accuracy": 0.21922473311424256,
      "num_tokens": 45037956.0,
      "step": 17765
    },
    {
      "entropy": 6.242060136795044,
      "epoch": 2.0507789959607616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004587639956139237,
      "loss": 5.7303,
      "mean_token_accuracy": 0.19783318042755127,
      "num_tokens": 45052074.0,
      "step": 17770
    },
    {
      "entropy": 6.252948188781739,
      "epoch": 2.0513560300057705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004587401821003624,
      "loss": 5.6724,
      "mean_token_accuracy": 0.1973973512649536,
      "num_tokens": 45064408.0,
      "step": 17775
    },
    {
      "entropy": 6.170503902435303,
      "epoch": 2.051933064050779,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045871636240679133,
      "loss": 5.5838,
      "mean_token_accuracy": 0.2190801367163658,
      "num_tokens": 45077944.0,
      "step": 17780
    },
    {
      "entropy": 6.20067777633667,
      "epoch": 2.0525100980957878,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004586925365340117,
      "loss": 5.6246,
      "mean_token_accuracy": 0.20417091846466065,
      "num_tokens": 45090682.0,
      "step": 17785
    },
    {
      "entropy": 6.181760978698731,
      "epoch": 2.053087132140796,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004586687044828247,
      "loss": 5.6361,
      "mean_token_accuracy": 0.20489009618759155,
      "num_tokens": 45102567.0,
      "step": 17790
    },
    {
      "entropy": 6.1823687076568605,
      "epoch": 2.053664166185805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004586448662540322,
      "loss": 5.6176,
      "mean_token_accuracy": 0.20492803901433945,
      "num_tokens": 45114585.0,
      "step": 17795
    },
    {
      "entropy": 6.313984107971192,
      "epoch": 2.0542412002308135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004586210218484358,
      "loss": 5.7203,
      "mean_token_accuracy": 0.20136224031448363,
      "num_tokens": 45128491.0,
      "step": 17800
    },
    {
      "entropy": 6.231251192092896,
      "epoch": 2.0548182342758223,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045859717126683745,
      "loss": 5.74,
      "mean_token_accuracy": 0.19338741451501845,
      "num_tokens": 45141646.0,
      "step": 17805
    },
    {
      "entropy": 6.1753373622894285,
      "epoch": 2.0553952683208307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045857331451003953,
      "loss": 5.5714,
      "mean_token_accuracy": 0.21142702400684357,
      "num_tokens": 45153797.0,
      "step": 17810
    },
    {
      "entropy": 6.218253326416016,
      "epoch": 2.0559723023658396,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045854945157884435,
      "loss": 5.6837,
      "mean_token_accuracy": 0.20009419620037078,
      "num_tokens": 45166626.0,
      "step": 17815
    },
    {
      "entropy": 6.224326848983765,
      "epoch": 2.056549336410848,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045852558247405455,
      "loss": 5.647,
      "mean_token_accuracy": 0.2017377108335495,
      "num_tokens": 45181974.0,
      "step": 17820
    },
    {
      "entropy": 6.340387392044067,
      "epoch": 2.057126370455857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045850170719647287,
      "loss": 5.7348,
      "mean_token_accuracy": 0.20140082240104676,
      "num_tokens": 45194406.0,
      "step": 17825
    },
    {
      "entropy": 6.219389533996582,
      "epoch": 2.0577034045008658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045847782574690254,
      "loss": 5.6531,
      "mean_token_accuracy": 0.20172075927257538,
      "num_tokens": 45206777.0,
      "step": 17830
    },
    {
      "entropy": 6.188518714904785,
      "epoch": 2.058280438545874,
      "grad_norm": 1.0,
      "learning_rate": 0.00045845393812614664,
      "loss": 5.6378,
      "mean_token_accuracy": 0.20293809473514557,
      "num_tokens": 45219718.0,
      "step": 17835
    },
    {
      "entropy": 6.3026697635650635,
      "epoch": 2.058857472590883,
      "grad_norm": 1.0,
      "learning_rate": 0.0004584300443350086,
      "loss": 5.691,
      "mean_token_accuracy": 0.2011967644095421,
      "num_tokens": 45231038.0,
      "step": 17840
    },
    {
      "entropy": 6.24705286026001,
      "epoch": 2.0594345066358914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004584061443742922,
      "loss": 5.7316,
      "mean_token_accuracy": 0.19505251199007034,
      "num_tokens": 45242835.0,
      "step": 17845
    },
    {
      "entropy": 6.253177499771118,
      "epoch": 2.0600115406809003,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045838223824480124,
      "loss": 5.7064,
      "mean_token_accuracy": 0.199451245367527,
      "num_tokens": 45254857.0,
      "step": 17850
    },
    {
      "entropy": 6.225717353820801,
      "epoch": 2.0605885747259087,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004583583259473397,
      "loss": 5.6696,
      "mean_token_accuracy": 0.20797770023345946,
      "num_tokens": 45267897.0,
      "step": 17855
    },
    {
      "entropy": 6.245395660400391,
      "epoch": 2.0611656087709176,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045833440748271203,
      "loss": 5.6785,
      "mean_token_accuracy": 0.21114442497491837,
      "num_tokens": 45280209.0,
      "step": 17860
    },
    {
      "entropy": 6.1997260570526125,
      "epoch": 2.061742642815926,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045831048285172266,
      "loss": 5.6612,
      "mean_token_accuracy": 0.21122022122144699,
      "num_tokens": 45291834.0,
      "step": 17865
    },
    {
      "entropy": 6.285340356826782,
      "epoch": 2.062319676860935,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004582865520551762,
      "loss": 5.7766,
      "mean_token_accuracy": 0.19351355582475663,
      "num_tokens": 45305057.0,
      "step": 17870
    },
    {
      "entropy": 6.176423025131226,
      "epoch": 2.0628967109059433,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045826261509387755,
      "loss": 5.6371,
      "mean_token_accuracy": 0.20038788616657258,
      "num_tokens": 45318212.0,
      "step": 17875
    },
    {
      "entropy": 6.325709581375122,
      "epoch": 2.063473744950952,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045823867196863197,
      "loss": 5.77,
      "mean_token_accuracy": 0.19209024012088777,
      "num_tokens": 45330981.0,
      "step": 17880
    },
    {
      "entropy": 6.174748849868775,
      "epoch": 2.0640507789959606,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004582147226802446,
      "loss": 5.6032,
      "mean_token_accuracy": 0.2046228528022766,
      "num_tokens": 45344404.0,
      "step": 17885
    },
    {
      "entropy": 6.151966142654419,
      "epoch": 2.0646278130409694,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004581907672295211,
      "loss": 5.6484,
      "mean_token_accuracy": 0.20869276523590088,
      "num_tokens": 45357708.0,
      "step": 17890
    },
    {
      "entropy": 6.098145818710327,
      "epoch": 2.065204847085978,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045816680561726716,
      "loss": 5.5352,
      "mean_token_accuracy": 0.218269482254982,
      "num_tokens": 45370625.0,
      "step": 17895
    },
    {
      "entropy": 6.24966869354248,
      "epoch": 2.0657818811309867,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004581428378442886,
      "loss": 5.6288,
      "mean_token_accuracy": 0.20364745259284972,
      "num_tokens": 45381864.0,
      "step": 17900
    },
    {
      "entropy": 6.217425107955933,
      "epoch": 2.0663589151759956,
      "grad_norm": 0.875,
      "learning_rate": 0.00045811886391139173,
      "loss": 5.6796,
      "mean_token_accuracy": 0.19824484437704087,
      "num_tokens": 45394624.0,
      "step": 17905
    },
    {
      "entropy": 6.161578035354614,
      "epoch": 2.066935949221004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045809488381938284,
      "loss": 5.5429,
      "mean_token_accuracy": 0.21560531556606294,
      "num_tokens": 45408551.0,
      "step": 17910
    },
    {
      "entropy": 6.289581727981568,
      "epoch": 2.067512983266013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004580708975690684,
      "loss": 5.7304,
      "mean_token_accuracy": 0.19615779519081117,
      "num_tokens": 45420432.0,
      "step": 17915
    },
    {
      "entropy": 6.305605745315551,
      "epoch": 2.0680900173110213,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004580469051612554,
      "loss": 5.8345,
      "mean_token_accuracy": 0.18389061093330383,
      "num_tokens": 45432914.0,
      "step": 17920
    },
    {
      "entropy": 6.253471946716308,
      "epoch": 2.06866705135603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045802290659675057,
      "loss": 5.6407,
      "mean_token_accuracy": 0.2047659322619438,
      "num_tokens": 45445501.0,
      "step": 17925
    },
    {
      "entropy": 6.153088617324829,
      "epoch": 2.0692440854010385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045799890187636123,
      "loss": 5.5754,
      "mean_token_accuracy": 0.20373598635196685,
      "num_tokens": 45458920.0,
      "step": 17930
    },
    {
      "entropy": 6.337396955490112,
      "epoch": 2.0698211194460474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045797489100089464,
      "loss": 5.7796,
      "mean_token_accuracy": 0.1933256357908249,
      "num_tokens": 45471438.0,
      "step": 17935
    },
    {
      "entropy": 6.25494384765625,
      "epoch": 2.070398153491056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004579508739711585,
      "loss": 5.6558,
      "mean_token_accuracy": 0.20210896134376527,
      "num_tokens": 45483956.0,
      "step": 17940
    },
    {
      "entropy": 6.243045330047607,
      "epoch": 2.0709751875360647,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045792685078796075,
      "loss": 5.7526,
      "mean_token_accuracy": 0.19296427965164184,
      "num_tokens": 45496499.0,
      "step": 17945
    },
    {
      "entropy": 6.265172386169434,
      "epoch": 2.071552221581073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000457902821452109,
      "loss": 5.6741,
      "mean_token_accuracy": 0.20286631286144258,
      "num_tokens": 45509069.0,
      "step": 17950
    },
    {
      "entropy": 6.3254681587219235,
      "epoch": 2.072129255626082,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045787878596441193,
      "loss": 5.805,
      "mean_token_accuracy": 0.1901290699839592,
      "num_tokens": 45521707.0,
      "step": 17955
    },
    {
      "entropy": 6.193796300888062,
      "epoch": 2.0727062896710904,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004578547443256776,
      "loss": 5.5911,
      "mean_token_accuracy": 0.21940589249134063,
      "num_tokens": 45533618.0,
      "step": 17960
    },
    {
      "entropy": 6.244109869003296,
      "epoch": 2.0732833237160992,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004578306965367148,
      "loss": 5.6728,
      "mean_token_accuracy": 0.21010765731334685,
      "num_tokens": 45546487.0,
      "step": 17965
    },
    {
      "entropy": 6.250371217727661,
      "epoch": 2.073860357761108,
      "grad_norm": 0.90625,
      "learning_rate": 0.00045780664259833235,
      "loss": 5.6895,
      "mean_token_accuracy": 0.19820765703916549,
      "num_tokens": 45560258.0,
      "step": 17970
    },
    {
      "entropy": 6.307592678070068,
      "epoch": 2.0744373918061165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045778258251133924,
      "loss": 5.7417,
      "mean_token_accuracy": 0.1971103221178055,
      "num_tokens": 45573152.0,
      "step": 17975
    },
    {
      "entropy": 6.385138463973999,
      "epoch": 2.0750144258511254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045775851627654474,
      "loss": 5.7478,
      "mean_token_accuracy": 0.19005440473556517,
      "num_tokens": 45586042.0,
      "step": 17980
    },
    {
      "entropy": 6.349015569686889,
      "epoch": 2.075591459896134,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004577344438947584,
      "loss": 5.7937,
      "mean_token_accuracy": 0.19286778420209885,
      "num_tokens": 45598761.0,
      "step": 17985
    },
    {
      "entropy": 6.200612878799438,
      "epoch": 2.0761684939411427,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045771036536678984,
      "loss": 5.6221,
      "mean_token_accuracy": 0.2056603580713272,
      "num_tokens": 45610988.0,
      "step": 17990
    },
    {
      "entropy": 6.2174876689910885,
      "epoch": 2.076745527986151,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045768628069344885,
      "loss": 5.6207,
      "mean_token_accuracy": 0.21314657479524612,
      "num_tokens": 45623689.0,
      "step": 17995
    },
    {
      "entropy": 6.287812757492065,
      "epoch": 2.07732256203116,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004576621898755455,
      "loss": 5.7087,
      "mean_token_accuracy": 0.20828912109136583,
      "num_tokens": 45637283.0,
      "step": 18000
    },
    {
      "epoch": 2.07732256203116,
      "eval_entropy": 6.0618410021668305,
      "eval_loss": 5.801326751708984,
      "eval_mean_token_accuracy": 0.20273184249448206,
      "eval_num_tokens": 45637283.0,
      "eval_runtime": 17.4769,
      "eval_samples_per_second": 1609.898,
      "eval_steps_per_second": 201.237,
      "step": 18000
    },
    {
      "entropy": 6.28868522644043,
      "epoch": 2.0778995960761684,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045763809291389024,
      "loss": 5.7312,
      "mean_token_accuracy": 0.1916009932756424,
      "num_tokens": 45651139.0,
      "step": 18005
    },
    {
      "entropy": 6.298526048660278,
      "epoch": 2.0784766301211772,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004576139898092933,
      "loss": 5.6915,
      "mean_token_accuracy": 0.19758338779211043,
      "num_tokens": 45663696.0,
      "step": 18010
    },
    {
      "entropy": 6.262426853179932,
      "epoch": 2.0790536641661856,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004575898805625657,
      "loss": 5.6891,
      "mean_token_accuracy": 0.20868945121765137,
      "num_tokens": 45676272.0,
      "step": 18015
    },
    {
      "entropy": 6.2215540409088135,
      "epoch": 2.0796306982111945,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045756576517451804,
      "loss": 5.7004,
      "mean_token_accuracy": 0.20417592227458953,
      "num_tokens": 45688031.0,
      "step": 18020
    },
    {
      "entropy": 6.294672107696533,
      "epoch": 2.080207732256203,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045754164364596156,
      "loss": 5.7672,
      "mean_token_accuracy": 0.19347960650920867,
      "num_tokens": 45700286.0,
      "step": 18025
    },
    {
      "entropy": 6.25861701965332,
      "epoch": 2.080784766301212,
      "grad_norm": 1.25,
      "learning_rate": 0.0004575175159777076,
      "loss": 5.6737,
      "mean_token_accuracy": 0.2059756338596344,
      "num_tokens": 45713935.0,
      "step": 18030
    },
    {
      "entropy": 6.222640943527222,
      "epoch": 2.0813618003462206,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004574933821705676,
      "loss": 5.6693,
      "mean_token_accuracy": 0.19638945013284684,
      "num_tokens": 45726815.0,
      "step": 18035
    },
    {
      "entropy": 6.237837457656861,
      "epoch": 2.081938834391229,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004574692422253534,
      "loss": 5.6471,
      "mean_token_accuracy": 0.20091569125652314,
      "num_tokens": 45738727.0,
      "step": 18040
    },
    {
      "entropy": 6.218392562866211,
      "epoch": 2.082515868436238,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045744509614287687,
      "loss": 5.6804,
      "mean_token_accuracy": 0.1998003140091896,
      "num_tokens": 45751221.0,
      "step": 18045
    },
    {
      "entropy": 6.293504953384399,
      "epoch": 2.0830929024812463,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004574209439239501,
      "loss": 5.6476,
      "mean_token_accuracy": 0.2031890094280243,
      "num_tokens": 45764680.0,
      "step": 18050
    },
    {
      "entropy": 6.25413498878479,
      "epoch": 2.083669936526255,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045739678556938563,
      "loss": 5.6747,
      "mean_token_accuracy": 0.19882526993751526,
      "num_tokens": 45778324.0,
      "step": 18055
    },
    {
      "entropy": 6.1591027736663815,
      "epoch": 2.0842469705712636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045737262107999575,
      "loss": 5.5853,
      "mean_token_accuracy": 0.20381422191858292,
      "num_tokens": 45789412.0,
      "step": 18060
    },
    {
      "entropy": 6.2949779510498045,
      "epoch": 2.0848240046162725,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004573484504565934,
      "loss": 5.8047,
      "mean_token_accuracy": 0.19288643598556518,
      "num_tokens": 45802264.0,
      "step": 18065
    },
    {
      "entropy": 6.28584623336792,
      "epoch": 2.085401038661281,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004573242736999915,
      "loss": 5.6552,
      "mean_token_accuracy": 0.206164214015007,
      "num_tokens": 45815411.0,
      "step": 18070
    },
    {
      "entropy": 6.296340274810791,
      "epoch": 2.0859780727062898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045730009081100314,
      "loss": 5.8108,
      "mean_token_accuracy": 0.1905052751302719,
      "num_tokens": 45828001.0,
      "step": 18075
    },
    {
      "entropy": 6.243520784378052,
      "epoch": 2.086555106751298,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045727590179044195,
      "loss": 5.7247,
      "mean_token_accuracy": 0.20086035430431365,
      "num_tokens": 45841001.0,
      "step": 18080
    },
    {
      "entropy": 6.289536714553833,
      "epoch": 2.087132140796307,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004572517066391211,
      "loss": 5.7516,
      "mean_token_accuracy": 0.19374656230211257,
      "num_tokens": 45855160.0,
      "step": 18085
    },
    {
      "entropy": 6.1252281665802,
      "epoch": 2.0877091748413155,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045722750535785484,
      "loss": 5.5337,
      "mean_token_accuracy": 0.20700814574956894,
      "num_tokens": 45867822.0,
      "step": 18090
    },
    {
      "entropy": 6.297888326644897,
      "epoch": 2.0882862088863243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045720329794745685,
      "loss": 5.7618,
      "mean_token_accuracy": 0.1892239511013031,
      "num_tokens": 45878990.0,
      "step": 18095
    },
    {
      "entropy": 6.262788391113281,
      "epoch": 2.0888632429313327,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004571790844087415,
      "loss": 5.6803,
      "mean_token_accuracy": 0.19784329384565352,
      "num_tokens": 45891411.0,
      "step": 18100
    },
    {
      "entropy": 6.198941898345947,
      "epoch": 2.0894402769763416,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004571548647425231,
      "loss": 5.626,
      "mean_token_accuracy": 0.20518599897623063,
      "num_tokens": 45903589.0,
      "step": 18105
    },
    {
      "entropy": 6.252762317657471,
      "epoch": 2.0900173110213505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004571306389496164,
      "loss": 5.6433,
      "mean_token_accuracy": 0.20128325670957564,
      "num_tokens": 45915454.0,
      "step": 18110
    },
    {
      "entropy": 6.27607102394104,
      "epoch": 2.090594345066359,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045710640703083594,
      "loss": 5.653,
      "mean_token_accuracy": 0.2016550987958908,
      "num_tokens": 45928974.0,
      "step": 18115
    },
    {
      "entropy": 6.306045436859131,
      "epoch": 2.0911713791113677,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045708216898699707,
      "loss": 5.6853,
      "mean_token_accuracy": 0.19986952394247054,
      "num_tokens": 45941040.0,
      "step": 18120
    },
    {
      "entropy": 6.2621222019195555,
      "epoch": 2.091748413156376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045705792481891483,
      "loss": 5.7093,
      "mean_token_accuracy": 0.2000724896788597,
      "num_tokens": 45954760.0,
      "step": 18125
    },
    {
      "entropy": 6.261735582351685,
      "epoch": 2.092325447201385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045703367452740477,
      "loss": 5.7491,
      "mean_token_accuracy": 0.20223553329706193,
      "num_tokens": 45967735.0,
      "step": 18130
    },
    {
      "entropy": 6.3031106948852536,
      "epoch": 2.0929024812463934,
      "grad_norm": 1.0,
      "learning_rate": 0.00045700941811328247,
      "loss": 5.8065,
      "mean_token_accuracy": 0.18789809793233872,
      "num_tokens": 45979318.0,
      "step": 18135
    },
    {
      "entropy": 6.224335527420044,
      "epoch": 2.0934795152914023,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045698515557736374,
      "loss": 5.6726,
      "mean_token_accuracy": 0.19954797476530076,
      "num_tokens": 45992487.0,
      "step": 18140
    },
    {
      "entropy": 6.327079820632934,
      "epoch": 2.0940565493364107,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045696088692046477,
      "loss": 5.734,
      "mean_token_accuracy": 0.1965554729104042,
      "num_tokens": 46004103.0,
      "step": 18145
    },
    {
      "entropy": 6.327071285247802,
      "epoch": 2.0946335833814196,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045693661214340174,
      "loss": 5.7361,
      "mean_token_accuracy": 0.19441136717796326,
      "num_tokens": 46016498.0,
      "step": 18150
    },
    {
      "entropy": 6.238581228256225,
      "epoch": 2.095210617426428,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045691233124699117,
      "loss": 5.7044,
      "mean_token_accuracy": 0.19793362468481063,
      "num_tokens": 46028179.0,
      "step": 18155
    },
    {
      "entropy": 6.26166353225708,
      "epoch": 2.095787651471437,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004568880442320497,
      "loss": 5.6595,
      "mean_token_accuracy": 0.19274355322122574,
      "num_tokens": 46040152.0,
      "step": 18160
    },
    {
      "entropy": 6.257787895202637,
      "epoch": 2.0963646855164453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045686375109939417,
      "loss": 5.7103,
      "mean_token_accuracy": 0.20011239796876906,
      "num_tokens": 46052663.0,
      "step": 18165
    },
    {
      "entropy": 6.183369827270508,
      "epoch": 2.096941719561454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004568394518498417,
      "loss": 5.6618,
      "mean_token_accuracy": 0.20111042261123657,
      "num_tokens": 46066793.0,
      "step": 18170
    },
    {
      "entropy": 6.263807964324951,
      "epoch": 2.0975187536064626,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045681514648420965,
      "loss": 5.7278,
      "mean_token_accuracy": 0.20204240679740906,
      "num_tokens": 46080913.0,
      "step": 18175
    },
    {
      "entropy": 6.2909845352172855,
      "epoch": 2.0980957876514714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004567908350033154,
      "loss": 5.7234,
      "mean_token_accuracy": 0.19252129793167114,
      "num_tokens": 46093344.0,
      "step": 18180
    },
    {
      "entropy": 6.2630698680877686,
      "epoch": 2.0986728216964803,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004567665174079767,
      "loss": 5.652,
      "mean_token_accuracy": 0.2011677071452141,
      "num_tokens": 46106967.0,
      "step": 18185
    },
    {
      "entropy": 6.2989636898040775,
      "epoch": 2.0992498557414887,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045674219369901153,
      "loss": 5.6851,
      "mean_token_accuracy": 0.19925054758787156,
      "num_tokens": 46118564.0,
      "step": 18190
    },
    {
      "entropy": 6.3145936012268065,
      "epoch": 2.0998268897864976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004567178638772379,
      "loss": 5.7863,
      "mean_token_accuracy": 0.189616097509861,
      "num_tokens": 46130057.0,
      "step": 18195
    },
    {
      "entropy": 6.190846967697143,
      "epoch": 2.100403923831506,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004566935279434742,
      "loss": 5.6231,
      "mean_token_accuracy": 0.20957980006933213,
      "num_tokens": 46142730.0,
      "step": 18200
    },
    {
      "entropy": 6.272794342041015,
      "epoch": 2.100980957876515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045666918589853895,
      "loss": 5.6973,
      "mean_token_accuracy": 0.2015947714447975,
      "num_tokens": 46153669.0,
      "step": 18205
    },
    {
      "entropy": 6.274526357650757,
      "epoch": 2.1015579919215233,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045664483774325094,
      "loss": 5.7012,
      "mean_token_accuracy": 0.19649210274219514,
      "num_tokens": 46166472.0,
      "step": 18210
    },
    {
      "entropy": 6.287811183929444,
      "epoch": 2.102135025966532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004566204834784289,
      "loss": 5.7291,
      "mean_token_accuracy": 0.19581724107265472,
      "num_tokens": 46178417.0,
      "step": 18215
    },
    {
      "entropy": 6.229691886901856,
      "epoch": 2.1027120600115405,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00045659612310489225,
      "loss": 5.6373,
      "mean_token_accuracy": 0.20344559103250504,
      "num_tokens": 46191537.0,
      "step": 18220
    },
    {
      "entropy": 6.2027663230896,
      "epoch": 2.1032890940565494,
      "grad_norm": 2.890625,
      "learning_rate": 0.00045657175662346014,
      "loss": 5.5484,
      "mean_token_accuracy": 0.21059780716896057,
      "num_tokens": 46205716.0,
      "step": 18225
    },
    {
      "entropy": 6.30950140953064,
      "epoch": 2.103866128101558,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004565473840349522,
      "loss": 5.7227,
      "mean_token_accuracy": 0.19705621153116226,
      "num_tokens": 46219050.0,
      "step": 18230
    },
    {
      "entropy": 6.30272855758667,
      "epoch": 2.1044431621465667,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045652300534018816,
      "loss": 5.7076,
      "mean_token_accuracy": 0.20135183334350587,
      "num_tokens": 46231045.0,
      "step": 18235
    },
    {
      "entropy": 6.362217426300049,
      "epoch": 2.105020196191575,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000456498620539988,
      "loss": 5.8006,
      "mean_token_accuracy": 0.19347463995218278,
      "num_tokens": 46244629.0,
      "step": 18240
    },
    {
      "entropy": 6.300135946273803,
      "epoch": 2.105597230236584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004564742296351719,
      "loss": 5.7074,
      "mean_token_accuracy": 0.2049865558743477,
      "num_tokens": 46256960.0,
      "step": 18245
    },
    {
      "entropy": 6.197177410125732,
      "epoch": 2.106174264281593,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045644983262656014,
      "loss": 5.6526,
      "mean_token_accuracy": 0.2082975000143051,
      "num_tokens": 46268558.0,
      "step": 18250
    },
    {
      "entropy": 6.2558331966400145,
      "epoch": 2.1067512983266012,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004564254295149735,
      "loss": 5.6997,
      "mean_token_accuracy": 0.20218954384326934,
      "num_tokens": 46281529.0,
      "step": 18255
    },
    {
      "entropy": 6.2200517654418945,
      "epoch": 2.10732833237161,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045640102030123264,
      "loss": 5.673,
      "mean_token_accuracy": 0.20111980736255647,
      "num_tokens": 46293598.0,
      "step": 18260
    },
    {
      "entropy": 6.2302264213562015,
      "epoch": 2.1079053664166185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045637660498615865,
      "loss": 5.6389,
      "mean_token_accuracy": 0.20525590777397157,
      "num_tokens": 46304252.0,
      "step": 18265
    },
    {
      "entropy": 6.275518941879272,
      "epoch": 2.1084824004616274,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004563521835705725,
      "loss": 5.7342,
      "mean_token_accuracy": 0.20362050384283065,
      "num_tokens": 46316295.0,
      "step": 18270
    },
    {
      "entropy": 6.22071533203125,
      "epoch": 2.109059434506636,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045632775605529587,
      "loss": 5.6476,
      "mean_token_accuracy": 0.2018867999315262,
      "num_tokens": 46328929.0,
      "step": 18275
    },
    {
      "entropy": 6.297320795059204,
      "epoch": 2.1096364685516447,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004563033224411501,
      "loss": 5.7266,
      "mean_token_accuracy": 0.1970434606075287,
      "num_tokens": 46340784.0,
      "step": 18280
    },
    {
      "entropy": 6.252612686157226,
      "epoch": 2.110213502596653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004562788827289572,
      "loss": 5.6702,
      "mean_token_accuracy": 0.2006689801812172,
      "num_tokens": 46354188.0,
      "step": 18285
    },
    {
      "entropy": 6.1701353073120115,
      "epoch": 2.110790536641662,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045625443691953914,
      "loss": 5.5704,
      "mean_token_accuracy": 0.21132206469774245,
      "num_tokens": 46366076.0,
      "step": 18290
    },
    {
      "entropy": 6.237895107269287,
      "epoch": 2.1113675706866704,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004562299850137181,
      "loss": 5.6038,
      "mean_token_accuracy": 0.20463516116142272,
      "num_tokens": 46376954.0,
      "step": 18295
    },
    {
      "entropy": 6.234227848052979,
      "epoch": 2.111944604731679,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004562055270123166,
      "loss": 5.6485,
      "mean_token_accuracy": 0.20295715630054473,
      "num_tokens": 46387951.0,
      "step": 18300
    },
    {
      "entropy": 6.2616410732269285,
      "epoch": 2.1125216387766876,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045618106291615715,
      "loss": 5.7406,
      "mean_token_accuracy": 0.20073629319667816,
      "num_tokens": 46399889.0,
      "step": 18305
    },
    {
      "entropy": 6.2183459281921385,
      "epoch": 2.1130986728216965,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004561565927260627,
      "loss": 5.7388,
      "mean_token_accuracy": 0.19874854236841202,
      "num_tokens": 46412420.0,
      "step": 18310
    },
    {
      "entropy": 6.256087827682495,
      "epoch": 2.1136757068667054,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004561321164428561,
      "loss": 5.7217,
      "mean_token_accuracy": 0.19498515874147415,
      "num_tokens": 46424758.0,
      "step": 18315
    },
    {
      "entropy": 6.239375400543213,
      "epoch": 2.114252740911714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004561076340673609,
      "loss": 5.7333,
      "mean_token_accuracy": 0.1950185477733612,
      "num_tokens": 46437163.0,
      "step": 18320
    },
    {
      "entropy": 6.249647712707519,
      "epoch": 2.1148297749567226,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004560831456004003,
      "loss": 5.6212,
      "mean_token_accuracy": 0.2016495406627655,
      "num_tokens": 46450780.0,
      "step": 18325
    },
    {
      "entropy": 6.277923822402954,
      "epoch": 2.115406809001731,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004560586510427981,
      "loss": 5.6951,
      "mean_token_accuracy": 0.20224761217832565,
      "num_tokens": 46463879.0,
      "step": 18330
    },
    {
      "entropy": 6.243698453903198,
      "epoch": 2.11598384304674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004560341503953781,
      "loss": 5.7105,
      "mean_token_accuracy": 0.19896586388349533,
      "num_tokens": 46477624.0,
      "step": 18335
    },
    {
      "entropy": 6.270669269561767,
      "epoch": 2.1165608770917483,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004560096436589643,
      "loss": 5.6949,
      "mean_token_accuracy": 0.19874074310064316,
      "num_tokens": 46489745.0,
      "step": 18340
    },
    {
      "entropy": 6.2555046558380125,
      "epoch": 2.117137911136757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045598513083438115,
      "loss": 5.6488,
      "mean_token_accuracy": 0.20351042598485947,
      "num_tokens": 46501429.0,
      "step": 18345
    },
    {
      "entropy": 6.256388521194458,
      "epoch": 2.1177149451817656,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045596061192245297,
      "loss": 5.7773,
      "mean_token_accuracy": 0.1907344341278076,
      "num_tokens": 46515094.0,
      "step": 18350
    },
    {
      "entropy": 6.246770191192627,
      "epoch": 2.1182919792267745,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004559360869240045,
      "loss": 5.6509,
      "mean_token_accuracy": 0.20008650571107864,
      "num_tokens": 46527659.0,
      "step": 18355
    },
    {
      "entropy": 6.233044385910034,
      "epoch": 2.118869013271783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004559115558398606,
      "loss": 5.6263,
      "mean_token_accuracy": 0.2005867749452591,
      "num_tokens": 46540190.0,
      "step": 18360
    },
    {
      "entropy": 6.341499280929566,
      "epoch": 2.1194460473167918,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004558870186708465,
      "loss": 5.7424,
      "mean_token_accuracy": 0.19382450729608536,
      "num_tokens": 46552349.0,
      "step": 18365
    },
    {
      "entropy": 6.307049655914307,
      "epoch": 2.1200230813618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045586247541778724,
      "loss": 5.7353,
      "mean_token_accuracy": 0.19565657079219817,
      "num_tokens": 46564693.0,
      "step": 18370
    },
    {
      "entropy": 6.177653217315674,
      "epoch": 2.120600115406809,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004558379260815085,
      "loss": 5.621,
      "mean_token_accuracy": 0.2031398132443428,
      "num_tokens": 46577035.0,
      "step": 18375
    },
    {
      "entropy": 6.2531898498535154,
      "epoch": 2.1211771494518175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004558133706628359,
      "loss": 5.6692,
      "mean_token_accuracy": 0.19893585443496703,
      "num_tokens": 46589670.0,
      "step": 18380
    },
    {
      "entropy": 6.366259336471558,
      "epoch": 2.1217541834968263,
      "grad_norm": 1.0,
      "learning_rate": 0.00045578880916259554,
      "loss": 5.7295,
      "mean_token_accuracy": 0.1948225110769272,
      "num_tokens": 46601320.0,
      "step": 18385
    },
    {
      "entropy": 6.273925352096557,
      "epoch": 2.122331217541835,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004557642415816132,
      "loss": 5.6811,
      "mean_token_accuracy": 0.20070391446352004,
      "num_tokens": 46615065.0,
      "step": 18390
    },
    {
      "entropy": 6.210146808624268,
      "epoch": 2.1229082515868436,
      "grad_norm": 1.0,
      "learning_rate": 0.0004557396679207155,
      "loss": 5.6336,
      "mean_token_accuracy": 0.2028342992067337,
      "num_tokens": 46628847.0,
      "step": 18395
    },
    {
      "entropy": 6.265949630737305,
      "epoch": 2.1234852856318525,
      "grad_norm": 1.0,
      "learning_rate": 0.00045571508818072887,
      "loss": 5.6467,
      "mean_token_accuracy": 0.19950714856386184,
      "num_tokens": 46642214.0,
      "step": 18400
    },
    {
      "entropy": 6.293567085266114,
      "epoch": 2.124062319676861,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004556905023624799,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20038115233182907,
      "num_tokens": 46655920.0,
      "step": 18405
    },
    {
      "entropy": 6.276736164093018,
      "epoch": 2.1246393537218697,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045566591046679577,
      "loss": 5.74,
      "mean_token_accuracy": 0.1909123882651329,
      "num_tokens": 46668006.0,
      "step": 18410
    },
    {
      "entropy": 6.175089740753174,
      "epoch": 2.125216387766878,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045564131249450343,
      "loss": 5.5669,
      "mean_token_accuracy": 0.21258433014154435,
      "num_tokens": 46681523.0,
      "step": 18415
    },
    {
      "entropy": 6.304235172271729,
      "epoch": 2.125793421811887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004556167084464302,
      "loss": 5.656,
      "mean_token_accuracy": 0.20029591768980026,
      "num_tokens": 46693904.0,
      "step": 18420
    },
    {
      "entropy": 6.229953956604004,
      "epoch": 2.1263704558568954,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004555920983234038,
      "loss": 5.6934,
      "mean_token_accuracy": 0.19166387915611266,
      "num_tokens": 46705210.0,
      "step": 18425
    },
    {
      "entropy": 6.223976993560791,
      "epoch": 2.1269474899019043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045556748212625183,
      "loss": 5.6352,
      "mean_token_accuracy": 0.20527464747428895,
      "num_tokens": 46717594.0,
      "step": 18430
    },
    {
      "entropy": 6.3230180740356445,
      "epoch": 2.1275245239469127,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004555428598558023,
      "loss": 5.8124,
      "mean_token_accuracy": 0.19022324681282043,
      "num_tokens": 46730363.0,
      "step": 18435
    },
    {
      "entropy": 6.174237775802612,
      "epoch": 2.1281015579919216,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004555182315128833,
      "loss": 5.6656,
      "mean_token_accuracy": 0.2000950887799263,
      "num_tokens": 46742554.0,
      "step": 18440
    },
    {
      "entropy": 6.285221242904663,
      "epoch": 2.12867859203693,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004554935970983234,
      "loss": 5.7417,
      "mean_token_accuracy": 0.20270133316516875,
      "num_tokens": 46754907.0,
      "step": 18445
    },
    {
      "entropy": 6.302076530456543,
      "epoch": 2.129255626081939,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004554689566129509,
      "loss": 5.7203,
      "mean_token_accuracy": 0.19416812807321548,
      "num_tokens": 46767463.0,
      "step": 18450
    },
    {
      "entropy": 6.310379457473755,
      "epoch": 2.1298326601269473,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045544431005759467,
      "loss": 5.7624,
      "mean_token_accuracy": 0.19574179947376252,
      "num_tokens": 46781523.0,
      "step": 18455
    },
    {
      "entropy": 6.279450273513794,
      "epoch": 2.130409694171956,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045541965743308376,
      "loss": 5.744,
      "mean_token_accuracy": 0.20246258825063707,
      "num_tokens": 46796557.0,
      "step": 18460
    },
    {
      "entropy": 6.307262992858886,
      "epoch": 2.130986728216965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004553949987402473,
      "loss": 5.7202,
      "mean_token_accuracy": 0.19556867927312852,
      "num_tokens": 46809142.0,
      "step": 18465
    },
    {
      "entropy": 6.289131546020508,
      "epoch": 2.1315637622619734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004553703339799146,
      "loss": 5.75,
      "mean_token_accuracy": 0.19643975049257278,
      "num_tokens": 46821654.0,
      "step": 18470
    },
    {
      "entropy": 6.199937438964843,
      "epoch": 2.1321407963069823,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004553456631529154,
      "loss": 5.6318,
      "mean_token_accuracy": 0.2066415637731552,
      "num_tokens": 46833262.0,
      "step": 18475
    },
    {
      "entropy": 6.311725378036499,
      "epoch": 2.1327178303519907,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004553209862600794,
      "loss": 5.7396,
      "mean_token_accuracy": 0.19841923862695693,
      "num_tokens": 46846194.0,
      "step": 18480
    },
    {
      "entropy": 6.2138237953186035,
      "epoch": 2.1332948643969996,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004552963033022365,
      "loss": 5.6354,
      "mean_token_accuracy": 0.20588660687208177,
      "num_tokens": 46858395.0,
      "step": 18485
    },
    {
      "entropy": 6.304419469833374,
      "epoch": 2.133871898442008,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00045527161428021706,
      "loss": 5.7163,
      "mean_token_accuracy": 0.205438332259655,
      "num_tokens": 46871609.0,
      "step": 18490
    },
    {
      "entropy": 6.276527547836304,
      "epoch": 2.134448932487017,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004552469191948514,
      "loss": 5.6917,
      "mean_token_accuracy": 0.20007949620485305,
      "num_tokens": 46883177.0,
      "step": 18495
    },
    {
      "entropy": 6.2978020191192625,
      "epoch": 2.1350259665320253,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004552222180469702,
      "loss": 5.6071,
      "mean_token_accuracy": 0.20520227104425431,
      "num_tokens": 46895678.0,
      "step": 18500
    },
    {
      "entropy": 6.238296318054199,
      "epoch": 2.135603000577034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00045519751083740413,
      "loss": 5.6506,
      "mean_token_accuracy": 0.19977222234010697,
      "num_tokens": 46909041.0,
      "step": 18505
    },
    {
      "entropy": 6.247663259506226,
      "epoch": 2.1361800346220425,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045517279756698435,
      "loss": 5.6051,
      "mean_token_accuracy": 0.2110324129462242,
      "num_tokens": 46921386.0,
      "step": 18510
    },
    {
      "entropy": 6.221551132202149,
      "epoch": 2.1367570686670514,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045514807823654203,
      "loss": 5.6646,
      "mean_token_accuracy": 0.2050367072224617,
      "num_tokens": 46933627.0,
      "step": 18515
    },
    {
      "entropy": 6.28520245552063,
      "epoch": 2.13733410271206,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004551233528469086,
      "loss": 5.739,
      "mean_token_accuracy": 0.19368458390235901,
      "num_tokens": 46945108.0,
      "step": 18520
    },
    {
      "entropy": 6.245186710357666,
      "epoch": 2.1379111367570687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045509862139891553,
      "loss": 5.6301,
      "mean_token_accuracy": 0.21137718260288238,
      "num_tokens": 46958366.0,
      "step": 18525
    },
    {
      "entropy": 6.275752162933349,
      "epoch": 2.138488170802077,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045507388389339496,
      "loss": 5.639,
      "mean_token_accuracy": 0.2036982625722885,
      "num_tokens": 46972001.0,
      "step": 18530
    },
    {
      "entropy": 6.229646396636963,
      "epoch": 2.139065204847086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045504914033117866,
      "loss": 5.6519,
      "mean_token_accuracy": 0.20372820198535918,
      "num_tokens": 46984365.0,
      "step": 18535
    },
    {
      "entropy": 6.233608388900757,
      "epoch": 2.139642238892095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045502439071309897,
      "loss": 5.5983,
      "mean_token_accuracy": 0.21546489745378494,
      "num_tokens": 46996838.0,
      "step": 18540
    },
    {
      "entropy": 6.284310817718506,
      "epoch": 2.1402192729371032,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045499963503998835,
      "loss": 5.743,
      "mean_token_accuracy": 0.19546830505132676,
      "num_tokens": 47008979.0,
      "step": 18545
    },
    {
      "entropy": 6.243851613998413,
      "epoch": 2.140796306982112,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004549748733126794,
      "loss": 5.6882,
      "mean_token_accuracy": 0.20074973106384278,
      "num_tokens": 47021092.0,
      "step": 18550
    },
    {
      "entropy": 6.145237064361572,
      "epoch": 2.1413733410271205,
      "grad_norm": 0.921875,
      "learning_rate": 0.000454950105532005,
      "loss": 5.6219,
      "mean_token_accuracy": 0.20671469271183013,
      "num_tokens": 47034277.0,
      "step": 18555
    },
    {
      "entropy": 6.228475093841553,
      "epoch": 2.1419503750721294,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045492533169879816,
      "loss": 5.6513,
      "mean_token_accuracy": 0.20313584953546523,
      "num_tokens": 47046783.0,
      "step": 18560
    },
    {
      "entropy": 6.21305365562439,
      "epoch": 2.142527409117138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045490055181389216,
      "loss": 5.6965,
      "mean_token_accuracy": 0.1951570063829422,
      "num_tokens": 47059358.0,
      "step": 18565
    },
    {
      "entropy": 6.284527730941773,
      "epoch": 2.1431044431621467,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045487576587812046,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19542501270771026,
      "num_tokens": 47071492.0,
      "step": 18570
    },
    {
      "entropy": 6.255841827392578,
      "epoch": 2.143681477207155,
      "grad_norm": 1.0,
      "learning_rate": 0.00045485097389231675,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19911949932575226,
      "num_tokens": 47083196.0,
      "step": 18575
    },
    {
      "entropy": 6.29262056350708,
      "epoch": 2.144258511252164,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004548261758573149,
      "loss": 5.7269,
      "mean_token_accuracy": 0.2044931173324585,
      "num_tokens": 47095193.0,
      "step": 18580
    },
    {
      "entropy": 6.302646541595459,
      "epoch": 2.1448355452971724,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004548013717739489,
      "loss": 5.7115,
      "mean_token_accuracy": 0.20298240184783936,
      "num_tokens": 47107751.0,
      "step": 18585
    },
    {
      "entropy": 6.237378740310669,
      "epoch": 2.145412579342181,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004547765616430531,
      "loss": 5.688,
      "mean_token_accuracy": 0.20369435995817184,
      "num_tokens": 47121336.0,
      "step": 18590
    },
    {
      "entropy": 6.2130636215209964,
      "epoch": 2.14598961338719,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004547517454654619,
      "loss": 5.6793,
      "mean_token_accuracy": 0.2008383244276047,
      "num_tokens": 47135565.0,
      "step": 18595
    },
    {
      "entropy": 6.291384267807007,
      "epoch": 2.1465666474321985,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045472692324201005,
      "loss": 5.6957,
      "mean_token_accuracy": 0.20216792821884155,
      "num_tokens": 47147703.0,
      "step": 18600
    },
    {
      "entropy": 6.293354225158692,
      "epoch": 2.1471436814772074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045470209497353243,
      "loss": 5.6633,
      "mean_token_accuracy": 0.2024347424507141,
      "num_tokens": 47159309.0,
      "step": 18605
    },
    {
      "entropy": 6.30399169921875,
      "epoch": 2.1477207155222158,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004546772606608641,
      "loss": 5.7444,
      "mean_token_accuracy": 0.20062008053064345,
      "num_tokens": 47171275.0,
      "step": 18610
    },
    {
      "entropy": 6.211285257339478,
      "epoch": 2.1482977495672246,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004546524203048404,
      "loss": 5.6571,
      "mean_token_accuracy": 0.20227411836385728,
      "num_tokens": 47183556.0,
      "step": 18615
    },
    {
      "entropy": 6.22853364944458,
      "epoch": 2.148874783612233,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004546275739062967,
      "loss": 5.669,
      "mean_token_accuracy": 0.20271932184696198,
      "num_tokens": 47196863.0,
      "step": 18620
    },
    {
      "entropy": 6.329017543792725,
      "epoch": 2.149451817657242,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004546027214660688,
      "loss": 5.7311,
      "mean_token_accuracy": 0.19531259983778,
      "num_tokens": 47208717.0,
      "step": 18625
    },
    {
      "entropy": 6.247023296356201,
      "epoch": 2.1500288517022503,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004545778629849926,
      "loss": 5.7223,
      "mean_token_accuracy": 0.2007393419742584,
      "num_tokens": 47222404.0,
      "step": 18630
    },
    {
      "entropy": 6.273023414611816,
      "epoch": 2.150605885747259,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004545529984639042,
      "loss": 5.6805,
      "mean_token_accuracy": 0.20067598670721054,
      "num_tokens": 47235005.0,
      "step": 18635
    },
    {
      "entropy": 6.2968450546264645,
      "epoch": 2.1511829197922676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004545281279036398,
      "loss": 5.7481,
      "mean_token_accuracy": 0.20012474209070205,
      "num_tokens": 47248591.0,
      "step": 18640
    },
    {
      "entropy": 6.226691770553589,
      "epoch": 2.1517599538372765,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004545032513050361,
      "loss": 5.6341,
      "mean_token_accuracy": 0.1996340572834015,
      "num_tokens": 47261231.0,
      "step": 18645
    },
    {
      "entropy": 6.282091283798218,
      "epoch": 2.152336987882285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004544783686689296,
      "loss": 5.6738,
      "mean_token_accuracy": 0.1930384963750839,
      "num_tokens": 47272891.0,
      "step": 18650
    },
    {
      "entropy": 6.313755941390991,
      "epoch": 2.1529140219272938,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045445347999615736,
      "loss": 5.7131,
      "mean_token_accuracy": 0.19727104604244233,
      "num_tokens": 47284730.0,
      "step": 18655
    },
    {
      "entropy": 6.310417556762696,
      "epoch": 2.153491055972302,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045442858528755653,
      "loss": 5.6751,
      "mean_token_accuracy": 0.19848490059375762,
      "num_tokens": 47297485.0,
      "step": 18660
    },
    {
      "entropy": 6.122275304794312,
      "epoch": 2.154068090017311,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045440368454396435,
      "loss": 5.6044,
      "mean_token_accuracy": 0.20649804025888444,
      "num_tokens": 47310925.0,
      "step": 18665
    },
    {
      "entropy": 6.22798523902893,
      "epoch": 2.15464512406232,
      "grad_norm": 1.0,
      "learning_rate": 0.0004543787777662183,
      "loss": 5.671,
      "mean_token_accuracy": 0.1950157806277275,
      "num_tokens": 47322645.0,
      "step": 18670
    },
    {
      "entropy": 6.286021995544433,
      "epoch": 2.1552221581073283,
      "grad_norm": 1.0,
      "learning_rate": 0.00045435386495515617,
      "loss": 5.7249,
      "mean_token_accuracy": 0.20187656581401825,
      "num_tokens": 47334582.0,
      "step": 18675
    },
    {
      "entropy": 6.1864299297332765,
      "epoch": 2.155799192152337,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004543289461116159,
      "loss": 5.6039,
      "mean_token_accuracy": 0.2020887851715088,
      "num_tokens": 47347792.0,
      "step": 18680
    },
    {
      "entropy": 6.2291182518005375,
      "epoch": 2.1563762261973456,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004543040212364356,
      "loss": 5.6996,
      "mean_token_accuracy": 0.19588663429021835,
      "num_tokens": 47361064.0,
      "step": 18685
    },
    {
      "entropy": 6.2985340595245365,
      "epoch": 2.1569532602423545,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004542790903304536,
      "loss": 5.7183,
      "mean_token_accuracy": 0.19610128402709961,
      "num_tokens": 47374329.0,
      "step": 18690
    },
    {
      "entropy": 6.242371416091919,
      "epoch": 2.157530294287363,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004542541533945085,
      "loss": 5.701,
      "mean_token_accuracy": 0.19981221705675126,
      "num_tokens": 47387755.0,
      "step": 18695
    },
    {
      "entropy": 6.254823017120361,
      "epoch": 2.1581073283323717,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045422921042943885,
      "loss": 5.717,
      "mean_token_accuracy": 0.19312427043914795,
      "num_tokens": 47400970.0,
      "step": 18700
    },
    {
      "entropy": 6.142841243743897,
      "epoch": 2.15868436237738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000454204261436084,
      "loss": 5.566,
      "mean_token_accuracy": 0.2102985680103302,
      "num_tokens": 47414091.0,
      "step": 18705
    },
    {
      "entropy": 6.1790680408477785,
      "epoch": 2.159261396422389,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045417930641528255,
      "loss": 5.4555,
      "mean_token_accuracy": 0.23294289112091066,
      "num_tokens": 47428485.0,
      "step": 18710
    },
    {
      "entropy": 6.248079776763916,
      "epoch": 2.1598384304673974,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045415434536787424,
      "loss": 5.7336,
      "mean_token_accuracy": 0.19813150763511658,
      "num_tokens": 47442474.0,
      "step": 18715
    },
    {
      "entropy": 6.224754095077515,
      "epoch": 2.1604154645124063,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004541293782946985,
      "loss": 5.6672,
      "mean_token_accuracy": 0.20551702976226807,
      "num_tokens": 47456286.0,
      "step": 18720
    },
    {
      "entropy": 6.213001012802124,
      "epoch": 2.1609924985574147,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004541044051965952,
      "loss": 5.6239,
      "mean_token_accuracy": 0.21122219711542128,
      "num_tokens": 47468946.0,
      "step": 18725
    },
    {
      "entropy": 6.310295629501343,
      "epoch": 2.1615695326024236,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004540794260744041,
      "loss": 5.6047,
      "mean_token_accuracy": 0.20749778300523758,
      "num_tokens": 47481877.0,
      "step": 18730
    },
    {
      "entropy": 6.238765621185303,
      "epoch": 2.162146566647432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004540544409289655,
      "loss": 5.6579,
      "mean_token_accuracy": 0.20858812630176543,
      "num_tokens": 47493564.0,
      "step": 18735
    },
    {
      "entropy": 6.213624429702759,
      "epoch": 2.162723600692441,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004540294497611197,
      "loss": 5.7315,
      "mean_token_accuracy": 0.1966701254248619,
      "num_tokens": 47506565.0,
      "step": 18740
    },
    {
      "entropy": 6.264366149902344,
      "epoch": 2.1633006347374497,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045400445257170725,
      "loss": 5.6561,
      "mean_token_accuracy": 0.20192568004131317,
      "num_tokens": 47518384.0,
      "step": 18745
    },
    {
      "entropy": 6.196795129776001,
      "epoch": 2.163877668782458,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004539794493615689,
      "loss": 5.6292,
      "mean_token_accuracy": 0.20243414342403412,
      "num_tokens": 47530260.0,
      "step": 18750
    },
    {
      "entropy": 6.2966076850891115,
      "epoch": 2.164454702827467,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004539544401315458,
      "loss": 5.7374,
      "mean_token_accuracy": 0.19430261552333833,
      "num_tokens": 47544651.0,
      "step": 18755
    },
    {
      "entropy": 6.297804117202759,
      "epoch": 2.1650317368724754,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000453929424882479,
      "loss": 5.7097,
      "mean_token_accuracy": 0.19538832008838652,
      "num_tokens": 47556854.0,
      "step": 18760
    },
    {
      "entropy": 6.268271732330322,
      "epoch": 2.1656087709174843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045390440361520987,
      "loss": 5.7241,
      "mean_token_accuracy": 0.19777745604515076,
      "num_tokens": 47569372.0,
      "step": 18765
    },
    {
      "entropy": 6.232343482971191,
      "epoch": 2.1661858049624927,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004538793763305799,
      "loss": 5.6616,
      "mean_token_accuracy": 0.19672017097473143,
      "num_tokens": 47581111.0,
      "step": 18770
    },
    {
      "entropy": 6.218725872039795,
      "epoch": 2.1667628390075016,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045385434302943104,
      "loss": 5.6949,
      "mean_token_accuracy": 0.20155950784683227,
      "num_tokens": 47593224.0,
      "step": 18775
    },
    {
      "entropy": 6.279178619384766,
      "epoch": 2.16733987305251,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004538293037126052,
      "loss": 5.6193,
      "mean_token_accuracy": 0.21189335584640503,
      "num_tokens": 47606959.0,
      "step": 18780
    },
    {
      "entropy": 6.225417518615723,
      "epoch": 2.167916907097519,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045380425838094444,
      "loss": 5.6624,
      "mean_token_accuracy": 0.1994910717010498,
      "num_tokens": 47620159.0,
      "step": 18785
    },
    {
      "entropy": 6.298720932006836,
      "epoch": 2.1684939411425272,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045377920703529133,
      "loss": 5.7613,
      "mean_token_accuracy": 0.193289178609848,
      "num_tokens": 47632325.0,
      "step": 18790
    },
    {
      "entropy": 6.265146207809448,
      "epoch": 2.169070975187536,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004537541496764885,
      "loss": 5.6245,
      "mean_token_accuracy": 0.2127773344516754,
      "num_tokens": 47644247.0,
      "step": 18795
    },
    {
      "entropy": 6.244962692260742,
      "epoch": 2.1696480092325445,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004537290863053786,
      "loss": 5.7176,
      "mean_token_accuracy": 0.2019757241010666,
      "num_tokens": 47655724.0,
      "step": 18800
    },
    {
      "entropy": 6.156143569946289,
      "epoch": 2.1702250432775534,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045370401692280455,
      "loss": 5.6246,
      "mean_token_accuracy": 0.19927211254835128,
      "num_tokens": 47667298.0,
      "step": 18805
    },
    {
      "entropy": 6.246666622161865,
      "epoch": 2.170802077322562,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045367894152960976,
      "loss": 5.6839,
      "mean_token_accuracy": 0.20487935543060304,
      "num_tokens": 47680393.0,
      "step": 18810
    },
    {
      "entropy": 6.165303516387939,
      "epoch": 2.1713791113675707,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045365386012663744,
      "loss": 5.6292,
      "mean_token_accuracy": 0.20821626633405685,
      "num_tokens": 47691766.0,
      "step": 18815
    },
    {
      "entropy": 6.282140684127808,
      "epoch": 2.1719561454125795,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045362877271473137,
      "loss": 5.732,
      "mean_token_accuracy": 0.1970954343676567,
      "num_tokens": 47704710.0,
      "step": 18820
    },
    {
      "entropy": 6.221932697296142,
      "epoch": 2.172533179457588,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045360367929473517,
      "loss": 5.6142,
      "mean_token_accuracy": 0.20292829871177673,
      "num_tokens": 47716869.0,
      "step": 18825
    },
    {
      "entropy": 6.202327156066895,
      "epoch": 2.173110213502597,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000453578579867493,
      "loss": 5.6419,
      "mean_token_accuracy": 0.1979321852326393,
      "num_tokens": 47729294.0,
      "step": 18830
    },
    {
      "entropy": 6.2667515754699705,
      "epoch": 2.1736872475476052,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045355347443384896,
      "loss": 5.7186,
      "mean_token_accuracy": 0.19669476002454758,
      "num_tokens": 47741130.0,
      "step": 18835
    },
    {
      "entropy": 6.233776521682739,
      "epoch": 2.174264281592614,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045352836299464755,
      "loss": 5.7159,
      "mean_token_accuracy": 0.19977934360504152,
      "num_tokens": 47753337.0,
      "step": 18840
    },
    {
      "entropy": 6.2708995819091795,
      "epoch": 2.1748413156376225,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004535032455507333,
      "loss": 5.6679,
      "mean_token_accuracy": 0.19924899637699128,
      "num_tokens": 47766105.0,
      "step": 18845
    },
    {
      "entropy": 6.270053482055664,
      "epoch": 2.1754183496826314,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045347812210295107,
      "loss": 5.6352,
      "mean_token_accuracy": 0.20986029505729675,
      "num_tokens": 47777887.0,
      "step": 18850
    },
    {
      "entropy": 6.275932168960571,
      "epoch": 2.17599538372764,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045345299265214583,
      "loss": 5.7206,
      "mean_token_accuracy": 0.20390598326921464,
      "num_tokens": 47789860.0,
      "step": 18855
    },
    {
      "entropy": 6.273596382141113,
      "epoch": 2.1765724177726486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045342785719916284,
      "loss": 5.6659,
      "mean_token_accuracy": 0.19911953657865525,
      "num_tokens": 47801987.0,
      "step": 18860
    },
    {
      "entropy": 6.327261590957642,
      "epoch": 2.177149451817657,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045340271574484755,
      "loss": 5.7524,
      "mean_token_accuracy": 0.19749284535646439,
      "num_tokens": 47814586.0,
      "step": 18865
    },
    {
      "entropy": 6.212128210067749,
      "epoch": 2.177726485862666,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004533775682900454,
      "loss": 5.5931,
      "mean_token_accuracy": 0.21043590158224107,
      "num_tokens": 47827138.0,
      "step": 18870
    },
    {
      "entropy": 6.302350044250488,
      "epoch": 2.1783035199076743,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004533524148356025,
      "loss": 5.7495,
      "mean_token_accuracy": 0.19709007292985917,
      "num_tokens": 47839748.0,
      "step": 18875
    },
    {
      "entropy": 6.10845308303833,
      "epoch": 2.178880553952683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004533272553823646,
      "loss": 5.5122,
      "mean_token_accuracy": 0.21126923710107803,
      "num_tokens": 47852542.0,
      "step": 18880
    },
    {
      "entropy": 6.263958311080932,
      "epoch": 2.179457587997692,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045330208993117816,
      "loss": 5.6829,
      "mean_token_accuracy": 0.2061159685254097,
      "num_tokens": 47866192.0,
      "step": 18885
    },
    {
      "entropy": 6.189452505111694,
      "epoch": 2.1800346220427005,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004532769184828894,
      "loss": 5.6048,
      "mean_token_accuracy": 0.2127915531396866,
      "num_tokens": 47878164.0,
      "step": 18890
    },
    {
      "entropy": 6.217138195037842,
      "epoch": 2.1806116560877093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004532517410383451,
      "loss": 5.6544,
      "mean_token_accuracy": 0.19837775379419326,
      "num_tokens": 47890040.0,
      "step": 18895
    },
    {
      "entropy": 6.258141040802002,
      "epoch": 2.1811886901327178,
      "grad_norm": 0.953125,
      "learning_rate": 0.000453226557598392,
      "loss": 5.6892,
      "mean_token_accuracy": 0.19527127593755722,
      "num_tokens": 47902949.0,
      "step": 18900
    },
    {
      "entropy": 6.275432014465332,
      "epoch": 2.1817657241777266,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004532013681638771,
      "loss": 5.6886,
      "mean_token_accuracy": 0.19948685616254808,
      "num_tokens": 47915674.0,
      "step": 18905
    },
    {
      "entropy": 6.223680925369263,
      "epoch": 2.182342758222735,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004531761727356478,
      "loss": 5.5729,
      "mean_token_accuracy": 0.21355077177286147,
      "num_tokens": 47929358.0,
      "step": 18910
    },
    {
      "entropy": 6.319423770904541,
      "epoch": 2.182919792267744,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004531509713145514,
      "loss": 5.7252,
      "mean_token_accuracy": 0.2019405335187912,
      "num_tokens": 47940790.0,
      "step": 18915
    },
    {
      "entropy": 6.284353590011596,
      "epoch": 2.1834968263127523,
      "grad_norm": 1.0,
      "learning_rate": 0.00045312576390143557,
      "loss": 5.6894,
      "mean_token_accuracy": 0.20282304286956787,
      "num_tokens": 47952651.0,
      "step": 18920
    },
    {
      "entropy": 6.320902490615845,
      "epoch": 2.184073860357761,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045310055049714815,
      "loss": 5.7058,
      "mean_token_accuracy": 0.1965101346373558,
      "num_tokens": 47964378.0,
      "step": 18925
    },
    {
      "entropy": 6.345327281951905,
      "epoch": 2.1846508944027696,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045307533110253726,
      "loss": 5.7993,
      "mean_token_accuracy": 0.19440625458955765,
      "num_tokens": 47976724.0,
      "step": 18930
    },
    {
      "entropy": 6.28765459060669,
      "epoch": 2.1852279284477785,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00045305010571845096,
      "loss": 5.7235,
      "mean_token_accuracy": 0.19789077192544938,
      "num_tokens": 47989364.0,
      "step": 18935
    },
    {
      "entropy": 6.204644727706909,
      "epoch": 2.185804962492787,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004530248743457378,
      "loss": 5.6966,
      "mean_token_accuracy": 0.2033308982849121,
      "num_tokens": 48001845.0,
      "step": 18940
    },
    {
      "entropy": 6.25111494064331,
      "epoch": 2.1863819965377957,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004529996369852465,
      "loss": 5.7554,
      "mean_token_accuracy": 0.19693850427865983,
      "num_tokens": 48016789.0,
      "step": 18945
    },
    {
      "entropy": 6.361500978469849,
      "epoch": 2.1869590305828046,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045297439363782576,
      "loss": 5.7593,
      "mean_token_accuracy": 0.186754110455513,
      "num_tokens": 48029016.0,
      "step": 18950
    },
    {
      "entropy": 6.151066541671753,
      "epoch": 2.187536064627813,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004529491443043247,
      "loss": 5.6131,
      "mean_token_accuracy": 0.20860934406518936,
      "num_tokens": 48041998.0,
      "step": 18955
    },
    {
      "entropy": 6.276990461349487,
      "epoch": 2.188113098672822,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004529238889855926,
      "loss": 5.6768,
      "mean_token_accuracy": 0.2061452731490135,
      "num_tokens": 48054716.0,
      "step": 18960
    },
    {
      "entropy": 6.238919019699097,
      "epoch": 2.1886901327178303,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004528986276824789,
      "loss": 5.6371,
      "mean_token_accuracy": 0.2060963839292526,
      "num_tokens": 48068444.0,
      "step": 18965
    },
    {
      "entropy": 6.261365222930908,
      "epoch": 2.189267166762839,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004528733603958331,
      "loss": 5.6417,
      "mean_token_accuracy": 0.20814335495233535,
      "num_tokens": 48081494.0,
      "step": 18970
    },
    {
      "entropy": 6.2345024108886715,
      "epoch": 2.1898442008078476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004528480871265053,
      "loss": 5.6654,
      "mean_token_accuracy": 0.20001519322395325,
      "num_tokens": 48094092.0,
      "step": 18975
    },
    {
      "entropy": 6.197571468353272,
      "epoch": 2.1904212348528564,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045282280787534537,
      "loss": 5.603,
      "mean_token_accuracy": 0.2113773763179779,
      "num_tokens": 48106061.0,
      "step": 18980
    },
    {
      "entropy": 6.219502210617065,
      "epoch": 2.190998268897865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004527975226432036,
      "loss": 5.625,
      "mean_token_accuracy": 0.20656011402606964,
      "num_tokens": 48118686.0,
      "step": 18985
    },
    {
      "entropy": 6.2857430934906,
      "epoch": 2.1915753029428737,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045277223143093057,
      "loss": 5.6722,
      "mean_token_accuracy": 0.20431246161460875,
      "num_tokens": 48130946.0,
      "step": 18990
    },
    {
      "entropy": 6.157081365585327,
      "epoch": 2.192152336987882,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045274693423937674,
      "loss": 5.5501,
      "mean_token_accuracy": 0.21694402694702147,
      "num_tokens": 48143923.0,
      "step": 18995
    },
    {
      "entropy": 6.302501821517945,
      "epoch": 2.192729371032891,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045272163106939314,
      "loss": 5.7226,
      "mean_token_accuracy": 0.19368312805891036,
      "num_tokens": 48156089.0,
      "step": 19000
    },
    {
      "entropy": 6.252831029891968,
      "epoch": 2.1933064050778994,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045269632192183076,
      "loss": 5.6113,
      "mean_token_accuracy": 0.20662413984537126,
      "num_tokens": 48169930.0,
      "step": 19005
    },
    {
      "entropy": 6.2990837574005125,
      "epoch": 2.1938834391229083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045267100679754075,
      "loss": 5.6962,
      "mean_token_accuracy": 0.1961020290851593,
      "num_tokens": 48181378.0,
      "step": 19010
    },
    {
      "entropy": 6.209407949447632,
      "epoch": 2.1944604731679167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004526456856973748,
      "loss": 5.7024,
      "mean_token_accuracy": 0.20240456908941268,
      "num_tokens": 48193141.0,
      "step": 19015
    },
    {
      "entropy": 6.204896640777588,
      "epoch": 2.1950375072129256,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004526203586221844,
      "loss": 5.5881,
      "mean_token_accuracy": 0.2130624085664749,
      "num_tokens": 48206817.0,
      "step": 19020
    },
    {
      "entropy": 6.307939291000366,
      "epoch": 2.1956145412579344,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045259502557282145,
      "loss": 5.6867,
      "mean_token_accuracy": 0.20451921075582505,
      "num_tokens": 48220104.0,
      "step": 19025
    },
    {
      "entropy": 6.279133415222168,
      "epoch": 2.196191575302943,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004525696865501381,
      "loss": 5.7114,
      "mean_token_accuracy": 0.1997212588787079,
      "num_tokens": 48232446.0,
      "step": 19030
    },
    {
      "entropy": 6.240639019012451,
      "epoch": 2.1967686093479517,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004525443415549865,
      "loss": 5.7317,
      "mean_token_accuracy": 0.19849334806203842,
      "num_tokens": 48244071.0,
      "step": 19035
    },
    {
      "entropy": 6.306009721755982,
      "epoch": 2.19734564339296,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045251899058821915,
      "loss": 5.7031,
      "mean_token_accuracy": 0.19233881533145905,
      "num_tokens": 48256808.0,
      "step": 19040
    },
    {
      "entropy": 6.276043224334717,
      "epoch": 2.197922677437969,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004524936336506887,
      "loss": 5.5768,
      "mean_token_accuracy": 0.21169555485248565,
      "num_tokens": 48269635.0,
      "step": 19045
    },
    {
      "entropy": 6.2518473148345945,
      "epoch": 2.1984997114829774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004524682707432482,
      "loss": 5.6898,
      "mean_token_accuracy": 0.19934749901294707,
      "num_tokens": 48282211.0,
      "step": 19050
    },
    {
      "entropy": 6.264818429946899,
      "epoch": 2.1990767455279863,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045244290186675034,
      "loss": 5.7374,
      "mean_token_accuracy": 0.19707529842853547,
      "num_tokens": 48295617.0,
      "step": 19055
    },
    {
      "entropy": 6.329167270660401,
      "epoch": 2.1996537795729947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004524175270220489,
      "loss": 5.7192,
      "mean_token_accuracy": 0.19702392667531968,
      "num_tokens": 48307885.0,
      "step": 19060
    },
    {
      "entropy": 6.274250888824463,
      "epoch": 2.2002308136180035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045239214620999683,
      "loss": 5.6627,
      "mean_token_accuracy": 0.20779596120119095,
      "num_tokens": 48320451.0,
      "step": 19065
    },
    {
      "entropy": 6.2617395401000975,
      "epoch": 2.200807847663012,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004523667594314481,
      "loss": 5.6355,
      "mean_token_accuracy": 0.2021152526140213,
      "num_tokens": 48333283.0,
      "step": 19070
    },
    {
      "entropy": 6.336185550689697,
      "epoch": 2.201384881708021,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00045234136668725655,
      "loss": 5.7638,
      "mean_token_accuracy": 0.1943548083305359,
      "num_tokens": 48347197.0,
      "step": 19075
    },
    {
      "entropy": 6.341498899459839,
      "epoch": 2.2019619157530292,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045231596797827616,
      "loss": 5.7902,
      "mean_token_accuracy": 0.19257287085056304,
      "num_tokens": 48359356.0,
      "step": 19080
    },
    {
      "entropy": 6.238107395172119,
      "epoch": 2.202538949798038,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045229056330536134,
      "loss": 5.6799,
      "mean_token_accuracy": 0.20391951948404313,
      "num_tokens": 48373343.0,
      "step": 19085
    },
    {
      "entropy": 6.221538734436035,
      "epoch": 2.2031159838430465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045226515266936644,
      "loss": 5.6084,
      "mean_token_accuracy": 0.20651209652423858,
      "num_tokens": 48385614.0,
      "step": 19090
    },
    {
      "entropy": 6.2835705280303955,
      "epoch": 2.2036930178880554,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004522397360711462,
      "loss": 5.6729,
      "mean_token_accuracy": 0.20113176852464676,
      "num_tokens": 48397814.0,
      "step": 19095
    },
    {
      "entropy": 6.248925256729126,
      "epoch": 2.2042700519330642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004522143135115555,
      "loss": 5.68,
      "mean_token_accuracy": 0.21124763935804367,
      "num_tokens": 48411092.0,
      "step": 19100
    },
    {
      "entropy": 6.241598129272461,
      "epoch": 2.2048470859780727,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045218888499144933,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20067670643329621,
      "num_tokens": 48423203.0,
      "step": 19105
    },
    {
      "entropy": 6.267763710021972,
      "epoch": 2.2054241200230815,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045216345051168314,
      "loss": 5.68,
      "mean_token_accuracy": 0.20845802426338195,
      "num_tokens": 48435562.0,
      "step": 19110
    },
    {
      "entropy": 6.254148292541504,
      "epoch": 2.20600115406809,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004521380100731122,
      "loss": 5.6786,
      "mean_token_accuracy": 0.19906038343906401,
      "num_tokens": 48449021.0,
      "step": 19115
    },
    {
      "entropy": 6.201479148864746,
      "epoch": 2.206578188113099,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045211256367659234,
      "loss": 5.6201,
      "mean_token_accuracy": 0.20985983461141586,
      "num_tokens": 48462411.0,
      "step": 19120
    },
    {
      "entropy": 6.295700883865356,
      "epoch": 2.207155222158107,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004520871113229793,
      "loss": 5.7067,
      "mean_token_accuracy": 0.1943223148584366,
      "num_tokens": 48474397.0,
      "step": 19125
    },
    {
      "entropy": 6.339552354812622,
      "epoch": 2.207732256203116,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00045206165301312927,
      "loss": 5.7408,
      "mean_token_accuracy": 0.20293704867362977,
      "num_tokens": 48490549.0,
      "step": 19130
    },
    {
      "entropy": 6.302863502502442,
      "epoch": 2.2083092902481245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004520361887478985,
      "loss": 5.7037,
      "mean_token_accuracy": 0.2045843243598938,
      "num_tokens": 48503692.0,
      "step": 19135
    },
    {
      "entropy": 6.193946981430054,
      "epoch": 2.2088863242931334,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004520107185281435,
      "loss": 5.6408,
      "mean_token_accuracy": 0.20386012345552446,
      "num_tokens": 48515574.0,
      "step": 19140
    },
    {
      "entropy": 6.274809741973877,
      "epoch": 2.209463358338142,
      "grad_norm": 1.0,
      "learning_rate": 0.0004519852423547208,
      "loss": 5.7523,
      "mean_token_accuracy": 0.19230310767889022,
      "num_tokens": 48527832.0,
      "step": 19145
    },
    {
      "entropy": 6.273365688323975,
      "epoch": 2.2100403923831506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004519597602284875,
      "loss": 5.666,
      "mean_token_accuracy": 0.20533026605844498,
      "num_tokens": 48539558.0,
      "step": 19150
    },
    {
      "entropy": 6.21296010017395,
      "epoch": 2.210617426428159,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004519342721503005,
      "loss": 5.6615,
      "mean_token_accuracy": 0.20489732325077056,
      "num_tokens": 48553360.0,
      "step": 19155
    },
    {
      "entropy": 6.291312265396118,
      "epoch": 2.211194460473168,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004519087781210171,
      "loss": 5.7077,
      "mean_token_accuracy": 0.1957475036382675,
      "num_tokens": 48567416.0,
      "step": 19160
    },
    {
      "entropy": 6.274664068222046,
      "epoch": 2.2117714945181763,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000451883278141495,
      "loss": 5.6947,
      "mean_token_accuracy": 0.20433863997459412,
      "num_tokens": 48580086.0,
      "step": 19165
    },
    {
      "entropy": 6.242549419403076,
      "epoch": 2.212348528563185,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00045185777221259157,
      "loss": 5.5993,
      "mean_token_accuracy": 0.21109416633844375,
      "num_tokens": 48591758.0,
      "step": 19170
    },
    {
      "entropy": 6.060356712341308,
      "epoch": 2.212925562608194,
      "grad_norm": 1.25,
      "learning_rate": 0.0004518322603351648,
      "loss": 5.496,
      "mean_token_accuracy": 0.22072903364896773,
      "num_tokens": 48604994.0,
      "step": 19175
    },
    {
      "entropy": 6.25623631477356,
      "epoch": 2.2135025966532025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045180674251007287,
      "loss": 5.7411,
      "mean_token_accuracy": 0.1990656226873398,
      "num_tokens": 48617652.0,
      "step": 19180
    },
    {
      "entropy": 6.245084238052368,
      "epoch": 2.2140796306982113,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045178121873817395,
      "loss": 5.5989,
      "mean_token_accuracy": 0.21215347498655318,
      "num_tokens": 48629696.0,
      "step": 19185
    },
    {
      "entropy": 6.224291801452637,
      "epoch": 2.2146566647432198,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004517556890203265,
      "loss": 5.6761,
      "mean_token_accuracy": 0.20050013661384583,
      "num_tokens": 48641297.0,
      "step": 19190
    },
    {
      "entropy": 6.272027492523193,
      "epoch": 2.2152336987882286,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045173015335738925,
      "loss": 5.6856,
      "mean_token_accuracy": 0.20090647637844086,
      "num_tokens": 48654384.0,
      "step": 19195
    },
    {
      "entropy": 6.223822450637817,
      "epoch": 2.215810732833237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045170461175022116,
      "loss": 5.6019,
      "mean_token_accuracy": 0.20782951563596724,
      "num_tokens": 48666868.0,
      "step": 19200
    },
    {
      "entropy": 6.305508947372436,
      "epoch": 2.216387766878246,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004516790641996812,
      "loss": 5.7428,
      "mean_token_accuracy": 0.1938718244433403,
      "num_tokens": 48679780.0,
      "step": 19205
    },
    {
      "entropy": 6.280866909027099,
      "epoch": 2.2169648009232543,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004516535107066286,
      "loss": 5.6844,
      "mean_token_accuracy": 0.19734510630369187,
      "num_tokens": 48692438.0,
      "step": 19210
    },
    {
      "entropy": 6.257282304763794,
      "epoch": 2.217541834968263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045162795127192296,
      "loss": 5.7049,
      "mean_token_accuracy": 0.20301399379968643,
      "num_tokens": 48705627.0,
      "step": 19215
    },
    {
      "entropy": 6.287993717193603,
      "epoch": 2.2181188690132716,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000451602385896424,
      "loss": 5.6589,
      "mean_token_accuracy": 0.19510169327259064,
      "num_tokens": 48717481.0,
      "step": 19220
    },
    {
      "entropy": 6.228791570663452,
      "epoch": 2.2186959030582805,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045157681458099145,
      "loss": 5.6949,
      "mean_token_accuracy": 0.19680411070585252,
      "num_tokens": 48729821.0,
      "step": 19225
    },
    {
      "entropy": 6.3001447200775145,
      "epoch": 2.2192729371032893,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004515512373264854,
      "loss": 5.6651,
      "mean_token_accuracy": 0.2024334952235222,
      "num_tokens": 48742632.0,
      "step": 19230
    },
    {
      "entropy": 6.230295610427857,
      "epoch": 2.2198499711482977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045152565413376623,
      "loss": 5.6323,
      "mean_token_accuracy": 0.2053336888551712,
      "num_tokens": 48754712.0,
      "step": 19235
    },
    {
      "entropy": 6.202464532852173,
      "epoch": 2.2204270051933066,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004515000650036944,
      "loss": 5.6668,
      "mean_token_accuracy": 0.20371556729078294,
      "num_tokens": 48767170.0,
      "step": 19240
    },
    {
      "entropy": 6.327952337265015,
      "epoch": 2.221004039238315,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004514744699371305,
      "loss": 5.7652,
      "mean_token_accuracy": 0.19156857430934907,
      "num_tokens": 48780392.0,
      "step": 19245
    },
    {
      "entropy": 6.357401990890503,
      "epoch": 2.221581073283324,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045144886893493547,
      "loss": 5.6769,
      "mean_token_accuracy": 0.20644160807132722,
      "num_tokens": 48792100.0,
      "step": 19250
    },
    {
      "entropy": 6.226708841323853,
      "epoch": 2.2221581073283323,
      "grad_norm": 0.859375,
      "learning_rate": 0.00045142326199797047,
      "loss": 5.7148,
      "mean_token_accuracy": 0.20027264654636384,
      "num_tokens": 48805192.0,
      "step": 19255
    },
    {
      "entropy": 6.233144998550415,
      "epoch": 2.222735141373341,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045139764912709656,
      "loss": 5.5933,
      "mean_token_accuracy": 0.20631292909383775,
      "num_tokens": 48819735.0,
      "step": 19260
    },
    {
      "entropy": 6.276242685317993,
      "epoch": 2.2233121754183496,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004513720303231754,
      "loss": 5.6662,
      "mean_token_accuracy": 0.20086461901664734,
      "num_tokens": 48833785.0,
      "step": 19265
    },
    {
      "entropy": 6.269479131698608,
      "epoch": 2.2238892094633584,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045134640558706864,
      "loss": 5.6589,
      "mean_token_accuracy": 0.2045081302523613,
      "num_tokens": 48846491.0,
      "step": 19270
    },
    {
      "entropy": 6.250634765625,
      "epoch": 2.224466243508367,
      "grad_norm": 0.96875,
      "learning_rate": 0.000451320774919638,
      "loss": 5.6105,
      "mean_token_accuracy": 0.19953082650899887,
      "num_tokens": 48858269.0,
      "step": 19275
    },
    {
      "entropy": 6.310326385498047,
      "epoch": 2.2250432775533757,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045129513832174587,
      "loss": 5.7779,
      "mean_token_accuracy": 0.18675852566957474,
      "num_tokens": 48870051.0,
      "step": 19280
    },
    {
      "entropy": 6.271349382400513,
      "epoch": 2.225620311598384,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045126949579425414,
      "loss": 5.6886,
      "mean_token_accuracy": 0.20334390103816985,
      "num_tokens": 48882933.0,
      "step": 19285
    },
    {
      "entropy": 6.261395645141602,
      "epoch": 2.226197345643393,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045124384733802563,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19847685545682908,
      "num_tokens": 48895641.0,
      "step": 19290
    },
    {
      "entropy": 6.335473823547363,
      "epoch": 2.2267743796884014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004512181929539228,
      "loss": 5.7684,
      "mean_token_accuracy": 0.19306793361902236,
      "num_tokens": 48907509.0,
      "step": 19295
    },
    {
      "entropy": 6.295061254501343,
      "epoch": 2.2273514137334103,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045119253264280855,
      "loss": 5.6351,
      "mean_token_accuracy": 0.20542253851890563,
      "num_tokens": 48920270.0,
      "step": 19300
    },
    {
      "entropy": 6.228653335571289,
      "epoch": 2.227928447778419,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045116686640554607,
      "loss": 5.6242,
      "mean_token_accuracy": 0.20741915255784987,
      "num_tokens": 48932630.0,
      "step": 19305
    },
    {
      "entropy": 6.273880863189698,
      "epoch": 2.2285054818234276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045114119424299843,
      "loss": 5.6637,
      "mean_token_accuracy": 0.2061673879623413,
      "num_tokens": 48944759.0,
      "step": 19310
    },
    {
      "entropy": 6.149727296829224,
      "epoch": 2.2290825158684364,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004511155161560293,
      "loss": 5.5869,
      "mean_token_accuracy": 0.21210285127162934,
      "num_tokens": 48956835.0,
      "step": 19315
    },
    {
      "entropy": 6.196287775039673,
      "epoch": 2.229659549913445,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045108983214550225,
      "loss": 5.5799,
      "mean_token_accuracy": 0.21053734570741653,
      "num_tokens": 48970552.0,
      "step": 19320
    },
    {
      "entropy": 6.29644193649292,
      "epoch": 2.2302365839584537,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004510641422122811,
      "loss": 5.7469,
      "mean_token_accuracy": 0.19769887924194335,
      "num_tokens": 48981206.0,
      "step": 19325
    },
    {
      "entropy": 6.276036977767944,
      "epoch": 2.230813618003462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045103844635723005,
      "loss": 5.6596,
      "mean_token_accuracy": 0.19861132204532622,
      "num_tokens": 48992849.0,
      "step": 19330
    },
    {
      "entropy": 6.3406047344207765,
      "epoch": 2.231390652048471,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045101274458121335,
      "loss": 5.8327,
      "mean_token_accuracy": 0.19742682427167893,
      "num_tokens": 49006094.0,
      "step": 19335
    },
    {
      "entropy": 6.184029054641724,
      "epoch": 2.2319676860934794,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045098703688509537,
      "loss": 5.6755,
      "mean_token_accuracy": 0.19771206974983216,
      "num_tokens": 49019118.0,
      "step": 19340
    },
    {
      "entropy": 6.219599342346191,
      "epoch": 2.2325447201384883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004509613232697408,
      "loss": 5.5792,
      "mean_token_accuracy": 0.2099671855568886,
      "num_tokens": 49031217.0,
      "step": 19345
    },
    {
      "entropy": 6.290146541595459,
      "epoch": 2.2331217541834967,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004509356037360145,
      "loss": 5.6939,
      "mean_token_accuracy": 0.20073580741882324,
      "num_tokens": 49043251.0,
      "step": 19350
    },
    {
      "entropy": 6.301710510253907,
      "epoch": 2.2336987882285055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004509098782847816,
      "loss": 5.7,
      "mean_token_accuracy": 0.19606435894966126,
      "num_tokens": 49055227.0,
      "step": 19355
    },
    {
      "entropy": 6.311263179779052,
      "epoch": 2.234275822273514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004508841469169073,
      "loss": 5.7527,
      "mean_token_accuracy": 0.19532299488782884,
      "num_tokens": 49067810.0,
      "step": 19360
    },
    {
      "entropy": 6.238333511352539,
      "epoch": 2.234852856318523,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045085840963325716,
      "loss": 5.6178,
      "mean_token_accuracy": 0.21025243103504182,
      "num_tokens": 49079898.0,
      "step": 19365
    },
    {
      "entropy": 6.329520845413208,
      "epoch": 2.2354298903635312,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004508326664346967,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1979602813720703,
      "num_tokens": 49092961.0,
      "step": 19370
    },
    {
      "entropy": 6.2410078048706055,
      "epoch": 2.23600692440854,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00045080691732209176,
      "loss": 5.6713,
      "mean_token_accuracy": 0.20437232702970504,
      "num_tokens": 49106670.0,
      "step": 19375
    },
    {
      "entropy": 6.121372985839844,
      "epoch": 2.236583958453549,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045078116229630864,
      "loss": 5.5283,
      "mean_token_accuracy": 0.2150515392422676,
      "num_tokens": 49119947.0,
      "step": 19380
    },
    {
      "entropy": 6.262368440628052,
      "epoch": 2.2371609924985574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045075540135821343,
      "loss": 5.7051,
      "mean_token_accuracy": 0.1984892264008522,
      "num_tokens": 49132486.0,
      "step": 19385
    },
    {
      "entropy": 6.293896579742432,
      "epoch": 2.2377380265435662,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004507296345086725,
      "loss": 5.6914,
      "mean_token_accuracy": 0.1952108010649681,
      "num_tokens": 49145121.0,
      "step": 19390
    },
    {
      "entropy": 6.278210210800171,
      "epoch": 2.2383150605885747,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004507038617485526,
      "loss": 5.724,
      "mean_token_accuracy": 0.19411944448947907,
      "num_tokens": 49157247.0,
      "step": 19395
    },
    {
      "entropy": 6.338027620315552,
      "epoch": 2.2388920946335835,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045067808307872064,
      "loss": 5.7718,
      "mean_token_accuracy": 0.19023239314556123,
      "num_tokens": 49170886.0,
      "step": 19400
    },
    {
      "entropy": 6.253342914581299,
      "epoch": 2.239469128678592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045065229850004365,
      "loss": 5.7105,
      "mean_token_accuracy": 0.19242435544729233,
      "num_tokens": 49183087.0,
      "step": 19405
    },
    {
      "entropy": 6.306363534927368,
      "epoch": 2.240046162723601,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004506265080133888,
      "loss": 5.6813,
      "mean_token_accuracy": 0.1980726808309555,
      "num_tokens": 49194703.0,
      "step": 19410
    },
    {
      "entropy": 6.2686042308807375,
      "epoch": 2.240623196768609,
      "grad_norm": 1.0,
      "learning_rate": 0.00045060071161962364,
      "loss": 5.7444,
      "mean_token_accuracy": 0.20032382011413574,
      "num_tokens": 49208459.0,
      "step": 19415
    },
    {
      "entropy": 6.189984512329102,
      "epoch": 2.241200230813618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004505749093196158,
      "loss": 5.5755,
      "mean_token_accuracy": 0.20888419896364213,
      "num_tokens": 49221043.0,
      "step": 19420
    },
    {
      "entropy": 6.353961038589477,
      "epoch": 2.2417772648586265,
      "grad_norm": 0.9375,
      "learning_rate": 0.000450549101114233,
      "loss": 5.7168,
      "mean_token_accuracy": 0.19870360046625138,
      "num_tokens": 49233738.0,
      "step": 19425
    },
    {
      "entropy": 6.2713377475738525,
      "epoch": 2.2423542989036354,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004505232870043434,
      "loss": 5.7081,
      "mean_token_accuracy": 0.19887772500514983,
      "num_tokens": 49246244.0,
      "step": 19430
    },
    {
      "entropy": 6.272959470748901,
      "epoch": 2.2429313329486438,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004504974669908152,
      "loss": 5.7495,
      "mean_token_accuracy": 0.1927038237452507,
      "num_tokens": 49258126.0,
      "step": 19435
    },
    {
      "entropy": 6.265785169601441,
      "epoch": 2.2435083669936526,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045047164107451696,
      "loss": 5.7207,
      "mean_token_accuracy": 0.20121547281742097,
      "num_tokens": 49270303.0,
      "step": 19440
    },
    {
      "entropy": 6.151367950439453,
      "epoch": 2.244085401038661,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004504458092563172,
      "loss": 5.4968,
      "mean_token_accuracy": 0.21375908106565475,
      "num_tokens": 49283295.0,
      "step": 19445
    },
    {
      "entropy": 6.281554460525513,
      "epoch": 2.24466243508367,
      "grad_norm": 1.0,
      "learning_rate": 0.00045041997153708475,
      "loss": 5.6964,
      "mean_token_accuracy": 0.19873198866844177,
      "num_tokens": 49296003.0,
      "step": 19450
    },
    {
      "entropy": 6.293779563903809,
      "epoch": 2.2452394691286788,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045039412791768877,
      "loss": 5.7586,
      "mean_token_accuracy": 0.20105987340211867,
      "num_tokens": 49307982.0,
      "step": 19455
    },
    {
      "entropy": 6.268418407440185,
      "epoch": 2.245816503173687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045036827839899833,
      "loss": 5.7034,
      "mean_token_accuracy": 0.20017822533845903,
      "num_tokens": 49319220.0,
      "step": 19460
    },
    {
      "entropy": 6.1963825702667235,
      "epoch": 2.246393537218696,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045034242298188303,
      "loss": 5.6481,
      "mean_token_accuracy": 0.20568958967924117,
      "num_tokens": 49331683.0,
      "step": 19465
    },
    {
      "entropy": 6.2171236038208,
      "epoch": 2.2469705712637045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004503165616672124,
      "loss": 5.5769,
      "mean_token_accuracy": 0.21518644094467163,
      "num_tokens": 49344815.0,
      "step": 19470
    },
    {
      "entropy": 6.334026908874511,
      "epoch": 2.2475476053087133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004502906944558563,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19766660332679747,
      "num_tokens": 49357399.0,
      "step": 19475
    },
    {
      "entropy": 6.291712617874145,
      "epoch": 2.2481246393537218,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004502648213486848,
      "loss": 5.7392,
      "mean_token_accuracy": 0.1950973317027092,
      "num_tokens": 49369641.0,
      "step": 19480
    },
    {
      "entropy": 6.244687843322754,
      "epoch": 2.2487016733987306,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045023894234656807,
      "loss": 5.6788,
      "mean_token_accuracy": 0.2035685196518898,
      "num_tokens": 49381043.0,
      "step": 19485
    },
    {
      "entropy": 6.304348802566528,
      "epoch": 2.249278707443739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004502130574503766,
      "loss": 5.7175,
      "mean_token_accuracy": 0.20120338052511216,
      "num_tokens": 49392862.0,
      "step": 19490
    },
    {
      "entropy": 6.348339700698853,
      "epoch": 2.249855741488748,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004501871666609809,
      "loss": 5.7619,
      "mean_token_accuracy": 0.19692918062210082,
      "num_tokens": 49405566.0,
      "step": 19495
    },
    {
      "entropy": 6.2860541343688965,
      "epoch": 2.2504327755337563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004501612699792519,
      "loss": 5.6944,
      "mean_token_accuracy": 0.19930025041103364,
      "num_tokens": 49418048.0,
      "step": 19500
    },
    {
      "entropy": 6.206205940246582,
      "epoch": 2.251009809578765,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004501353674060606,
      "loss": 5.6799,
      "mean_token_accuracy": 0.2036149799823761,
      "num_tokens": 49430129.0,
      "step": 19505
    },
    {
      "entropy": 6.293288707733154,
      "epoch": 2.251586843623774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004501094589422782,
      "loss": 5.7495,
      "mean_token_accuracy": 0.19339427947998047,
      "num_tokens": 49443047.0,
      "step": 19510
    },
    {
      "entropy": 6.321244239807129,
      "epoch": 2.2521638776687825,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045008354458877614,
      "loss": 5.7239,
      "mean_token_accuracy": 0.1999143436551094,
      "num_tokens": 49456061.0,
      "step": 19515
    },
    {
      "entropy": 6.147455263137817,
      "epoch": 2.252740911713791,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000450057624346426,
      "loss": 5.4799,
      "mean_token_accuracy": 0.21652720719575883,
      "num_tokens": 49470044.0,
      "step": 19520
    },
    {
      "entropy": 6.240294694900513,
      "epoch": 2.2533179457587997,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045003169821609967,
      "loss": 5.7055,
      "mean_token_accuracy": 0.19474950730800628,
      "num_tokens": 49482232.0,
      "step": 19525
    },
    {
      "entropy": 6.2413819313049315,
      "epoch": 2.2538949798038086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045000576619866914,
      "loss": 5.6198,
      "mean_token_accuracy": 0.20373505204916,
      "num_tokens": 49493639.0,
      "step": 19530
    },
    {
      "entropy": 6.344704818725586,
      "epoch": 2.254472013848817,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044997982829500657,
      "loss": 5.7817,
      "mean_token_accuracy": 0.1936420902609825,
      "num_tokens": 49506233.0,
      "step": 19535
    },
    {
      "entropy": 6.242608976364136,
      "epoch": 2.255049047893826,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044995388450598436,
      "loss": 5.7051,
      "mean_token_accuracy": 0.19923149347305297,
      "num_tokens": 49518479.0,
      "step": 19540
    },
    {
      "entropy": 6.284429264068604,
      "epoch": 2.2556260819388343,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004499279348324751,
      "loss": 5.7145,
      "mean_token_accuracy": 0.19815237522125245,
      "num_tokens": 49531293.0,
      "step": 19545
    },
    {
      "entropy": 6.22764024734497,
      "epoch": 2.256203115983843,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044990197927535173,
      "loss": 5.6071,
      "mean_token_accuracy": 0.20217270106077195,
      "num_tokens": 49543492.0,
      "step": 19550
    },
    {
      "entropy": 6.291266059875488,
      "epoch": 2.2567801500288516,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044987601783548703,
      "loss": 5.7307,
      "mean_token_accuracy": 0.19845923781394958,
      "num_tokens": 49555460.0,
      "step": 19555
    },
    {
      "entropy": 6.148696851730347,
      "epoch": 2.2573571840738604,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004498500505137545,
      "loss": 5.5574,
      "mean_token_accuracy": 0.2042447879910469,
      "num_tokens": 49568077.0,
      "step": 19560
    },
    {
      "entropy": 6.2508501529693605,
      "epoch": 2.257934218118869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004498240773110273,
      "loss": 5.6799,
      "mean_token_accuracy": 0.19796499013900756,
      "num_tokens": 49580201.0,
      "step": 19565
    },
    {
      "entropy": 6.22829647064209,
      "epoch": 2.2585112521638777,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004497980982281791,
      "loss": 5.6522,
      "mean_token_accuracy": 0.1994529828429222,
      "num_tokens": 49592469.0,
      "step": 19570
    },
    {
      "entropy": 6.281613969802857,
      "epoch": 2.259088286208886,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004497721132660837,
      "loss": 5.7306,
      "mean_token_accuracy": 0.1930721014738083,
      "num_tokens": 49604961.0,
      "step": 19575
    },
    {
      "entropy": 6.291258430480957,
      "epoch": 2.259665320253895,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044974612242561516,
      "loss": 5.7096,
      "mean_token_accuracy": 0.19761281013488768,
      "num_tokens": 49617435.0,
      "step": 19580
    },
    {
      "entropy": 6.309053659439087,
      "epoch": 2.260242354298904,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004497201257076475,
      "loss": 5.7003,
      "mean_token_accuracy": 0.19988716542720794,
      "num_tokens": 49629799.0,
      "step": 19585
    },
    {
      "entropy": 6.165639066696167,
      "epoch": 2.2608193883439123,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004496941231130552,
      "loss": 5.5452,
      "mean_token_accuracy": 0.20381347239017486,
      "num_tokens": 49642937.0,
      "step": 19590
    },
    {
      "entropy": 6.264139080047608,
      "epoch": 2.261396422388921,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004496681146427129,
      "loss": 5.6587,
      "mean_token_accuracy": 0.2007022649049759,
      "num_tokens": 49655731.0,
      "step": 19595
    },
    {
      "entropy": 6.33866138458252,
      "epoch": 2.2619734564339296,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044964210029749523,
      "loss": 5.6796,
      "mean_token_accuracy": 0.19902746975421906,
      "num_tokens": 49668515.0,
      "step": 19600
    },
    {
      "entropy": 6.3006068706512455,
      "epoch": 2.2625504904789384,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044961608007827736,
      "loss": 5.7223,
      "mean_token_accuracy": 0.19971500486135482,
      "num_tokens": 49681246.0,
      "step": 19605
    },
    {
      "entropy": 6.258845043182373,
      "epoch": 2.263127524523947,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004495900539859344,
      "loss": 5.7072,
      "mean_token_accuracy": 0.19777692407369613,
      "num_tokens": 49694234.0,
      "step": 19610
    },
    {
      "entropy": 6.229405355453491,
      "epoch": 2.2637045585689557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044956402202134157,
      "loss": 5.5755,
      "mean_token_accuracy": 0.21066538989543915,
      "num_tokens": 49707511.0,
      "step": 19615
    },
    {
      "entropy": 6.229258632659912,
      "epoch": 2.264281592613964,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044953798418537465,
      "loss": 5.6788,
      "mean_token_accuracy": 0.20957430303096772,
      "num_tokens": 49720348.0,
      "step": 19620
    },
    {
      "entropy": 6.162171697616577,
      "epoch": 2.264858626658973,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004495119404789093,
      "loss": 5.5329,
      "mean_token_accuracy": 0.2054566487669945,
      "num_tokens": 49733777.0,
      "step": 19625
    },
    {
      "entropy": 6.266354131698608,
      "epoch": 2.2654356607039814,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004494858909028216,
      "loss": 5.6268,
      "mean_token_accuracy": 0.2031772255897522,
      "num_tokens": 49746962.0,
      "step": 19630
    },
    {
      "entropy": 6.334909582138062,
      "epoch": 2.2660126947489903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004494598354579875,
      "loss": 5.7383,
      "mean_token_accuracy": 0.19375489205121993,
      "num_tokens": 49760211.0,
      "step": 19635
    },
    {
      "entropy": 6.1849085807800295,
      "epoch": 2.2665897287939987,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004494337741452836,
      "loss": 5.6095,
      "mean_token_accuracy": 0.21134635508060456,
      "num_tokens": 49773215.0,
      "step": 19640
    },
    {
      "entropy": 6.261235141754151,
      "epoch": 2.2671667628390075,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004494077069655863,
      "loss": 5.6471,
      "mean_token_accuracy": 0.20450802892446518,
      "num_tokens": 49785521.0,
      "step": 19645
    },
    {
      "entropy": 6.223216867446899,
      "epoch": 2.267743796884016,
      "grad_norm": 1.0,
      "learning_rate": 0.0004493816339197724,
      "loss": 5.6542,
      "mean_token_accuracy": 0.20172957181930543,
      "num_tokens": 49796449.0,
      "step": 19650
    },
    {
      "entropy": 6.256512594223023,
      "epoch": 2.268320830929025,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044935555500871897,
      "loss": 5.6981,
      "mean_token_accuracy": 0.20591560751199722,
      "num_tokens": 49810222.0,
      "step": 19655
    },
    {
      "entropy": 6.272163391113281,
      "epoch": 2.2688978649740337,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000449329470233303,
      "loss": 5.6737,
      "mean_token_accuracy": 0.20172365754842758,
      "num_tokens": 49822799.0,
      "step": 19660
    },
    {
      "entropy": 6.253868532180786,
      "epoch": 2.269474899019042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044930337959440183,
      "loss": 5.6592,
      "mean_token_accuracy": 0.1968416303396225,
      "num_tokens": 49835782.0,
      "step": 19665
    },
    {
      "entropy": 6.306293582916259,
      "epoch": 2.270051933064051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004492772830928931,
      "loss": 5.6753,
      "mean_token_accuracy": 0.20224833190441133,
      "num_tokens": 49849204.0,
      "step": 19670
    },
    {
      "entropy": 6.298428726196289,
      "epoch": 2.2706289671090594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044925118072965456,
      "loss": 5.7578,
      "mean_token_accuracy": 0.19448983371257783,
      "num_tokens": 49860769.0,
      "step": 19675
    },
    {
      "entropy": 6.26639838218689,
      "epoch": 2.2712060011540682,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000449225072505564,
      "loss": 5.675,
      "mean_token_accuracy": 0.2055061027407646,
      "num_tokens": 49873923.0,
      "step": 19680
    },
    {
      "entropy": 6.273349142074585,
      "epoch": 2.2717830351990767,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044919895842149976,
      "loss": 5.6467,
      "mean_token_accuracy": 0.20548494160175323,
      "num_tokens": 49886968.0,
      "step": 19685
    },
    {
      "entropy": 6.182073926925659,
      "epoch": 2.2723600692440855,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044917283847834005,
      "loss": 5.6087,
      "mean_token_accuracy": 0.20875933170318603,
      "num_tokens": 49899096.0,
      "step": 19690
    },
    {
      "entropy": 6.215857076644897,
      "epoch": 2.272937103289094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004491467126769635,
      "loss": 5.6242,
      "mean_token_accuracy": 0.19774646759033204,
      "num_tokens": 49911887.0,
      "step": 19695
    },
    {
      "entropy": 6.243524026870728,
      "epoch": 2.273514137334103,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044912058101824866,
      "loss": 5.616,
      "mean_token_accuracy": 0.20824216306209564,
      "num_tokens": 49925188.0,
      "step": 19700
    },
    {
      "entropy": 6.197985458374023,
      "epoch": 2.274091171379111,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044909444350307463,
      "loss": 5.6368,
      "mean_token_accuracy": 0.20444974899291993,
      "num_tokens": 49938568.0,
      "step": 19705
    },
    {
      "entropy": 6.199601888656616,
      "epoch": 2.27466820542412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004490683001323205,
      "loss": 5.5664,
      "mean_token_accuracy": 0.21142929047346115,
      "num_tokens": 49950630.0,
      "step": 19710
    },
    {
      "entropy": 6.239274263381958,
      "epoch": 2.2752452394691285,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044904215090686554,
      "loss": 5.6207,
      "mean_token_accuracy": 0.21090169250965118,
      "num_tokens": 49962242.0,
      "step": 19715
    },
    {
      "entropy": 6.279738235473633,
      "epoch": 2.2758222735141374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044901599582758926,
      "loss": 5.7232,
      "mean_token_accuracy": 0.19315190464258195,
      "num_tokens": 49974604.0,
      "step": 19720
    },
    {
      "entropy": 6.263618850708008,
      "epoch": 2.2763993075591458,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044898983489537143,
      "loss": 5.6535,
      "mean_token_accuracy": 0.20598474591970445,
      "num_tokens": 49986326.0,
      "step": 19725
    },
    {
      "entropy": 6.25758547782898,
      "epoch": 2.2769763416041546,
      "grad_norm": 1.03125,
      "learning_rate": 0.000448963668111092,
      "loss": 5.6286,
      "mean_token_accuracy": 0.21237566769123079,
      "num_tokens": 49998752.0,
      "step": 19730
    },
    {
      "entropy": 6.2858623504638675,
      "epoch": 2.2775533756491635,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044893749547563085,
      "loss": 5.7179,
      "mean_token_accuracy": 0.1985759273171425,
      "num_tokens": 50011839.0,
      "step": 19735
    },
    {
      "entropy": 6.2531860828399655,
      "epoch": 2.278130409694172,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044891131698986846,
      "loss": 5.6083,
      "mean_token_accuracy": 0.20769107043743135,
      "num_tokens": 50024554.0,
      "step": 19740
    },
    {
      "entropy": 6.287480163574219,
      "epoch": 2.2787074437391808,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488851326546854,
      "loss": 5.7384,
      "mean_token_accuracy": 0.19784743040800096,
      "num_tokens": 50038868.0,
      "step": 19745
    },
    {
      "entropy": 6.300860214233398,
      "epoch": 2.279284477784189,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004488589424709622,
      "loss": 5.6848,
      "mean_token_accuracy": 0.20356245785951615,
      "num_tokens": 50050556.0,
      "step": 19750
    },
    {
      "entropy": 6.322600412368774,
      "epoch": 2.279861511829198,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004488327464395799,
      "loss": 5.7133,
      "mean_token_accuracy": 0.1967867985367775,
      "num_tokens": 50063169.0,
      "step": 19755
    },
    {
      "entropy": 6.267973327636719,
      "epoch": 2.2804385458742065,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004488065445614195,
      "loss": 5.7346,
      "mean_token_accuracy": 0.20178399682044984,
      "num_tokens": 50076563.0,
      "step": 19760
    },
    {
      "entropy": 6.241406488418579,
      "epoch": 2.2810155799192153,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004487803368373623,
      "loss": 5.6867,
      "mean_token_accuracy": 0.20351874083280563,
      "num_tokens": 50089584.0,
      "step": 19765
    },
    {
      "entropy": 6.212295293807983,
      "epoch": 2.2815926139642237,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004487541232682898,
      "loss": 5.6177,
      "mean_token_accuracy": 0.20406360924243927,
      "num_tokens": 50101566.0,
      "step": 19770
    },
    {
      "entropy": 6.289868545532227,
      "epoch": 2.2821696480092326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004487279038550836,
      "loss": 5.6975,
      "mean_token_accuracy": 0.20299792289733887,
      "num_tokens": 50114378.0,
      "step": 19775
    },
    {
      "entropy": 6.22979850769043,
      "epoch": 2.282746682054241,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487016785986258,
      "loss": 5.6501,
      "mean_token_accuracy": 0.2072186663746834,
      "num_tokens": 50126658.0,
      "step": 19780
    },
    {
      "entropy": 6.2693885326385494,
      "epoch": 2.28332371609925,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044867544749979817,
      "loss": 5.7034,
      "mean_token_accuracy": 0.20219238996505737,
      "num_tokens": 50139802.0,
      "step": 19785
    },
    {
      "entropy": 6.336936330795288,
      "epoch": 2.2839007501442588,
      "grad_norm": 1.0,
      "learning_rate": 0.00044864921055948317,
      "loss": 5.7811,
      "mean_token_accuracy": 0.18792867064476013,
      "num_tokens": 50152158.0,
      "step": 19790
    },
    {
      "entropy": 6.300091648101807,
      "epoch": 2.284477784189267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044862296777856326,
      "loss": 5.6825,
      "mean_token_accuracy": 0.20104401260614396,
      "num_tokens": 50165099.0,
      "step": 19795
    },
    {
      "entropy": 6.213589143753052,
      "epoch": 2.2850548182342756,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485967191579211,
      "loss": 5.6505,
      "mean_token_accuracy": 0.20115123093128204,
      "num_tokens": 50177073.0,
      "step": 19800
    },
    {
      "entropy": 6.29067587852478,
      "epoch": 2.2856318522792844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004485704646984394,
      "loss": 5.7091,
      "mean_token_accuracy": 0.20162361562252046,
      "num_tokens": 50190458.0,
      "step": 19805
    },
    {
      "entropy": 6.247507810592651,
      "epoch": 2.2862088863242933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004485442044010015,
      "loss": 5.6373,
      "mean_token_accuracy": 0.20112094432115554,
      "num_tokens": 50201726.0,
      "step": 19810
    },
    {
      "entropy": 6.226012134552002,
      "epoch": 2.2867859203693017,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004485179382664904,
      "loss": 5.6275,
      "mean_token_accuracy": 0.21108348816633224,
      "num_tokens": 50215550.0,
      "step": 19815
    },
    {
      "entropy": 6.242710971832276,
      "epoch": 2.2873629544143106,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004484916662957896,
      "loss": 5.6531,
      "mean_token_accuracy": 0.2013290137052536,
      "num_tokens": 50227982.0,
      "step": 19820
    },
    {
      "entropy": 6.287098169326782,
      "epoch": 2.287939988459319,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004484653884897829,
      "loss": 5.697,
      "mean_token_accuracy": 0.19716725051403045,
      "num_tokens": 50239804.0,
      "step": 19825
    },
    {
      "entropy": 6.296359920501709,
      "epoch": 2.288517022504328,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044843910484935394,
      "loss": 5.7051,
      "mean_token_accuracy": 0.20337263941764833,
      "num_tokens": 50252192.0,
      "step": 19830
    },
    {
      "entropy": 6.3207886695861815,
      "epoch": 2.2890940565493363,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004484128153753868,
      "loss": 5.7775,
      "mean_token_accuracy": 0.19393563866615296,
      "num_tokens": 50265194.0,
      "step": 19835
    },
    {
      "entropy": 6.248172855377197,
      "epoch": 2.289671090594345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044838652006876583,
      "loss": 5.6234,
      "mean_token_accuracy": 0.20863277316093445,
      "num_tokens": 50277704.0,
      "step": 19840
    },
    {
      "entropy": 6.16427059173584,
      "epoch": 2.2902481246393536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044836021893037537,
      "loss": 5.6121,
      "mean_token_accuracy": 0.2052631974220276,
      "num_tokens": 50290670.0,
      "step": 19845
    },
    {
      "entropy": 6.189370059967041,
      "epoch": 2.2908251586843624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004483339119611001,
      "loss": 5.6138,
      "mean_token_accuracy": 0.20742383003234863,
      "num_tokens": 50302764.0,
      "step": 19850
    },
    {
      "entropy": 6.247882652282715,
      "epoch": 2.291402192729371,
      "grad_norm": 1.0,
      "learning_rate": 0.00044830759916182476,
      "loss": 5.6724,
      "mean_token_accuracy": 0.20733534246683122,
      "num_tokens": 50316824.0,
      "step": 19855
    },
    {
      "entropy": 6.367726135253906,
      "epoch": 2.2919792267743797,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004482812805334344,
      "loss": 5.735,
      "mean_token_accuracy": 0.1979316368699074,
      "num_tokens": 50329259.0,
      "step": 19860
    },
    {
      "entropy": 6.2147620677947994,
      "epoch": 2.2925562608193886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004482549560768142,
      "loss": 5.6072,
      "mean_token_accuracy": 0.21409889310598373,
      "num_tokens": 50341165.0,
      "step": 19865
    },
    {
      "entropy": 6.1384663581848145,
      "epoch": 2.293133294864397,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004482286257928497,
      "loss": 5.6229,
      "mean_token_accuracy": 0.20753696858882903,
      "num_tokens": 50354655.0,
      "step": 19870
    },
    {
      "entropy": 6.2925012588500975,
      "epoch": 2.293710328909406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004482022896824263,
      "loss": 5.7101,
      "mean_token_accuracy": 0.19972767382860185,
      "num_tokens": 50366917.0,
      "step": 19875
    },
    {
      "entropy": 6.283646869659424,
      "epoch": 2.2942873629544143,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044817594774643004,
      "loss": 5.7299,
      "mean_token_accuracy": 0.20198842734098435,
      "num_tokens": 50380415.0,
      "step": 19880
    },
    {
      "entropy": 6.252005958557129,
      "epoch": 2.294864396999423,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044814959998574675,
      "loss": 5.6455,
      "mean_token_accuracy": 0.20236456990242005,
      "num_tokens": 50393051.0,
      "step": 19885
    },
    {
      "entropy": 6.267017650604248,
      "epoch": 2.2954414310444315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00044812324640126265,
      "loss": 5.7291,
      "mean_token_accuracy": 0.19582874774932862,
      "num_tokens": 50407517.0,
      "step": 19890
    },
    {
      "entropy": 6.304991579055786,
      "epoch": 2.2960184650894404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004480968869938642,
      "loss": 5.768,
      "mean_token_accuracy": 0.19451882392168046,
      "num_tokens": 50419656.0,
      "step": 19895
    },
    {
      "entropy": 6.23197455406189,
      "epoch": 2.296595499134449,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044807052176443793,
      "loss": 5.5888,
      "mean_token_accuracy": 0.20050704181194307,
      "num_tokens": 50433798.0,
      "step": 19900
    },
    {
      "entropy": 6.326214170455932,
      "epoch": 2.2971725331794577,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044804415071387067,
      "loss": 5.7539,
      "mean_token_accuracy": 0.19597139209508896,
      "num_tokens": 50445459.0,
      "step": 19905
    },
    {
      "entropy": 6.253977108001709,
      "epoch": 2.297749567224466,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004480177738430492,
      "loss": 5.6304,
      "mean_token_accuracy": 0.2138488247990608,
      "num_tokens": 50458649.0,
      "step": 19910
    },
    {
      "entropy": 6.094612073898316,
      "epoch": 2.298326601269475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044799139115286104,
      "loss": 5.5016,
      "mean_token_accuracy": 0.22122004479169846,
      "num_tokens": 50471878.0,
      "step": 19915
    },
    {
      "entropy": 6.3028308868408205,
      "epoch": 2.2989036353144834,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004479650026441933,
      "loss": 5.7464,
      "mean_token_accuracy": 0.19905219376087188,
      "num_tokens": 50484033.0,
      "step": 19920
    },
    {
      "entropy": 6.292723274230957,
      "epoch": 2.2994806693594922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044793860831793356,
      "loss": 5.6897,
      "mean_token_accuracy": 0.2036846920847893,
      "num_tokens": 50496072.0,
      "step": 19925
    },
    {
      "entropy": 6.292513847351074,
      "epoch": 2.3000577034045007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044791220817496963,
      "loss": 5.6397,
      "mean_token_accuracy": 0.211128394305706,
      "num_tokens": 50508572.0,
      "step": 19930
    },
    {
      "entropy": 6.258067464828491,
      "epoch": 2.3006347374495095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004478858022161895,
      "loss": 5.7699,
      "mean_token_accuracy": 0.19765519201755524,
      "num_tokens": 50521946.0,
      "step": 19935
    },
    {
      "entropy": 6.286145114898682,
      "epoch": 2.3012117714945184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478593904424813,
      "loss": 5.6429,
      "mean_token_accuracy": 0.2058190479874611,
      "num_tokens": 50534517.0,
      "step": 19940
    },
    {
      "entropy": 6.258847951889038,
      "epoch": 2.301788805539527,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004478329728547334,
      "loss": 5.6853,
      "mean_token_accuracy": 0.20199194103479384,
      "num_tokens": 50546479.0,
      "step": 19945
    },
    {
      "entropy": 6.306153774261475,
      "epoch": 2.3023658395845357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044780654945383424,
      "loss": 5.7625,
      "mean_token_accuracy": 0.1984587013721466,
      "num_tokens": 50559579.0,
      "step": 19950
    },
    {
      "entropy": 6.2746538639068605,
      "epoch": 2.302942873629544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044778012024067267,
      "loss": 5.6039,
      "mean_token_accuracy": 0.20293633341789247,
      "num_tokens": 50572810.0,
      "step": 19955
    },
    {
      "entropy": 6.2745934009552,
      "epoch": 2.303519907674553,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004477536852161376,
      "loss": 5.6872,
      "mean_token_accuracy": 0.19994460344314574,
      "num_tokens": 50585514.0,
      "step": 19960
    },
    {
      "entropy": 6.228223371505737,
      "epoch": 2.3040969417195614,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004477272443811181,
      "loss": 5.7051,
      "mean_token_accuracy": 0.2045487269759178,
      "num_tokens": 50597897.0,
      "step": 19965
    },
    {
      "entropy": 6.223145341873169,
      "epoch": 2.3046739757645702,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004477007977365035,
      "loss": 5.7032,
      "mean_token_accuracy": 0.20153484046459197,
      "num_tokens": 50610251.0,
      "step": 19970
    },
    {
      "entropy": 6.279410886764526,
      "epoch": 2.3052510098095786,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004476743452831834,
      "loss": 5.6709,
      "mean_token_accuracy": 0.19960313737392427,
      "num_tokens": 50622212.0,
      "step": 19975
    },
    {
      "entropy": 6.2553199291229244,
      "epoch": 2.3058280438545875,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044764788702204747,
      "loss": 5.6603,
      "mean_token_accuracy": 0.19849528670310973,
      "num_tokens": 50636648.0,
      "step": 19980
    },
    {
      "entropy": 6.333353281021118,
      "epoch": 2.306405077899596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004476214229539856,
      "loss": 5.7256,
      "mean_token_accuracy": 0.20371800065040588,
      "num_tokens": 50648033.0,
      "step": 19985
    },
    {
      "entropy": 6.275083923339844,
      "epoch": 2.306982111944605,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004475949530798879,
      "loss": 5.7493,
      "mean_token_accuracy": 0.19488436132669448,
      "num_tokens": 50660547.0,
      "step": 19990
    },
    {
      "entropy": 6.2665363311767575,
      "epoch": 2.307559145989613,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044756847740064475,
      "loss": 5.6434,
      "mean_token_accuracy": 0.20826811641454696,
      "num_tokens": 50672194.0,
      "step": 19995
    },
    {
      "entropy": 6.215688705444336,
      "epoch": 2.308136180034622,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004475419959171465,
      "loss": 5.6734,
      "mean_token_accuracy": 0.2031127005815506,
      "num_tokens": 50685896.0,
      "step": 20000
    },
    {
      "entropy": 6.182767391204834,
      "epoch": 2.3087132140796305,
      "grad_norm": 1.046875,
      "learning_rate": 0.000447515508630284,
      "loss": 5.6239,
      "mean_token_accuracy": 0.20299613624811172,
      "num_tokens": 50699255.0,
      "step": 20005
    },
    {
      "entropy": 6.237594938278198,
      "epoch": 2.3092902481246393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044748901554094806,
      "loss": 5.6161,
      "mean_token_accuracy": 0.20817889124155045,
      "num_tokens": 50711098.0,
      "step": 20010
    },
    {
      "entropy": 6.308900451660156,
      "epoch": 2.309867282169648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004474625166500297,
      "loss": 5.8127,
      "mean_token_accuracy": 0.19276440292596816,
      "num_tokens": 50724821.0,
      "step": 20015
    },
    {
      "entropy": 6.335447454452515,
      "epoch": 2.3104443162146566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044743601195842026,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1979809135198593,
      "num_tokens": 50736572.0,
      "step": 20020
    },
    {
      "entropy": 6.072242641448975,
      "epoch": 2.3110213502596655,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044740950146701127,
      "loss": 5.4582,
      "mean_token_accuracy": 0.21951938569545745,
      "num_tokens": 50749743.0,
      "step": 20025
    },
    {
      "entropy": 6.307686471939087,
      "epoch": 2.311598384304674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004473829851766943,
      "loss": 5.658,
      "mean_token_accuracy": 0.20932891070842743,
      "num_tokens": 50761558.0,
      "step": 20030
    },
    {
      "entropy": 6.29748854637146,
      "epoch": 2.3121754183496828,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004473564630883612,
      "loss": 5.7355,
      "mean_token_accuracy": 0.19531415402889252,
      "num_tokens": 50773963.0,
      "step": 20035
    },
    {
      "entropy": 6.236698007583618,
      "epoch": 2.312752452394691,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004473299352029042,
      "loss": 5.6699,
      "mean_token_accuracy": 0.20165673196315764,
      "num_tokens": 50787160.0,
      "step": 20040
    },
    {
      "entropy": 6.2561298370361325,
      "epoch": 2.3133294864397,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004473034015212154,
      "loss": 5.5511,
      "mean_token_accuracy": 0.21679649949073793,
      "num_tokens": 50800014.0,
      "step": 20045
    },
    {
      "entropy": 6.24334454536438,
      "epoch": 2.3139065204847085,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004472768620441872,
      "loss": 5.6008,
      "mean_token_accuracy": 0.2024207055568695,
      "num_tokens": 50812149.0,
      "step": 20050
    },
    {
      "entropy": 6.2762298583984375,
      "epoch": 2.3144835545297173,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044725031677271234,
      "loss": 5.7212,
      "mean_token_accuracy": 0.2026994124054909,
      "num_tokens": 50825044.0,
      "step": 20055
    },
    {
      "entropy": 6.355571365356445,
      "epoch": 2.3150605885747257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004472237657076837,
      "loss": 5.776,
      "mean_token_accuracy": 0.19817397147417068,
      "num_tokens": 50838813.0,
      "step": 20060
    },
    {
      "entropy": 6.267147254943848,
      "epoch": 2.3156376226197346,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004471972088499942,
      "loss": 5.5682,
      "mean_token_accuracy": 0.20367081761360167,
      "num_tokens": 50853029.0,
      "step": 20065
    },
    {
      "entropy": 6.255858373641968,
      "epoch": 2.3162146566647435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004471706462005371,
      "loss": 5.6604,
      "mean_token_accuracy": 0.20608419924974442,
      "num_tokens": 50866383.0,
      "step": 20070
    },
    {
      "entropy": 6.209093046188355,
      "epoch": 2.316791690709752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004471440777602059,
      "loss": 5.6679,
      "mean_token_accuracy": 0.20656839907169341,
      "num_tokens": 50880537.0,
      "step": 20075
    },
    {
      "entropy": 6.200661325454712,
      "epoch": 2.3173687247547603,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044711750352989407,
      "loss": 5.6458,
      "mean_token_accuracy": 0.20677527785301208,
      "num_tokens": 50892213.0,
      "step": 20080
    },
    {
      "entropy": 6.2801555633544925,
      "epoch": 2.317945758799769,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004470909235104956,
      "loss": 5.6822,
      "mean_token_accuracy": 0.20342328995466233,
      "num_tokens": 50904022.0,
      "step": 20085
    },
    {
      "entropy": 6.255757808685303,
      "epoch": 2.318522792844778,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470643377029043,
      "loss": 5.6522,
      "mean_token_accuracy": 0.19987345188856126,
      "num_tokens": 50916766.0,
      "step": 20090
    },
    {
      "entropy": 6.262523746490478,
      "epoch": 2.3190998268897864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470377461080145,
      "loss": 5.6339,
      "mean_token_accuracy": 0.20058869421482087,
      "num_tokens": 50928842.0,
      "step": 20095
    },
    {
      "entropy": 6.215834283828736,
      "epoch": 2.3196768609347953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470111487267206,
      "loss": 5.6589,
      "mean_token_accuracy": 0.20172934085130692,
      "num_tokens": 50941079.0,
      "step": 20100
    },
    {
      "entropy": 6.231136703491211,
      "epoch": 2.3202538949798037,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004469845455599171,
      "loss": 5.6413,
      "mean_token_accuracy": 0.20266380459070205,
      "num_tokens": 50953299.0,
      "step": 20105
    },
    {
      "entropy": 6.258836793899536,
      "epoch": 2.3208309290248126,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004469579366084989,
      "loss": 5.6269,
      "mean_token_accuracy": 0.20282865166664124,
      "num_tokens": 50965550.0,
      "step": 20110
    },
    {
      "entropy": 6.232774448394776,
      "epoch": 2.321407963069821,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044693132187336094,
      "loss": 5.6454,
      "mean_token_accuracy": 0.20486282110214232,
      "num_tokens": 50978454.0,
      "step": 20115
    },
    {
      "entropy": 6.271752834320068,
      "epoch": 2.32198499711483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004469047013553983,
      "loss": 5.6816,
      "mean_token_accuracy": 0.2025473728775978,
      "num_tokens": 50990473.0,
      "step": 20120
    },
    {
      "entropy": 6.234145259857177,
      "epoch": 2.3225620311598383,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044687807505550646,
      "loss": 5.7472,
      "mean_token_accuracy": 0.19835685491561889,
      "num_tokens": 51003914.0,
      "step": 20125
    },
    {
      "entropy": 6.334716796875,
      "epoch": 2.323139065204847,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044685144297458096,
      "loss": 5.7581,
      "mean_token_accuracy": 0.19428663253784179,
      "num_tokens": 51016874.0,
      "step": 20130
    },
    {
      "entropy": 6.260063934326172,
      "epoch": 2.3237160992498556,
      "grad_norm": 1.0,
      "learning_rate": 0.00044682480511351754,
      "loss": 5.6534,
      "mean_token_accuracy": 0.20682739466428757,
      "num_tokens": 51030568.0,
      "step": 20135
    },
    {
      "entropy": 6.315000581741333,
      "epoch": 2.3242931332948644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004467981614732121,
      "loss": 5.6874,
      "mean_token_accuracy": 0.19595109969377517,
      "num_tokens": 51043876.0,
      "step": 20140
    },
    {
      "entropy": 6.268725061416626,
      "epoch": 2.3248701673398733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044677151205456086,
      "loss": 5.7445,
      "mean_token_accuracy": 0.19285174161195756,
      "num_tokens": 51056693.0,
      "step": 20145
    },
    {
      "entropy": 6.3454419612884525,
      "epoch": 2.3254472013848817,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004467448568584601,
      "loss": 5.7272,
      "mean_token_accuracy": 0.20218962877988816,
      "num_tokens": 51068878.0,
      "step": 20150
    },
    {
      "entropy": 6.282096529006958,
      "epoch": 2.32602423542989,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004467181958858064,
      "loss": 5.7078,
      "mean_token_accuracy": 0.1994556576013565,
      "num_tokens": 51082234.0,
      "step": 20155
    },
    {
      "entropy": 6.242162370681763,
      "epoch": 2.326601269474899,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004466915291374965,
      "loss": 5.6892,
      "mean_token_accuracy": 0.20700761079788207,
      "num_tokens": 51095469.0,
      "step": 20160
    },
    {
      "entropy": 6.3454022884368895,
      "epoch": 2.327178303519908,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004466648566144273,
      "loss": 5.6762,
      "mean_token_accuracy": 0.20169818848371507,
      "num_tokens": 51107302.0,
      "step": 20165
    },
    {
      "entropy": 6.277090740203858,
      "epoch": 2.3277553375649163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000446638178317496,
      "loss": 5.7468,
      "mean_token_accuracy": 0.19192387759685517,
      "num_tokens": 51121222.0,
      "step": 20170
    },
    {
      "entropy": 6.314355182647705,
      "epoch": 2.328332371609925,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044661149424759974,
      "loss": 5.7318,
      "mean_token_accuracy": 0.20163864493370057,
      "num_tokens": 51134113.0,
      "step": 20175
    },
    {
      "entropy": 6.2640129089355465,
      "epoch": 2.3289094056549335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004465848044056361,
      "loss": 5.6579,
      "mean_token_accuracy": 0.20302291661500932,
      "num_tokens": 51145978.0,
      "step": 20180
    },
    {
      "entropy": 6.135003757476807,
      "epoch": 2.3294864396999424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004465581087925028,
      "loss": 5.566,
      "mean_token_accuracy": 0.20935563296079635,
      "num_tokens": 51158016.0,
      "step": 20185
    },
    {
      "entropy": 6.283479690551758,
      "epoch": 2.330063473744951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004465314074090978,
      "loss": 5.6693,
      "mean_token_accuracy": 0.2063070759177208,
      "num_tokens": 51170222.0,
      "step": 20190
    },
    {
      "entropy": 6.2931970119476315,
      "epoch": 2.3306405077899597,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044650470025631904,
      "loss": 5.6441,
      "mean_token_accuracy": 0.2030964732170105,
      "num_tokens": 51183560.0,
      "step": 20195
    },
    {
      "entropy": 6.227710723876953,
      "epoch": 2.331217541834968,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004464779873350649,
      "loss": 5.6565,
      "mean_token_accuracy": 0.2054636761546135,
      "num_tokens": 51195532.0,
      "step": 20200
    },
    {
      "entropy": 6.314197731018067,
      "epoch": 2.331794575879977,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004464512686462339,
      "loss": 5.7144,
      "mean_token_accuracy": 0.19894203096628188,
      "num_tokens": 51207078.0,
      "step": 20205
    },
    {
      "entropy": 6.2745904445648195,
      "epoch": 2.3323716099249854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044642454419072455,
      "loss": 5.7361,
      "mean_token_accuracy": 0.18972522765398026,
      "num_tokens": 51220642.0,
      "step": 20210
    },
    {
      "entropy": 6.279746675491333,
      "epoch": 2.3329486439699942,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004463978139694358,
      "loss": 5.7041,
      "mean_token_accuracy": 0.20027433782815934,
      "num_tokens": 51234127.0,
      "step": 20215
    },
    {
      "entropy": 6.264712810516357,
      "epoch": 2.333525678015003,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044637107798326675,
      "loss": 5.6664,
      "mean_token_accuracy": 0.20233578532934188,
      "num_tokens": 51246645.0,
      "step": 20220
    },
    {
      "entropy": 6.206380081176758,
      "epoch": 2.3341027120600115,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004463443362331166,
      "loss": 5.6032,
      "mean_token_accuracy": 0.21379945874214173,
      "num_tokens": 51259392.0,
      "step": 20225
    },
    {
      "entropy": 6.231292629241944,
      "epoch": 2.3346797461050204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044631758871988486,
      "loss": 5.6473,
      "mean_token_accuracy": 0.20762919187545775,
      "num_tokens": 51273073.0,
      "step": 20230
    },
    {
      "entropy": 6.290574645996093,
      "epoch": 2.335256780150029,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004462908354444711,
      "loss": 5.6746,
      "mean_token_accuracy": 0.20356673896312713,
      "num_tokens": 51286350.0,
      "step": 20235
    },
    {
      "entropy": 6.199232959747315,
      "epoch": 2.3358338141950377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004462640764077751,
      "loss": 5.5735,
      "mean_token_accuracy": 0.21453811079263688,
      "num_tokens": 51298262.0,
      "step": 20240
    },
    {
      "entropy": 6.280736064910888,
      "epoch": 2.336410848240046,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004462373116106971,
      "loss": 5.7144,
      "mean_token_accuracy": 0.19932273626327515,
      "num_tokens": 51310162.0,
      "step": 20245
    },
    {
      "entropy": 6.266496467590332,
      "epoch": 2.336987882285055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044621054105413704,
      "loss": 5.612,
      "mean_token_accuracy": 0.20644543766975404,
      "num_tokens": 51323468.0,
      "step": 20250
    },
    {
      "entropy": 6.3015974998474125,
      "epoch": 2.3375649163300634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044618376473899555,
      "loss": 5.6908,
      "mean_token_accuracy": 0.1973407030105591,
      "num_tokens": 51336674.0,
      "step": 20255
    },
    {
      "entropy": 6.286459064483642,
      "epoch": 2.338141950375072,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004461569826661732,
      "loss": 5.7339,
      "mean_token_accuracy": 0.20073004066944122,
      "num_tokens": 51349950.0,
      "step": 20260
    },
    {
      "entropy": 6.267491865158081,
      "epoch": 2.3387189844200806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004461301948365707,
      "loss": 5.7072,
      "mean_token_accuracy": 0.20379726439714432,
      "num_tokens": 51362191.0,
      "step": 20265
    },
    {
      "entropy": 6.227451467514038,
      "epoch": 2.3392960184650895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004461034012510891,
      "loss": 5.6166,
      "mean_token_accuracy": 0.20497980564832688,
      "num_tokens": 51374325.0,
      "step": 20270
    },
    {
      "entropy": 6.133911609649658,
      "epoch": 2.339873052510098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044607660191062963,
      "loss": 5.5029,
      "mean_token_accuracy": 0.21547901779413223,
      "num_tokens": 51386632.0,
      "step": 20275
    },
    {
      "entropy": 6.2999029636383055,
      "epoch": 2.340450086555107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044604979681609364,
      "loss": 5.7559,
      "mean_token_accuracy": 0.19529117792844772,
      "num_tokens": 51397876.0,
      "step": 20280
    },
    {
      "entropy": 6.262453269958496,
      "epoch": 2.341027120600115,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044602298596838264,
      "loss": 5.6767,
      "mean_token_accuracy": 0.20610849559307098,
      "num_tokens": 51411466.0,
      "step": 20285
    },
    {
      "entropy": 6.217646169662475,
      "epoch": 2.341604154645124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044599616936839853,
      "loss": 5.5861,
      "mean_token_accuracy": 0.2078832894563675,
      "num_tokens": 51423241.0,
      "step": 20290
    },
    {
      "entropy": 6.242675876617431,
      "epoch": 2.342181188690133,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004459693470170432,
      "loss": 5.6747,
      "mean_token_accuracy": 0.21084921658039094,
      "num_tokens": 51436056.0,
      "step": 20295
    },
    {
      "entropy": 6.2866333484649655,
      "epoch": 2.3427582227351413,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004459425189152187,
      "loss": 5.6865,
      "mean_token_accuracy": 0.20467903465032578,
      "num_tokens": 51449986.0,
      "step": 20300
    },
    {
      "entropy": 6.0869440078735355,
      "epoch": 2.34333525678015,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044591568506382757,
      "loss": 5.512,
      "mean_token_accuracy": 0.21996897757053374,
      "num_tokens": 51463784.0,
      "step": 20305
    },
    {
      "entropy": 6.161295318603516,
      "epoch": 2.3439122908251586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044588884546377226,
      "loss": 5.5563,
      "mean_token_accuracy": 0.21987365186214447,
      "num_tokens": 51476772.0,
      "step": 20310
    },
    {
      "entropy": 6.26782636642456,
      "epoch": 2.3444893248701675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004458620001159555,
      "loss": 5.6835,
      "mean_token_accuracy": 0.20060908943414688,
      "num_tokens": 51489889.0,
      "step": 20315
    },
    {
      "entropy": 6.231486797332764,
      "epoch": 2.345066358915176,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004458351490212803,
      "loss": 5.6875,
      "mean_token_accuracy": 0.20165486186742781,
      "num_tokens": 51503462.0,
      "step": 20320
    },
    {
      "entropy": 6.22360577583313,
      "epoch": 2.3456433929601848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044580829218064964,
      "loss": 5.5747,
      "mean_token_accuracy": 0.2104731783270836,
      "num_tokens": 51515338.0,
      "step": 20325
    },
    {
      "entropy": 6.211590671539307,
      "epoch": 2.346220427005193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000445781429594967,
      "loss": 5.6082,
      "mean_token_accuracy": 0.21087510585784913,
      "num_tokens": 51527990.0,
      "step": 20330
    },
    {
      "entropy": 6.263199234008789,
      "epoch": 2.346797461050202,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004457545612651357,
      "loss": 5.7009,
      "mean_token_accuracy": 0.1987906277179718,
      "num_tokens": 51542316.0,
      "step": 20335
    },
    {
      "entropy": 6.258873653411865,
      "epoch": 2.3473744950952105,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044572768719205964,
      "loss": 5.6455,
      "mean_token_accuracy": 0.19983627051115035,
      "num_tokens": 51556268.0,
      "step": 20340
    },
    {
      "entropy": 6.327695560455322,
      "epoch": 2.3479515291402193,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044570080737664264,
      "loss": 5.7313,
      "mean_token_accuracy": 0.20035652071237564,
      "num_tokens": 51568511.0,
      "step": 20345
    },
    {
      "entropy": 6.300572347640991,
      "epoch": 2.3485285631852277,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044567392181978874,
      "loss": 5.7322,
      "mean_token_accuracy": 0.2027884155511856,
      "num_tokens": 51581915.0,
      "step": 20350
    },
    {
      "entropy": 6.2482068061828615,
      "epoch": 2.3491055972302366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044564703052240223,
      "loss": 5.5632,
      "mean_token_accuracy": 0.206429186463356,
      "num_tokens": 51593544.0,
      "step": 20355
    },
    {
      "entropy": 6.225199222564697,
      "epoch": 2.349682631275245,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004456201334853876,
      "loss": 5.6447,
      "mean_token_accuracy": 0.20361943393945695,
      "num_tokens": 51605722.0,
      "step": 20360
    },
    {
      "entropy": 6.171707487106323,
      "epoch": 2.350259665320254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044559323070964956,
      "loss": 5.5619,
      "mean_token_accuracy": 0.21735349893569947,
      "num_tokens": 51618277.0,
      "step": 20365
    },
    {
      "entropy": 6.279447221755982,
      "epoch": 2.3508366993652627,
      "grad_norm": 1.015625,
      "learning_rate": 0.000445566322196093,
      "loss": 5.6386,
      "mean_token_accuracy": 0.20725711435079575,
      "num_tokens": 51629960.0,
      "step": 20370
    },
    {
      "entropy": 6.15372519493103,
      "epoch": 2.351413733410271,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004455394079456228,
      "loss": 5.5611,
      "mean_token_accuracy": 0.21145151555538177,
      "num_tokens": 51643089.0,
      "step": 20375
    },
    {
      "entropy": 6.301345109939575,
      "epoch": 2.35199076745528,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044551248795914446,
      "loss": 5.7577,
      "mean_token_accuracy": 0.19685018360614776,
      "num_tokens": 51655581.0,
      "step": 20380
    },
    {
      "entropy": 6.262241697311401,
      "epoch": 2.3525678015002884,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004454855622375632,
      "loss": 5.6393,
      "mean_token_accuracy": 0.19920710027217864,
      "num_tokens": 51667693.0,
      "step": 20385
    },
    {
      "entropy": 6.284205627441406,
      "epoch": 2.3531448355452973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004454586307817848,
      "loss": 5.7058,
      "mean_token_accuracy": 0.201881305873394,
      "num_tokens": 51680479.0,
      "step": 20390
    },
    {
      "entropy": 6.31897292137146,
      "epoch": 2.3537218695903057,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000445431693592715,
      "loss": 5.7588,
      "mean_token_accuracy": 0.19764790832996368,
      "num_tokens": 51693053.0,
      "step": 20395
    },
    {
      "entropy": 6.272164964675904,
      "epoch": 2.3542989036353146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004454047506712598,
      "loss": 5.7127,
      "mean_token_accuracy": 0.1975509226322174,
      "num_tokens": 51704974.0,
      "step": 20400
    },
    {
      "entropy": 6.269799041748047,
      "epoch": 2.354875937680323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044537780201832545,
      "loss": 5.6073,
      "mean_token_accuracy": 0.2105048656463623,
      "num_tokens": 51717190.0,
      "step": 20405
    },
    {
      "entropy": 6.223147296905518,
      "epoch": 2.355452971725332,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004453508476348184,
      "loss": 5.6756,
      "mean_token_accuracy": 0.20309789031744002,
      "num_tokens": 51730462.0,
      "step": 20410
    },
    {
      "entropy": 6.26445746421814,
      "epoch": 2.3560300057703403,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004453238875216452,
      "loss": 5.729,
      "mean_token_accuracy": 0.19753512591123581,
      "num_tokens": 51743723.0,
      "step": 20415
    },
    {
      "entropy": 6.272413158416748,
      "epoch": 2.356607039815349,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004452969216797127,
      "loss": 5.6361,
      "mean_token_accuracy": 0.20816679447889327,
      "num_tokens": 51756959.0,
      "step": 20420
    },
    {
      "entropy": 6.231748151779175,
      "epoch": 2.357184073860358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004452699501099277,
      "loss": 5.6892,
      "mean_token_accuracy": 0.1973074868321419,
      "num_tokens": 51768282.0,
      "step": 20425
    },
    {
      "entropy": 6.301117372512818,
      "epoch": 2.3577611079053664,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044524297281319766,
      "loss": 5.6995,
      "mean_token_accuracy": 0.20059530138969422,
      "num_tokens": 51780729.0,
      "step": 20430
    },
    {
      "entropy": 6.352606630325317,
      "epoch": 2.358338141950375,
      "grad_norm": 1.0,
      "learning_rate": 0.00044521598979042963,
      "loss": 5.7804,
      "mean_token_accuracy": 0.19381918460130693,
      "num_tokens": 51793661.0,
      "step": 20435
    },
    {
      "entropy": 6.334464693069458,
      "epoch": 2.3589151759953837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044518900104253154,
      "loss": 5.7166,
      "mean_token_accuracy": 0.20146120488643646,
      "num_tokens": 51806027.0,
      "step": 20440
    },
    {
      "entropy": 6.28181939125061,
      "epoch": 2.3594922100403926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004451620065704108,
      "loss": 5.7687,
      "mean_token_accuracy": 0.19285354763269424,
      "num_tokens": 51817495.0,
      "step": 20445
    },
    {
      "entropy": 6.267786502838135,
      "epoch": 2.360069244085401,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044513500637497546,
      "loss": 5.7067,
      "mean_token_accuracy": 0.20077207535505295,
      "num_tokens": 51829948.0,
      "step": 20450
    },
    {
      "entropy": 6.26269645690918,
      "epoch": 2.36064627813041,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044510800045713373,
      "loss": 5.763,
      "mean_token_accuracy": 0.1996346727013588,
      "num_tokens": 51842615.0,
      "step": 20455
    },
    {
      "entropy": 6.3423303127288815,
      "epoch": 2.3612233121754183,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044508098881779396,
      "loss": 5.6521,
      "mean_token_accuracy": 0.20848301351070403,
      "num_tokens": 51853785.0,
      "step": 20460
    },
    {
      "entropy": 6.240302801132202,
      "epoch": 2.361800346220427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004450539714578645,
      "loss": 5.5644,
      "mean_token_accuracy": 0.21347058415412903,
      "num_tokens": 51867427.0,
      "step": 20465
    },
    {
      "entropy": 6.330528354644775,
      "epoch": 2.3623773802654355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004450269483782543,
      "loss": 5.8257,
      "mean_token_accuracy": 0.1894884392619133,
      "num_tokens": 51880241.0,
      "step": 20470
    },
    {
      "entropy": 6.268656826019287,
      "epoch": 2.3629544143104444,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004449999195798721,
      "loss": 5.6768,
      "mean_token_accuracy": 0.1985153317451477,
      "num_tokens": 51891927.0,
      "step": 20475
    },
    {
      "entropy": 6.2747650146484375,
      "epoch": 2.363531448355453,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044497288506362706,
      "loss": 5.7059,
      "mean_token_accuracy": 0.19938716739416124,
      "num_tokens": 51905633.0,
      "step": 20480
    },
    {
      "entropy": 6.296726703643799,
      "epoch": 2.3641084824004617,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004449458448304284,
      "loss": 5.7228,
      "mean_token_accuracy": 0.19274458885192872,
      "num_tokens": 51919398.0,
      "step": 20485
    },
    {
      "entropy": 6.162575387954712,
      "epoch": 2.36468551644547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044491879888118574,
      "loss": 5.5151,
      "mean_token_accuracy": 0.21548093110322952,
      "num_tokens": 51932732.0,
      "step": 20490
    },
    {
      "entropy": 6.195258712768554,
      "epoch": 2.365262550490479,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004448917472168087,
      "loss": 5.6026,
      "mean_token_accuracy": 0.20719641596078872,
      "num_tokens": 51945830.0,
      "step": 20495
    },
    {
      "entropy": 6.2915153980255125,
      "epoch": 2.365839584535488,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044486468983820707,
      "loss": 5.6378,
      "mean_token_accuracy": 0.20723852813243865,
      "num_tokens": 51957643.0,
      "step": 20500
    },
    {
      "entropy": 6.228965425491333,
      "epoch": 2.3664166185804962,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000444837626746291,
      "loss": 5.6028,
      "mean_token_accuracy": 0.204250730574131,
      "num_tokens": 51969840.0,
      "step": 20505
    },
    {
      "entropy": 6.281548500061035,
      "epoch": 2.366993652625505,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004448105579419707,
      "loss": 5.7117,
      "mean_token_accuracy": 0.19711303263902663,
      "num_tokens": 51981250.0,
      "step": 20510
    },
    {
      "entropy": 6.310452318191528,
      "epoch": 2.3675706866705135,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004447834834261567,
      "loss": 5.6955,
      "mean_token_accuracy": 0.20483924001455306,
      "num_tokens": 51993312.0,
      "step": 20515
    },
    {
      "entropy": 6.2059783935546875,
      "epoch": 2.3681477207155224,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447564031997595,
      "loss": 5.6552,
      "mean_token_accuracy": 0.2059281960129738,
      "num_tokens": 52006697.0,
      "step": 20520
    },
    {
      "entropy": 6.220630645751953,
      "epoch": 2.368724754760531,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044472931726369,
      "loss": 5.6028,
      "mean_token_accuracy": 0.2104820802807808,
      "num_tokens": 52018938.0,
      "step": 20525
    },
    {
      "entropy": 6.184477710723877,
      "epoch": 2.3693017888055397,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044470222561885926,
      "loss": 5.6124,
      "mean_token_accuracy": 0.20970357060432435,
      "num_tokens": 52031309.0,
      "step": 20530
    },
    {
      "entropy": 6.2767784118652346,
      "epoch": 2.369878822850548,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044467512826617845,
      "loss": 5.6729,
      "mean_token_accuracy": 0.20308965742588042,
      "num_tokens": 52043661.0,
      "step": 20535
    },
    {
      "entropy": 6.242342472076416,
      "epoch": 2.370455856895557,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044464802520655897,
      "loss": 5.6567,
      "mean_token_accuracy": 0.20575435012578963,
      "num_tokens": 52055049.0,
      "step": 20540
    },
    {
      "entropy": 6.186476564407348,
      "epoch": 2.3710328909405654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004446209164409124,
      "loss": 5.658,
      "mean_token_accuracy": 0.20335730761289597,
      "num_tokens": 52068188.0,
      "step": 20545
    },
    {
      "entropy": 6.313190841674805,
      "epoch": 2.371609924985574,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004445938019701506,
      "loss": 5.6971,
      "mean_token_accuracy": 0.2033468797802925,
      "num_tokens": 52080662.0,
      "step": 20550
    },
    {
      "entropy": 6.317338705062866,
      "epoch": 2.3721869590305826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004445666817951855,
      "loss": 5.7162,
      "mean_token_accuracy": 0.20163785368204118,
      "num_tokens": 52093088.0,
      "step": 20555
    },
    {
      "entropy": 6.213424921035767,
      "epoch": 2.3727639930755915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004445395559169293,
      "loss": 5.6153,
      "mean_token_accuracy": 0.2092706009745598,
      "num_tokens": 52105796.0,
      "step": 20560
    },
    {
      "entropy": 6.273268795013427,
      "epoch": 2.3733410271206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445124243362943,
      "loss": 5.686,
      "mean_token_accuracy": 0.20361438393592834,
      "num_tokens": 52118146.0,
      "step": 20565
    },
    {
      "entropy": 6.201507329940796,
      "epoch": 2.3739180611656088,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004444852870541932,
      "loss": 5.5947,
      "mean_token_accuracy": 0.21228427141904832,
      "num_tokens": 52131033.0,
      "step": 20570
    },
    {
      "entropy": 6.2907613754272464,
      "epoch": 2.3744950952106176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004444581440715386,
      "loss": 5.7203,
      "mean_token_accuracy": 0.2021082565188408,
      "num_tokens": 52143742.0,
      "step": 20575
    },
    {
      "entropy": 6.230777454376221,
      "epoch": 2.375072129255626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044443099538924347,
      "loss": 5.5845,
      "mean_token_accuracy": 0.20782473981380462,
      "num_tokens": 52156091.0,
      "step": 20580
    },
    {
      "entropy": 6.29959774017334,
      "epoch": 2.375649163300635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444038410082211,
      "loss": 5.7438,
      "mean_token_accuracy": 0.19574818015098572,
      "num_tokens": 52168404.0,
      "step": 20585
    },
    {
      "entropy": 6.201791667938233,
      "epoch": 2.3762261973456433,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004443766809293846,
      "loss": 5.5777,
      "mean_token_accuracy": 0.211177359521389,
      "num_tokens": 52180660.0,
      "step": 20590
    },
    {
      "entropy": 6.262328720092773,
      "epoch": 2.376803231390652,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004443495151536475,
      "loss": 5.6617,
      "mean_token_accuracy": 0.20792763084173202,
      "num_tokens": 52193752.0,
      "step": 20595
    },
    {
      "entropy": 6.345590829849243,
      "epoch": 2.3773802654356606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004443223436819237,
      "loss": 5.7176,
      "mean_token_accuracy": 0.19998614639043807,
      "num_tokens": 52205686.0,
      "step": 20600
    },
    {
      "entropy": 6.256855201721192,
      "epoch": 2.3779572994806695,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044429516651512687,
      "loss": 5.7057,
      "mean_token_accuracy": 0.1990906223654747,
      "num_tokens": 52218743.0,
      "step": 20605
    },
    {
      "entropy": 6.284265804290771,
      "epoch": 2.378534333525678,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044426798365417133,
      "loss": 5.5578,
      "mean_token_accuracy": 0.20861430764198302,
      "num_tokens": 52231760.0,
      "step": 20610
    },
    {
      "entropy": 6.282824516296387,
      "epoch": 2.3791113675706868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044424079509997104,
      "loss": 5.6959,
      "mean_token_accuracy": 0.1966059237718582,
      "num_tokens": 52243146.0,
      "step": 20615
    },
    {
      "entropy": 6.2703900814056395,
      "epoch": 2.379688401615695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004442136008534409,
      "loss": 5.702,
      "mean_token_accuracy": 0.20122209936380386,
      "num_tokens": 52255330.0,
      "step": 20620
    },
    {
      "entropy": 6.279065322875977,
      "epoch": 2.380265435660704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044418640091549525,
      "loss": 5.7076,
      "mean_token_accuracy": 0.20006371140480042,
      "num_tokens": 52266563.0,
      "step": 20625
    },
    {
      "entropy": 6.275199604034424,
      "epoch": 2.3808424697057124,
      "grad_norm": 0.984375,
      "learning_rate": 0.000444159195287049,
      "loss": 5.7504,
      "mean_token_accuracy": 0.19873536825180055,
      "num_tokens": 52279137.0,
      "step": 20630
    },
    {
      "entropy": 6.206452178955078,
      "epoch": 2.3814195037507213,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004441319839690173,
      "loss": 5.6645,
      "mean_token_accuracy": 0.20957825034856797,
      "num_tokens": 52293317.0,
      "step": 20635
    },
    {
      "entropy": 6.326365232467651,
      "epoch": 2.3819965377957297,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004441047669623153,
      "loss": 5.7161,
      "mean_token_accuracy": 0.1966287985444069,
      "num_tokens": 52306383.0,
      "step": 20640
    },
    {
      "entropy": 6.226604080200195,
      "epoch": 2.3825735718407386,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044407754426785845,
      "loss": 5.6056,
      "mean_token_accuracy": 0.20549362599849702,
      "num_tokens": 52318446.0,
      "step": 20645
    },
    {
      "entropy": 6.285811233520508,
      "epoch": 2.3831506058857475,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004440503158865625,
      "loss": 5.6457,
      "mean_token_accuracy": 0.20058149844408035,
      "num_tokens": 52330761.0,
      "step": 20650
    },
    {
      "entropy": 6.372319030761719,
      "epoch": 2.383727639930756,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044402308181934295,
      "loss": 5.7142,
      "mean_token_accuracy": 0.1901174381375313,
      "num_tokens": 52342906.0,
      "step": 20655
    },
    {
      "entropy": 6.240784454345703,
      "epoch": 2.3843046739757647,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004439958420671162,
      "loss": 5.6703,
      "mean_token_accuracy": 0.2053072065114975,
      "num_tokens": 52355928.0,
      "step": 20660
    },
    {
      "entropy": 6.261500549316406,
      "epoch": 2.384881708020773,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044396859663079814,
      "loss": 5.6843,
      "mean_token_accuracy": 0.198262619972229,
      "num_tokens": 52368370.0,
      "step": 20665
    },
    {
      "entropy": 6.348115158081055,
      "epoch": 2.385458742065782,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00044394134551130533,
      "loss": 5.7175,
      "mean_token_accuracy": 0.19636475592851638,
      "num_tokens": 52382636.0,
      "step": 20670
    },
    {
      "entropy": 6.261222696304321,
      "epoch": 2.3860357761107904,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004439140887095542,
      "loss": 5.6412,
      "mean_token_accuracy": 0.21150686740875244,
      "num_tokens": 52395745.0,
      "step": 20675
    },
    {
      "entropy": 6.2865699291229244,
      "epoch": 2.3866128101557993,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044388682622646165,
      "loss": 5.6257,
      "mean_token_accuracy": 0.2079330489039421,
      "num_tokens": 52408929.0,
      "step": 20680
    },
    {
      "entropy": 6.236780977249145,
      "epoch": 2.3871898442008077,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004438595580629446,
      "loss": 5.6058,
      "mean_token_accuracy": 0.20728029310703278,
      "num_tokens": 52421626.0,
      "step": 20685
    },
    {
      "entropy": 6.3429218292236325,
      "epoch": 2.3877668782458166,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004438322842199202,
      "loss": 5.7593,
      "mean_token_accuracy": 0.19178105741739274,
      "num_tokens": 52434139.0,
      "step": 20690
    },
    {
      "entropy": 6.257878828048706,
      "epoch": 2.388343912290825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004438050046983057,
      "loss": 5.6522,
      "mean_token_accuracy": 0.20227349400520325,
      "num_tokens": 52446141.0,
      "step": 20695
    },
    {
      "entropy": 6.26620659828186,
      "epoch": 2.388920946335834,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044377771949901876,
      "loss": 5.6836,
      "mean_token_accuracy": 0.20275806486606598,
      "num_tokens": 52459110.0,
      "step": 20700
    },
    {
      "entropy": 6.2968615055084225,
      "epoch": 2.3894979803808427,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044375042862297703,
      "loss": 5.7898,
      "mean_token_accuracy": 0.19312844276428223,
      "num_tokens": 52471985.0,
      "step": 20705
    },
    {
      "entropy": 6.314662981033325,
      "epoch": 2.390075014425851,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044372313207109856,
      "loss": 5.7307,
      "mean_token_accuracy": 0.2024776130914688,
      "num_tokens": 52484226.0,
      "step": 20710
    },
    {
      "entropy": 6.280565023422241,
      "epoch": 2.3906520484708595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044369582984430127,
      "loss": 5.6635,
      "mean_token_accuracy": 0.20388756394386293,
      "num_tokens": 52496141.0,
      "step": 20715
    },
    {
      "entropy": 6.239601516723633,
      "epoch": 2.3912290825158684,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044366852194350354,
      "loss": 5.6939,
      "mean_token_accuracy": 0.19880139082670212,
      "num_tokens": 52509231.0,
      "step": 20720
    },
    {
      "entropy": 6.2407002449035645,
      "epoch": 2.3918061165608773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004436412083696239,
      "loss": 5.6169,
      "mean_token_accuracy": 0.20698375701904298,
      "num_tokens": 52522619.0,
      "step": 20725
    },
    {
      "entropy": 6.35409779548645,
      "epoch": 2.3923831506058857,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044361388912358095,
      "loss": 5.6889,
      "mean_token_accuracy": 0.19898567199707032,
      "num_tokens": 52534241.0,
      "step": 20730
    },
    {
      "entropy": 6.3576192378997805,
      "epoch": 2.3929601846508946,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004435865642062936,
      "loss": 5.6701,
      "mean_token_accuracy": 0.1995391771197319,
      "num_tokens": 52546306.0,
      "step": 20735
    },
    {
      "entropy": 6.314690685272216,
      "epoch": 2.393537218695903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004435592336186809,
      "loss": 5.7571,
      "mean_token_accuracy": 0.19136658757925035,
      "num_tokens": 52559855.0,
      "step": 20740
    },
    {
      "entropy": 6.176814842224121,
      "epoch": 2.394114252740912,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004435318973616622,
      "loss": 5.5541,
      "mean_token_accuracy": 0.21697945296764373,
      "num_tokens": 52574071.0,
      "step": 20745
    },
    {
      "entropy": 6.2600531578063965,
      "epoch": 2.3946912867859202,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044350455543615665,
      "loss": 5.6525,
      "mean_token_accuracy": 0.20137125849723816,
      "num_tokens": 52585983.0,
      "step": 20750
    },
    {
      "entropy": 6.256138610839844,
      "epoch": 2.395268320830929,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004434772078430843,
      "loss": 5.5896,
      "mean_token_accuracy": 0.20987087488174438,
      "num_tokens": 52598399.0,
      "step": 20755
    },
    {
      "entropy": 6.2769557476043705,
      "epoch": 2.3958453548759375,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004434498545833646,
      "loss": 5.7124,
      "mean_token_accuracy": 0.19558157175779342,
      "num_tokens": 52610172.0,
      "step": 20760
    },
    {
      "entropy": 6.26928448677063,
      "epoch": 2.3964223889209464,
      "grad_norm": 1.0,
      "learning_rate": 0.0004434224956579177,
      "loss": 5.6461,
      "mean_token_accuracy": 0.21473123431205748,
      "num_tokens": 52622626.0,
      "step": 20765
    },
    {
      "entropy": 6.291495132446289,
      "epoch": 2.396999422965955,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004433951310676639,
      "loss": 5.6781,
      "mean_token_accuracy": 0.1995982840657234,
      "num_tokens": 52633893.0,
      "step": 20770
    },
    {
      "entropy": 6.303919506072998,
      "epoch": 2.3975764570109637,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044336776081352347,
      "loss": 5.6773,
      "mean_token_accuracy": 0.2000240921974182,
      "num_tokens": 52645674.0,
      "step": 20775
    },
    {
      "entropy": 6.325625371932984,
      "epoch": 2.3981534910559725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044334038489641706,
      "loss": 5.7368,
      "mean_token_accuracy": 0.19996620267629622,
      "num_tokens": 52659030.0,
      "step": 20780
    },
    {
      "entropy": 6.242504549026489,
      "epoch": 2.398730525100981,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044331300331726544,
      "loss": 5.6485,
      "mean_token_accuracy": 0.20337757021188735,
      "num_tokens": 52670680.0,
      "step": 20785
    },
    {
      "entropy": 6.3164763927459715,
      "epoch": 2.3993075591459894,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044328561607698947,
      "loss": 5.7851,
      "mean_token_accuracy": 0.19662774354219437,
      "num_tokens": 52682860.0,
      "step": 20790
    },
    {
      "entropy": 6.218564176559449,
      "epoch": 2.3998845931909982,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004432582231765105,
      "loss": 5.6275,
      "mean_token_accuracy": 0.20385155081748962,
      "num_tokens": 52695566.0,
      "step": 20795
    },
    {
      "entropy": 6.248559188842774,
      "epoch": 2.400461627236007,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044323082461674974,
      "loss": 5.6073,
      "mean_token_accuracy": 0.20926202684640885,
      "num_tokens": 52708409.0,
      "step": 20800
    },
    {
      "entropy": 6.272957420349121,
      "epoch": 2.4010386612810155,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004432034203986287,
      "loss": 5.6408,
      "mean_token_accuracy": 0.20582594424486161,
      "num_tokens": 52721637.0,
      "step": 20805
    },
    {
      "entropy": 6.288475513458252,
      "epoch": 2.4016156953260244,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004431760105230693,
      "loss": 5.7064,
      "mean_token_accuracy": 0.19705838412046434,
      "num_tokens": 52732157.0,
      "step": 20810
    },
    {
      "entropy": 6.24582257270813,
      "epoch": 2.402192729371033,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044314859499099325,
      "loss": 5.6957,
      "mean_token_accuracy": 0.20342942029237748,
      "num_tokens": 52745038.0,
      "step": 20815
    },
    {
      "entropy": 6.2946062088012695,
      "epoch": 2.4027697634160416,
      "grad_norm": 1.0,
      "learning_rate": 0.00044312117380332274,
      "loss": 5.6468,
      "mean_token_accuracy": 0.20523704588413239,
      "num_tokens": 52757381.0,
      "step": 20820
    },
    {
      "entropy": 6.178961086273193,
      "epoch": 2.40334679746105,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044309374696098,
      "loss": 5.5987,
      "mean_token_accuracy": 0.2124863639473915,
      "num_tokens": 52770604.0,
      "step": 20825
    },
    {
      "entropy": 6.212749719619751,
      "epoch": 2.403923831506059,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004430663144648876,
      "loss": 5.6055,
      "mean_token_accuracy": 0.21524260938167572,
      "num_tokens": 52784634.0,
      "step": 20830
    },
    {
      "entropy": 6.252115392684937,
      "epoch": 2.4045008655510673,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044303887631596823,
      "loss": 5.6607,
      "mean_token_accuracy": 0.20367956459522246,
      "num_tokens": 52798417.0,
      "step": 20835
    },
    {
      "entropy": 6.386414480209351,
      "epoch": 2.405077899596076,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004430114325151447,
      "loss": 5.8127,
      "mean_token_accuracy": 0.19441916793584824,
      "num_tokens": 52810885.0,
      "step": 20840
    },
    {
      "entropy": 6.256485033035278,
      "epoch": 2.4056549336410846,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004429839830633401,
      "loss": 5.6499,
      "mean_token_accuracy": 0.20765568166971207,
      "num_tokens": 52824450.0,
      "step": 20845
    },
    {
      "entropy": 6.250825023651123,
      "epoch": 2.4062319676860935,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004429565279614777,
      "loss": 5.6702,
      "mean_token_accuracy": 0.20331302881240845,
      "num_tokens": 52837053.0,
      "step": 20850
    },
    {
      "entropy": 6.264067935943603,
      "epoch": 2.4068090017311023,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044292906721048094,
      "loss": 5.6769,
      "mean_token_accuracy": 0.20398466140031815,
      "num_tokens": 52851695.0,
      "step": 20855
    },
    {
      "entropy": 6.331319522857666,
      "epoch": 2.4073860357761108,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004429016008112733,
      "loss": 5.6837,
      "mean_token_accuracy": 0.19836791157722472,
      "num_tokens": 52862999.0,
      "step": 20860
    },
    {
      "entropy": 6.212860012054444,
      "epoch": 2.4079630698211196,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004428741287647788,
      "loss": 5.6076,
      "mean_token_accuracy": 0.2104809895157814,
      "num_tokens": 52875224.0,
      "step": 20865
    },
    {
      "entropy": 6.277280998229981,
      "epoch": 2.408540103866128,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044284665107192136,
      "loss": 5.7049,
      "mean_token_accuracy": 0.19749379605054856,
      "num_tokens": 52887296.0,
      "step": 20870
    },
    {
      "entropy": 6.261632585525513,
      "epoch": 2.409117137911137,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004428191677336251,
      "loss": 5.6026,
      "mean_token_accuracy": 0.20402903109788895,
      "num_tokens": 52900670.0,
      "step": 20875
    },
    {
      "entropy": 6.338109302520752,
      "epoch": 2.4096941719561453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004427916787508146,
      "loss": 5.7901,
      "mean_token_accuracy": 0.19941924065351485,
      "num_tokens": 52914515.0,
      "step": 20880
    },
    {
      "entropy": 6.187658214569092,
      "epoch": 2.410271206001154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004427641841244144,
      "loss": 5.5725,
      "mean_token_accuracy": 0.21277358680963515,
      "num_tokens": 52928448.0,
      "step": 20885
    },
    {
      "entropy": 6.275861692428589,
      "epoch": 2.4108482400461626,
      "grad_norm": 0.90625,
      "learning_rate": 0.000442736683855349,
      "loss": 5.6697,
      "mean_token_accuracy": 0.2048983320593834,
      "num_tokens": 52941381.0,
      "step": 20890
    },
    {
      "entropy": 6.235480117797851,
      "epoch": 2.4114252740911715,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004427091779445437,
      "loss": 5.6208,
      "mean_token_accuracy": 0.19731384217739106,
      "num_tokens": 52954959.0,
      "step": 20895
    },
    {
      "entropy": 6.272963142395019,
      "epoch": 2.41200230813618,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004426816663929235,
      "loss": 5.6502,
      "mean_token_accuracy": 0.20786180794239045,
      "num_tokens": 52967956.0,
      "step": 20900
    },
    {
      "entropy": 6.2870433807373045,
      "epoch": 2.4125793421811887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044265414920141366,
      "loss": 5.7109,
      "mean_token_accuracy": 0.20373185575008393,
      "num_tokens": 52980997.0,
      "step": 20905
    },
    {
      "entropy": 6.300219583511352,
      "epoch": 2.413156376226197,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00044262662637093987,
      "loss": 5.6935,
      "mean_token_accuracy": 0.19570810943841935,
      "num_tokens": 52994636.0,
      "step": 20910
    },
    {
      "entropy": 6.237855911254883,
      "epoch": 2.413733410271206,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044259909790242776,
      "loss": 5.6419,
      "mean_token_accuracy": 0.19943866580724717,
      "num_tokens": 53007065.0,
      "step": 20915
    },
    {
      "entropy": 6.204882478713989,
      "epoch": 2.4143104443162144,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004425715637968032,
      "loss": 5.6102,
      "mean_token_accuracy": 0.20188685655593872,
      "num_tokens": 53020479.0,
      "step": 20920
    },
    {
      "entropy": 6.294823837280274,
      "epoch": 2.4148874783612233,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004425440240549923,
      "loss": 5.729,
      "mean_token_accuracy": 0.20429362654685973,
      "num_tokens": 53036390.0,
      "step": 20925
    },
    {
      "entropy": 6.316632652282715,
      "epoch": 2.415464512406232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044251647867792147,
      "loss": 5.621,
      "mean_token_accuracy": 0.20305669605731963,
      "num_tokens": 53048604.0,
      "step": 20930
    },
    {
      "entropy": 6.298521041870117,
      "epoch": 2.4160415464512406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044248892766651706,
      "loss": 5.7039,
      "mean_token_accuracy": 0.20051255971193313,
      "num_tokens": 53062774.0,
      "step": 20935
    },
    {
      "entropy": 6.273114824295044,
      "epoch": 2.4166185804962494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004424613710217057,
      "loss": 5.6487,
      "mean_token_accuracy": 0.20796942263841628,
      "num_tokens": 53076701.0,
      "step": 20940
    },
    {
      "entropy": 6.320744132995605,
      "epoch": 2.417195614541258,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044243380874441437,
      "loss": 5.7188,
      "mean_token_accuracy": 0.20322779566049576,
      "num_tokens": 53089950.0,
      "step": 20945
    },
    {
      "entropy": 6.215546131134033,
      "epoch": 2.4177726485862667,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044240624083557,
      "loss": 5.5759,
      "mean_token_accuracy": 0.20332265943288802,
      "num_tokens": 53103622.0,
      "step": 20950
    },
    {
      "entropy": 6.292780828475952,
      "epoch": 2.418349682631275,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044237866729609994,
      "loss": 5.7713,
      "mean_token_accuracy": 0.20153989493846894,
      "num_tokens": 53116382.0,
      "step": 20955
    },
    {
      "entropy": 6.321187973022461,
      "epoch": 2.418926716676284,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004423510881269315,
      "loss": 5.6753,
      "mean_token_accuracy": 0.20557799339294433,
      "num_tokens": 53128759.0,
      "step": 20960
    },
    {
      "entropy": 6.264397716522216,
      "epoch": 2.4195037507212924,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004423235033289924,
      "loss": 5.6817,
      "mean_token_accuracy": 0.2095179796218872,
      "num_tokens": 53141131.0,
      "step": 20965
    },
    {
      "entropy": 6.260027551651001,
      "epoch": 2.4200807847663013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004422959129032103,
      "loss": 5.7133,
      "mean_token_accuracy": 0.20061449110507965,
      "num_tokens": 53152486.0,
      "step": 20970
    },
    {
      "entropy": 6.249807596206665,
      "epoch": 2.4206578188113097,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044226831685051333,
      "loss": 5.603,
      "mean_token_accuracy": 0.20373159945011138,
      "num_tokens": 53164897.0,
      "step": 20975
    },
    {
      "entropy": 6.303619050979615,
      "epoch": 2.4212348528563186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004422407151718296,
      "loss": 5.7379,
      "mean_token_accuracy": 0.1971864342689514,
      "num_tokens": 53177759.0,
      "step": 20980
    },
    {
      "entropy": 6.338917827606201,
      "epoch": 2.4218118869013274,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044221310786808746,
      "loss": 5.7051,
      "mean_token_accuracy": 0.20118940770626068,
      "num_tokens": 53191985.0,
      "step": 20985
    },
    {
      "entropy": 6.262404060363769,
      "epoch": 2.422388920946336,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004421854949402155,
      "loss": 5.5662,
      "mean_token_accuracy": 0.20347983986139298,
      "num_tokens": 53203531.0,
      "step": 20990
    },
    {
      "entropy": 6.258559465408325,
      "epoch": 2.4229659549913443,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044215787638914245,
      "loss": 5.6769,
      "mean_token_accuracy": 0.19556978344917297,
      "num_tokens": 53216440.0,
      "step": 20995
    },
    {
      "entropy": 6.309841728210449,
      "epoch": 2.423542989036353,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004421302522157973,
      "loss": 5.7298,
      "mean_token_accuracy": 0.1915926992893219,
      "num_tokens": 53230674.0,
      "step": 21000
    },
    {
      "epoch": 2.423542989036353,
      "eval_entropy": 6.090504400832496,
      "eval_loss": 5.771459579467773,
      "eval_mean_token_accuracy": 0.2048968357527361,
      "eval_num_tokens": 53230674.0,
      "eval_runtime": 17.5353,
      "eval_samples_per_second": 1604.536,
      "eval_steps_per_second": 200.567,
      "step": 21000
    }
  ],
  "logging_steps": 5,
  "max_steps": 86650,
  "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": 1.163850898046976e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}