hin-deva-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5763688760806917,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.905435800552368,
      "epoch": 0.0004803073967339097,
      "grad_norm": 15.375,
      "learning_rate": 2e-06,
      "loss": 14.4544,
      "mean_token_accuracy": 6.385695887729525e-05,
      "num_tokens": 10855.0,
      "step": 5
    },
    {
      "entropy": 4.913383293151855,
      "epoch": 0.0009606147934678194,
      "grad_norm": 16.625,
      "learning_rate": 4.5e-06,
      "loss": 14.5216,
      "mean_token_accuracy": 0.00014638935390394182,
      "num_tokens": 24110.0,
      "step": 10
    },
    {
      "entropy": 4.925370550155639,
      "epoch": 0.001440922190201729,
      "grad_norm": 18.125,
      "learning_rate": 7e-06,
      "loss": 14.412,
      "mean_token_accuracy": 6.793478387407958e-05,
      "num_tokens": 35984.0,
      "step": 15
    },
    {
      "entropy": 4.947782182693482,
      "epoch": 0.0019212295869356388,
      "grad_norm": 19.5,
      "learning_rate": 9.5e-06,
      "loss": 14.2345,
      "mean_token_accuracy": 9.48766595683992e-05,
      "num_tokens": 48152.0,
      "step": 20
    },
    {
      "entropy": 5.064438819885254,
      "epoch": 0.0024015369836695487,
      "grad_norm": 23.75,
      "learning_rate": 1.2e-05,
      "loss": 13.7842,
      "mean_token_accuracy": 0.00013405063073150812,
      "num_tokens": 59810.0,
      "step": 25
    },
    {
      "entropy": 5.404191589355468,
      "epoch": 0.002881844380403458,
      "grad_norm": 31.125,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.1568,
      "mean_token_accuracy": 8.628127980045975e-05,
      "num_tokens": 70852.0,
      "step": 30
    },
    {
      "entropy": 7.142218971252442,
      "epoch": 0.0033621517771373678,
      "grad_norm": 21.0,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9814,
      "mean_token_accuracy": 5.5741361575201156e-05,
      "num_tokens": 83378.0,
      "step": 35
    },
    {
      "entropy": 10.262308692932129,
      "epoch": 0.0038424591738712775,
      "grad_norm": 3.5625,
      "learning_rate": 1.95e-05,
      "loss": 10.9117,
      "mean_token_accuracy": 9.433962404727935e-05,
      "num_tokens": 95505.0,
      "step": 40
    },
    {
      "entropy": 10.698264122009277,
      "epoch": 0.004322766570605188,
      "grad_norm": 3.203125,
      "learning_rate": 2.2e-05,
      "loss": 10.7241,
      "mean_token_accuracy": 0.0005975825013592839,
      "num_tokens": 106812.0,
      "step": 45
    },
    {
      "entropy": 10.737651824951172,
      "epoch": 0.004803073967339097,
      "grad_norm": 2.78125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.5853,
      "mean_token_accuracy": 0.009875867154914886,
      "num_tokens": 118572.0,
      "step": 50
    },
    {
      "entropy": 10.73659906387329,
      "epoch": 0.005283381364073006,
      "grad_norm": 2.109375,
      "learning_rate": 2.7e-05,
      "loss": 10.3833,
      "mean_token_accuracy": 0.014529264252632856,
      "num_tokens": 130051.0,
      "step": 55
    },
    {
      "entropy": 10.701386642456054,
      "epoch": 0.005763688760806916,
      "grad_norm": 2.046875,
      "learning_rate": 2.95e-05,
      "loss": 10.2322,
      "mean_token_accuracy": 0.014093486312776803,
      "num_tokens": 141920.0,
      "step": 60
    },
    {
      "entropy": 10.711284351348876,
      "epoch": 0.006243996157540826,
      "grad_norm": 1.796875,
      "learning_rate": 3.2e-05,
      "loss": 10.045,
      "mean_token_accuracy": 0.015402647480368615,
      "num_tokens": 152706.0,
      "step": 65
    },
    {
      "entropy": 10.732075595855713,
      "epoch": 0.0067243035542747355,
      "grad_norm": 1.8125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.9817,
      "mean_token_accuracy": 0.013764715194702149,
      "num_tokens": 165253.0,
      "step": 70
    },
    {
      "entropy": 10.734086227416991,
      "epoch": 0.007204610951008645,
      "grad_norm": 1.8046875,
      "learning_rate": 3.7e-05,
      "loss": 9.8756,
      "mean_token_accuracy": 0.02153849443420768,
      "num_tokens": 176708.0,
      "step": 75
    },
    {
      "entropy": 10.729099750518799,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.8125,
      "learning_rate": 3.95e-05,
      "loss": 9.8268,
      "mean_token_accuracy": 0.020801819767802952,
      "num_tokens": 188240.0,
      "step": 80
    },
    {
      "entropy": 10.723050022125244,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.796875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6525,
      "mean_token_accuracy": 0.01858828719705343,
      "num_tokens": 198355.0,
      "step": 85
    },
    {
      "entropy": 10.719482040405273,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.71875,
      "learning_rate": 4.45e-05,
      "loss": 9.5811,
      "mean_token_accuracy": 0.01679358258843422,
      "num_tokens": 209497.0,
      "step": 90
    },
    {
      "entropy": 10.711874198913574,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.828125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5018,
      "mean_token_accuracy": 0.016100958082824947,
      "num_tokens": 220859.0,
      "step": 95
    },
    {
      "entropy": 10.695251655578613,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4214,
      "mean_token_accuracy": 0.017460816819220782,
      "num_tokens": 231550.0,
      "step": 100
    },
    {
      "entropy": 10.673291873931884,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.671875,
      "learning_rate": 5.2e-05,
      "loss": 9.2854,
      "mean_token_accuracy": 0.016554870828986167,
      "num_tokens": 244210.0,
      "step": 105
    },
    {
      "entropy": 10.652278709411622,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.7578125,
      "learning_rate": 5.45e-05,
      "loss": 9.1581,
      "mean_token_accuracy": 0.024293087422847748,
      "num_tokens": 255745.0,
      "step": 110
    },
    {
      "entropy": 10.617483043670655,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.5703125,
      "learning_rate": 5.7e-05,
      "loss": 9.055,
      "mean_token_accuracy": 0.03160325270146132,
      "num_tokens": 266180.0,
      "step": 115
    },
    {
      "entropy": 10.569834327697754,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.65625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.939,
      "mean_token_accuracy": 0.025451960042119026,
      "num_tokens": 277736.0,
      "step": 120
    },
    {
      "entropy": 10.52544527053833,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.59375,
      "learning_rate": 6.2e-05,
      "loss": 8.8203,
      "mean_token_accuracy": 0.034478765353560446,
      "num_tokens": 289069.0,
      "step": 125
    },
    {
      "entropy": 10.468996620178222,
      "epoch": 0.012487992315081652,
      "grad_norm": 1.5546875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.6972,
      "mean_token_accuracy": 0.03924299031496048,
      "num_tokens": 300240.0,
      "step": 130
    },
    {
      "entropy": 10.390470600128173,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.5390625,
      "learning_rate": 6.7e-05,
      "loss": 8.622,
      "mean_token_accuracy": 0.033455712348222734,
      "num_tokens": 311698.0,
      "step": 135
    },
    {
      "entropy": 10.308708190917969,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.359375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.4986,
      "mean_token_accuracy": 0.036465326137840745,
      "num_tokens": 322844.0,
      "step": 140
    },
    {
      "entropy": 10.209622859954834,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.296875,
      "learning_rate": 7.2e-05,
      "loss": 8.3864,
      "mean_token_accuracy": 0.04093883950263262,
      "num_tokens": 335092.0,
      "step": 145
    },
    {
      "entropy": 10.091598796844483,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.2578125,
      "learning_rate": 7.45e-05,
      "loss": 8.2513,
      "mean_token_accuracy": 0.041372421011328694,
      "num_tokens": 347033.0,
      "step": 150
    },
    {
      "entropy": 9.934578609466552,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 8.133,
      "mean_token_accuracy": 0.04138694703578949,
      "num_tokens": 358696.0,
      "step": 155
    },
    {
      "entropy": 9.762340259552001,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.1328125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0344,
      "mean_token_accuracy": 0.04901535548269749,
      "num_tokens": 369390.0,
      "step": 160
    },
    {
      "entropy": 9.557822990417481,
      "epoch": 0.01585014409221902,
      "grad_norm": 0.984375,
      "learning_rate": 8.2e-05,
      "loss": 7.9344,
      "mean_token_accuracy": 0.04700087420642376,
      "num_tokens": 380540.0,
      "step": 165
    },
    {
      "entropy": 9.355019855499268,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.015625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.8587,
      "mean_token_accuracy": 0.04431998413056135,
      "num_tokens": 391243.0,
      "step": 170
    },
    {
      "entropy": 9.114866542816163,
      "epoch": 0.01681075888568684,
      "grad_norm": 0.8671875,
      "learning_rate": 8.7e-05,
      "loss": 7.7947,
      "mean_token_accuracy": 0.05138894282281399,
      "num_tokens": 403336.0,
      "step": 175
    },
    {
      "entropy": 8.863803100585937,
      "epoch": 0.01729106628242075,
      "grad_norm": 0.83984375,
      "learning_rate": 8.95e-05,
      "loss": 7.693,
      "mean_token_accuracy": 0.054227447509765624,
      "num_tokens": 413886.0,
      "step": 180
    },
    {
      "entropy": 8.668179988861084,
      "epoch": 0.01777137367915466,
      "grad_norm": 0.95703125,
      "learning_rate": 9.2e-05,
      "loss": 7.7008,
      "mean_token_accuracy": 0.05373974815011025,
      "num_tokens": 425277.0,
      "step": 185
    },
    {
      "entropy": 8.488366603851318,
      "epoch": 0.01825168107588857,
      "grad_norm": 0.72265625,
      "learning_rate": 9.45e-05,
      "loss": 7.7708,
      "mean_token_accuracy": 0.05546994470059872,
      "num_tokens": 436868.0,
      "step": 190
    },
    {
      "entropy": 8.390744590759278,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.078125,
      "learning_rate": 9.7e-05,
      "loss": 7.5877,
      "mean_token_accuracy": 0.052850323542952535,
      "num_tokens": 448349.0,
      "step": 195
    },
    {
      "entropy": 8.283522891998292,
      "epoch": 0.01921229586935639,
      "grad_norm": 0.77734375,
      "learning_rate": 9.95e-05,
      "loss": 7.5809,
      "mean_token_accuracy": 0.05953923240303993,
      "num_tokens": 459447.0,
      "step": 200
    },
    {
      "entropy": 8.150887298583985,
      "epoch": 0.0196926032660903,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000102,
      "loss": 7.5463,
      "mean_token_accuracy": 0.05646048374474048,
      "num_tokens": 470734.0,
      "step": 205
    },
    {
      "entropy": 8.137212657928467,
      "epoch": 0.020172910662824207,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.5847,
      "mean_token_accuracy": 0.06030619703233242,
      "num_tokens": 482015.0,
      "step": 210
    },
    {
      "entropy": 8.123589897155762,
      "epoch": 0.020653218059558116,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000107,
      "loss": 7.5311,
      "mean_token_accuracy": 0.06505471915006637,
      "num_tokens": 493339.0,
      "step": 215
    },
    {
      "entropy": 8.091688442230225,
      "epoch": 0.021133525456292025,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0001095,
      "loss": 7.4806,
      "mean_token_accuracy": 0.06740148589015008,
      "num_tokens": 504924.0,
      "step": 220
    },
    {
      "entropy": 8.105245876312257,
      "epoch": 0.021613832853025938,
      "grad_norm": 1.25,
      "learning_rate": 0.000112,
      "loss": 7.4493,
      "mean_token_accuracy": 0.06430868282914162,
      "num_tokens": 516603.0,
      "step": 225
    },
    {
      "entropy": 8.071302318572998,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0001145,
      "loss": 7.3628,
      "mean_token_accuracy": 0.07100334689021111,
      "num_tokens": 528347.0,
      "step": 230
    },
    {
      "entropy": 8.018728876113892,
      "epoch": 0.022574447646493755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.4365,
      "mean_token_accuracy": 0.07649173960089684,
      "num_tokens": 539328.0,
      "step": 235
    },
    {
      "entropy": 8.024458360671996,
      "epoch": 0.023054755043227664,
      "grad_norm": 0.921875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.4121,
      "mean_token_accuracy": 0.07992132529616355,
      "num_tokens": 549297.0,
      "step": 240
    },
    {
      "entropy": 7.979285478591919,
      "epoch": 0.023535062439961577,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000122,
      "loss": 7.373,
      "mean_token_accuracy": 0.07391561307013035,
      "num_tokens": 560306.0,
      "step": 245
    },
    {
      "entropy": 8.05026822090149,
      "epoch": 0.024015369836695485,
      "grad_norm": 2.078125,
      "learning_rate": 0.0001245,
      "loss": 7.4298,
      "mean_token_accuracy": 0.07491427883505822,
      "num_tokens": 571972.0,
      "step": 250
    },
    {
      "entropy": 8.011105632781982,
      "epoch": 0.024495677233429394,
      "grad_norm": 0.984375,
      "learning_rate": 0.000127,
      "loss": 7.3324,
      "mean_token_accuracy": 0.07995120882987976,
      "num_tokens": 582962.0,
      "step": 255
    },
    {
      "entropy": 7.978642559051513,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001295,
      "loss": 7.29,
      "mean_token_accuracy": 0.07967569157481194,
      "num_tokens": 597193.0,
      "step": 260
    },
    {
      "entropy": 7.959802865982056,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.5625,
      "learning_rate": 0.000132,
      "loss": 7.3629,
      "mean_token_accuracy": 0.08052113763988018,
      "num_tokens": 608982.0,
      "step": 265
    },
    {
      "entropy": 7.93332347869873,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.265625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.3853,
      "mean_token_accuracy": 0.07839832231402397,
      "num_tokens": 619953.0,
      "step": 270
    },
    {
      "entropy": 8.012384223937989,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.46875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3273,
      "mean_token_accuracy": 0.08211004585027695,
      "num_tokens": 631039.0,
      "step": 275
    },
    {
      "entropy": 7.914879417419433,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.375,
      "learning_rate": 0.0001395,
      "loss": 7.3281,
      "mean_token_accuracy": 0.0799741841852665,
      "num_tokens": 642656.0,
      "step": 280
    },
    {
      "entropy": 7.884822368621826,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1901,
      "mean_token_accuracy": 0.09594471827149391,
      "num_tokens": 653748.0,
      "step": 285
    },
    {
      "entropy": 7.860710859298706,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001445,
      "loss": 7.1763,
      "mean_token_accuracy": 0.08746851906180382,
      "num_tokens": 665618.0,
      "step": 290
    },
    {
      "entropy": 7.799470329284668,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000147,
      "loss": 7.1972,
      "mean_token_accuracy": 0.08807335570454597,
      "num_tokens": 677329.0,
      "step": 295
    },
    {
      "entropy": 7.814197397232055,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001495,
      "loss": 7.1466,
      "mean_token_accuracy": 0.09343294948339462,
      "num_tokens": 688278.0,
      "step": 300
    },
    {
      "entropy": 7.766972303390503,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000152,
      "loss": 7.2009,
      "mean_token_accuracy": 0.09057235419750213,
      "num_tokens": 700739.0,
      "step": 305
    },
    {
      "entropy": 7.782851505279541,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.1242,
      "mean_token_accuracy": 0.09343240261077881,
      "num_tokens": 712527.0,
      "step": 310
    },
    {
      "entropy": 7.725392866134643,
      "epoch": 0.03025936599423631,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000157,
      "loss": 7.1564,
      "mean_token_accuracy": 0.09019288495182991,
      "num_tokens": 724514.0,
      "step": 315
    },
    {
      "entropy": 7.741910028457641,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0001595,
      "loss": 7.1047,
      "mean_token_accuracy": 0.09378963932394982,
      "num_tokens": 735679.0,
      "step": 320
    },
    {
      "entropy": 7.70381178855896,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000162,
      "loss": 7.0993,
      "mean_token_accuracy": 0.0981727458536625,
      "num_tokens": 747896.0,
      "step": 325
    },
    {
      "entropy": 7.6704418659210205,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.40625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1443,
      "mean_token_accuracy": 0.09557413160800934,
      "num_tokens": 759081.0,
      "step": 330
    },
    {
      "entropy": 7.715982484817505,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.546875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0815,
      "mean_token_accuracy": 0.09094736129045486,
      "num_tokens": 770459.0,
      "step": 335
    },
    {
      "entropy": 7.668689107894897,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.078125,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0395,
      "mean_token_accuracy": 0.09340968057513237,
      "num_tokens": 783960.0,
      "step": 340
    },
    {
      "entropy": 7.586823654174805,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.34375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.118,
      "mean_token_accuracy": 0.10006800666451454,
      "num_tokens": 795425.0,
      "step": 345
    },
    {
      "entropy": 7.6425714015960695,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.0715,
      "mean_token_accuracy": 0.09306764900684357,
      "num_tokens": 807536.0,
      "step": 350
    },
    {
      "entropy": 7.6408501148223875,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.171875,
      "learning_rate": 0.000177,
      "loss": 7.1133,
      "mean_token_accuracy": 0.09133664220571518,
      "num_tokens": 818801.0,
      "step": 355
    },
    {
      "entropy": 7.592654371261597,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001795,
      "loss": 6.997,
      "mean_token_accuracy": 0.09742675051093101,
      "num_tokens": 831497.0,
      "step": 360
    },
    {
      "entropy": 7.561050844192505,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000182,
      "loss": 6.9518,
      "mean_token_accuracy": 0.10334301143884658,
      "num_tokens": 842491.0,
      "step": 365
    },
    {
      "entropy": 7.5975159168243405,
      "epoch": 0.03554274735830932,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001845,
      "loss": 7.038,
      "mean_token_accuracy": 0.09360877349972725,
      "num_tokens": 854560.0,
      "step": 370
    },
    {
      "entropy": 7.535368490219116,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.140625,
      "learning_rate": 0.000187,
      "loss": 6.9465,
      "mean_token_accuracy": 0.1063318483531475,
      "num_tokens": 866688.0,
      "step": 375
    },
    {
      "entropy": 7.5550884246826175,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.515625,
      "learning_rate": 0.0001895,
      "loss": 7.0347,
      "mean_token_accuracy": 0.09658555164933205,
      "num_tokens": 879484.0,
      "step": 380
    },
    {
      "entropy": 7.572937870025635,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000192,
      "loss": 7.0152,
      "mean_token_accuracy": 0.10198549181222916,
      "num_tokens": 890807.0,
      "step": 385
    },
    {
      "entropy": 7.576615810394287,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0001945,
      "loss": 6.9605,
      "mean_token_accuracy": 0.10429177060723305,
      "num_tokens": 901759.0,
      "step": 390
    },
    {
      "entropy": 7.430867433547974,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.89,
      "mean_token_accuracy": 0.10844507589936256,
      "num_tokens": 912212.0,
      "step": 395
    },
    {
      "entropy": 7.53248782157898,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9851,
      "mean_token_accuracy": 0.10315561667084694,
      "num_tokens": 923947.0,
      "step": 400
    },
    {
      "entropy": 7.401272916793824,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000202,
      "loss": 6.9312,
      "mean_token_accuracy": 0.10761535987257957,
      "num_tokens": 935732.0,
      "step": 405
    },
    {
      "entropy": 7.374160099029541,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.296875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8693,
      "mean_token_accuracy": 0.10408192500472069,
      "num_tokens": 948261.0,
      "step": 410
    },
    {
      "entropy": 7.473669195175171,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.375,
      "learning_rate": 0.000207,
      "loss": 6.9775,
      "mean_token_accuracy": 0.10531274750828742,
      "num_tokens": 959574.0,
      "step": 415
    },
    {
      "entropy": 7.502210903167724,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002095,
      "loss": 7.0212,
      "mean_token_accuracy": 0.10413776263594628,
      "num_tokens": 970329.0,
      "step": 420
    },
    {
      "entropy": 7.457794141769409,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000212,
      "loss": 6.897,
      "mean_token_accuracy": 0.11288983076810837,
      "num_tokens": 982037.0,
      "step": 425
    },
    {
      "entropy": 7.320314025878906,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002145,
      "loss": 6.8488,
      "mean_token_accuracy": 0.10890574008226395,
      "num_tokens": 994612.0,
      "step": 430
    },
    {
      "entropy": 7.436053037643433,
      "epoch": 0.04178674351585014,
      "grad_norm": 0.96875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9562,
      "mean_token_accuracy": 0.11320589855313301,
      "num_tokens": 1005960.0,
      "step": 435
    },
    {
      "entropy": 7.342877960205078,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002195,
      "loss": 6.8312,
      "mean_token_accuracy": 0.10930031016469002,
      "num_tokens": 1017618.0,
      "step": 440
    },
    {
      "entropy": 7.3924178123474125,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.78125,
      "learning_rate": 0.000222,
      "loss": 6.8722,
      "mean_token_accuracy": 0.10684446468949318,
      "num_tokens": 1029307.0,
      "step": 445
    },
    {
      "entropy": 7.34854736328125,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002245,
      "loss": 6.8798,
      "mean_token_accuracy": 0.10711006596684455,
      "num_tokens": 1042027.0,
      "step": 450
    },
    {
      "entropy": 7.3988017559051515,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8815,
      "mean_token_accuracy": 0.10970013663172722,
      "num_tokens": 1053125.0,
      "step": 455
    },
    {
      "entropy": 7.270225954055786,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.8643,
      "mean_token_accuracy": 0.11234270855784416,
      "num_tokens": 1064908.0,
      "step": 460
    },
    {
      "entropy": 7.329568862915039,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7486,
      "mean_token_accuracy": 0.11601889058947563,
      "num_tokens": 1076328.0,
      "step": 465
    },
    {
      "entropy": 7.339612579345703,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.546875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9582,
      "mean_token_accuracy": 0.10401691868901253,
      "num_tokens": 1088469.0,
      "step": 470
    },
    {
      "entropy": 7.294072389602661,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.46875,
      "learning_rate": 0.000237,
      "loss": 6.764,
      "mean_token_accuracy": 0.11554281786084175,
      "num_tokens": 1099408.0,
      "step": 475
    },
    {
      "entropy": 7.170223951339722,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002395,
      "loss": 6.6937,
      "mean_token_accuracy": 0.11718685477972031,
      "num_tokens": 1111101.0,
      "step": 480
    },
    {
      "entropy": 7.192569828033447,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.4375,
      "learning_rate": 0.000242,
      "loss": 6.6999,
      "mean_token_accuracy": 0.11837229505181313,
      "num_tokens": 1122877.0,
      "step": 485
    },
    {
      "entropy": 7.259336137771607,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002445,
      "loss": 6.811,
      "mean_token_accuracy": 0.10224953144788743,
      "num_tokens": 1133956.0,
      "step": 490
    },
    {
      "entropy": 7.205342721939087,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000247,
      "loss": 6.7171,
      "mean_token_accuracy": 0.11888269484043121,
      "num_tokens": 1146101.0,
      "step": 495
    },
    {
      "entropy": 7.218681526184082,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002495,
      "loss": 6.8062,
      "mean_token_accuracy": 0.12109306752681732,
      "num_tokens": 1157432.0,
      "step": 500
    },
    {
      "entropy": 7.241644239425659,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000252,
      "loss": 6.7906,
      "mean_token_accuracy": 0.11429283022880554,
      "num_tokens": 1167601.0,
      "step": 505
    },
    {
      "entropy": 7.20673680305481,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002545,
      "loss": 6.6849,
      "mean_token_accuracy": 0.11796795651316642,
      "num_tokens": 1178818.0,
      "step": 510
    },
    {
      "entropy": 7.338191556930542,
      "epoch": 0.0494716618635927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000257,
      "loss": 6.9209,
      "mean_token_accuracy": 0.11272176653146744,
      "num_tokens": 1189977.0,
      "step": 515
    },
    {
      "entropy": 7.15383129119873,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002595,
      "loss": 6.7585,
      "mean_token_accuracy": 0.11906541213393211,
      "num_tokens": 1201039.0,
      "step": 520
    },
    {
      "entropy": 7.194189023971558,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000262,
      "loss": 6.7788,
      "mean_token_accuracy": 0.11931035146117211,
      "num_tokens": 1212573.0,
      "step": 525
    },
    {
      "entropy": 7.264106559753418,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.25,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.7181,
      "mean_token_accuracy": 0.12056513205170631,
      "num_tokens": 1223382.0,
      "step": 530
    },
    {
      "entropy": 7.170845222473145,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.203125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7927,
      "mean_token_accuracy": 0.1244619682431221,
      "num_tokens": 1236501.0,
      "step": 535
    },
    {
      "entropy": 7.22953052520752,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7818,
      "mean_token_accuracy": 0.11887621656060218,
      "num_tokens": 1246798.0,
      "step": 540
    },
    {
      "entropy": 7.1261049747467045,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.4375,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7113,
      "mean_token_accuracy": 0.12007794082164765,
      "num_tokens": 1258182.0,
      "step": 545
    },
    {
      "entropy": 7.136359167098999,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0002745,
      "loss": 6.6647,
      "mean_token_accuracy": 0.11331228986382484,
      "num_tokens": 1270273.0,
      "step": 550
    },
    {
      "entropy": 7.1538010597229,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.5625,
      "learning_rate": 0.000277,
      "loss": 6.7157,
      "mean_token_accuracy": 0.1148271806538105,
      "num_tokens": 1281136.0,
      "step": 555
    },
    {
      "entropy": 7.182579135894775,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002795,
      "loss": 6.6991,
      "mean_token_accuracy": 0.11754038706421852,
      "num_tokens": 1294488.0,
      "step": 560
    },
    {
      "entropy": 7.084699916839599,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6054,
      "mean_token_accuracy": 0.12230072245001793,
      "num_tokens": 1304113.0,
      "step": 565
    },
    {
      "entropy": 7.104630041122436,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002845,
      "loss": 6.6713,
      "mean_token_accuracy": 0.12584275975823403,
      "num_tokens": 1315417.0,
      "step": 570
    },
    {
      "entropy": 7.049445104598999,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.359375,
      "learning_rate": 0.000287,
      "loss": 6.6306,
      "mean_token_accuracy": 0.12282092794775963,
      "num_tokens": 1328084.0,
      "step": 575
    },
    {
      "entropy": 7.128454923629761,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002895,
      "loss": 6.7747,
      "mean_token_accuracy": 0.12278011664748192,
      "num_tokens": 1338696.0,
      "step": 580
    },
    {
      "entropy": 7.090286588668823,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.078125,
      "learning_rate": 0.000292,
      "loss": 6.7179,
      "mean_token_accuracy": 0.11662147119641304,
      "num_tokens": 1350240.0,
      "step": 585
    },
    {
      "entropy": 7.117737483978272,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002945,
      "loss": 6.6639,
      "mean_token_accuracy": 0.11824186667799949,
      "num_tokens": 1361720.0,
      "step": 590
    },
    {
      "entropy": 6.979755496978759,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000297,
      "loss": 6.6611,
      "mean_token_accuracy": 0.12714812606573106,
      "num_tokens": 1373286.0,
      "step": 595
    },
    {
      "entropy": 7.1012420654296875,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002995,
      "loss": 6.6116,
      "mean_token_accuracy": 0.12019829899072647,
      "num_tokens": 1384274.0,
      "step": 600
    },
    {
      "entropy": 7.009186458587647,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.484375,
      "learning_rate": 0.000302,
      "loss": 6.6625,
      "mean_token_accuracy": 0.12345866709947587,
      "num_tokens": 1395355.0,
      "step": 605
    },
    {
      "entropy": 6.984350728988647,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003045,
      "loss": 6.5861,
      "mean_token_accuracy": 0.1286102719604969,
      "num_tokens": 1406664.0,
      "step": 610
    },
    {
      "entropy": 7.138849353790283,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000307,
      "loss": 6.7404,
      "mean_token_accuracy": 0.11916537806391717,
      "num_tokens": 1418450.0,
      "step": 615
    },
    {
      "entropy": 7.0278784275054935,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003095,
      "loss": 6.6374,
      "mean_token_accuracy": 0.1211201086640358,
      "num_tokens": 1430048.0,
      "step": 620
    },
    {
      "entropy": 7.098714780807495,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.484375,
      "learning_rate": 0.000312,
      "loss": 6.6739,
      "mean_token_accuracy": 0.12407828345894814,
      "num_tokens": 1441820.0,
      "step": 625
    },
    {
      "entropy": 6.935336065292359,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003145,
      "loss": 6.5317,
      "mean_token_accuracy": 0.12386323809623719,
      "num_tokens": 1453209.0,
      "step": 630
    },
    {
      "entropy": 6.9461079120635985,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.046875,
      "learning_rate": 0.000317,
      "loss": 6.546,
      "mean_token_accuracy": 0.122261643409729,
      "num_tokens": 1465423.0,
      "step": 635
    },
    {
      "entropy": 6.992834091186523,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003195,
      "loss": 6.5324,
      "mean_token_accuracy": 0.12468917295336723,
      "num_tokens": 1476575.0,
      "step": 640
    },
    {
      "entropy": 6.9027880191802975,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.5,
      "learning_rate": 0.000322,
      "loss": 6.495,
      "mean_token_accuracy": 0.12275839298963546,
      "num_tokens": 1486932.0,
      "step": 645
    },
    {
      "entropy": 6.965601539611816,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.532,
      "mean_token_accuracy": 0.12491905614733696,
      "num_tokens": 1498494.0,
      "step": 650
    },
    {
      "entropy": 6.929765415191651,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5786,
      "mean_token_accuracy": 0.1209442898631096,
      "num_tokens": 1509937.0,
      "step": 655
    },
    {
      "entropy": 7.01481876373291,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.328125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6063,
      "mean_token_accuracy": 0.1320799134671688,
      "num_tokens": 1519823.0,
      "step": 660
    },
    {
      "entropy": 7.075559997558594,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.6301,
      "mean_token_accuracy": 0.12214003503322601,
      "num_tokens": 1529943.0,
      "step": 665
    },
    {
      "entropy": 6.827705144882202,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4589,
      "mean_token_accuracy": 0.12923690006136895,
      "num_tokens": 1540618.0,
      "step": 670
    },
    {
      "entropy": 6.919480466842652,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000337,
      "loss": 6.5779,
      "mean_token_accuracy": 0.12186570391058922,
      "num_tokens": 1553208.0,
      "step": 675
    },
    {
      "entropy": 6.962458515167237,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003395,
      "loss": 6.5472,
      "mean_token_accuracy": 0.13120265454053878,
      "num_tokens": 1563975.0,
      "step": 680
    },
    {
      "entropy": 7.013445615768433,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000342,
      "loss": 6.5969,
      "mean_token_accuracy": 0.12027480229735374,
      "num_tokens": 1575998.0,
      "step": 685
    },
    {
      "entropy": 6.872586297988891,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4957,
      "mean_token_accuracy": 0.12978472113609313,
      "num_tokens": 1586041.0,
      "step": 690
    },
    {
      "entropy": 6.915009069442749,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.546875,
      "learning_rate": 0.000347,
      "loss": 6.5464,
      "mean_token_accuracy": 0.12723507061600686,
      "num_tokens": 1597531.0,
      "step": 695
    },
    {
      "entropy": 6.859202480316162,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003495,
      "loss": 6.5324,
      "mean_token_accuracy": 0.1253533720970154,
      "num_tokens": 1609255.0,
      "step": 700
    },
    {
      "entropy": 6.813373184204101,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.34375,
      "learning_rate": 0.000352,
      "loss": 6.5279,
      "mean_token_accuracy": 0.1321180894970894,
      "num_tokens": 1621098.0,
      "step": 705
    },
    {
      "entropy": 6.89396710395813,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003545,
      "loss": 6.4187,
      "mean_token_accuracy": 0.14093777388334275,
      "num_tokens": 1631941.0,
      "step": 710
    },
    {
      "entropy": 6.811427450180053,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357,
      "loss": 6.4234,
      "mean_token_accuracy": 0.13498233929276465,
      "num_tokens": 1643117.0,
      "step": 715
    },
    {
      "entropy": 6.800242614746094,
      "epoch": 0.069164265129683,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003595,
      "loss": 6.4627,
      "mean_token_accuracy": 0.13018925338983536,
      "num_tokens": 1653595.0,
      "step": 720
    },
    {
      "entropy": 6.83650918006897,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000362,
      "loss": 6.5025,
      "mean_token_accuracy": 0.13135557770729064,
      "num_tokens": 1664495.0,
      "step": 725
    },
    {
      "entropy": 6.802758979797363,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003645,
      "loss": 6.4803,
      "mean_token_accuracy": 0.1322146400809288,
      "num_tokens": 1674923.0,
      "step": 730
    },
    {
      "entropy": 6.83389892578125,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.265625,
      "learning_rate": 0.000367,
      "loss": 6.3745,
      "mean_token_accuracy": 0.1347797356545925,
      "num_tokens": 1685904.0,
      "step": 735
    },
    {
      "entropy": 6.834182548522949,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003695,
      "loss": 6.5992,
      "mean_token_accuracy": 0.12240826413035392,
      "num_tokens": 1699133.0,
      "step": 740
    },
    {
      "entropy": 6.7865643978118895,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.046875,
      "learning_rate": 0.000372,
      "loss": 6.408,
      "mean_token_accuracy": 0.13333356380462646,
      "num_tokens": 1711559.0,
      "step": 745
    },
    {
      "entropy": 6.8253981590271,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003745,
      "loss": 6.3276,
      "mean_token_accuracy": 0.1389731504023075,
      "num_tokens": 1722526.0,
      "step": 750
    },
    {
      "entropy": 6.7026113986969,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.328125,
      "learning_rate": 0.000377,
      "loss": 6.3874,
      "mean_token_accuracy": 0.13594599291682244,
      "num_tokens": 1734261.0,
      "step": 755
    },
    {
      "entropy": 6.764468193054199,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.375,
      "learning_rate": 0.0003795,
      "loss": 6.3728,
      "mean_token_accuracy": 0.14337323382496833,
      "num_tokens": 1745151.0,
      "step": 760
    },
    {
      "entropy": 6.874772214889527,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000382,
      "loss": 6.5351,
      "mean_token_accuracy": 0.132026257365942,
      "num_tokens": 1755463.0,
      "step": 765
    },
    {
      "entropy": 6.830476856231689,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003845,
      "loss": 6.4887,
      "mean_token_accuracy": 0.13159892857074737,
      "num_tokens": 1767717.0,
      "step": 770
    },
    {
      "entropy": 6.660166883468628,
      "epoch": 0.074447646493756,
      "grad_norm": 1.421875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.321,
      "mean_token_accuracy": 0.1407831773161888,
      "num_tokens": 1779115.0,
      "step": 775
    },
    {
      "entropy": 6.616621780395508,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.2814,
      "mean_token_accuracy": 0.13830310627818107,
      "num_tokens": 1789644.0,
      "step": 780
    },
    {
      "entropy": 6.793073272705078,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.4934,
      "mean_token_accuracy": 0.13178734481334686,
      "num_tokens": 1800606.0,
      "step": 785
    },
    {
      "entropy": 6.812487077713013,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4962,
      "mean_token_accuracy": 0.130271115899086,
      "num_tokens": 1812168.0,
      "step": 790
    },
    {
      "entropy": 6.774988603591919,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4425,
      "mean_token_accuracy": 0.12558576837182045,
      "num_tokens": 1823830.0,
      "step": 795
    },
    {
      "entropy": 6.7873927593231205,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003995,
      "loss": 6.4269,
      "mean_token_accuracy": 0.13118278980255127,
      "num_tokens": 1835611.0,
      "step": 800
    },
    {
      "entropy": 6.705799150466919,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.328125,
      "learning_rate": 0.000402,
      "loss": 6.4715,
      "mean_token_accuracy": 0.12893390357494355,
      "num_tokens": 1847036.0,
      "step": 805
    },
    {
      "entropy": 6.743785047531128,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004045,
      "loss": 6.416,
      "mean_token_accuracy": 0.1317111983895302,
      "num_tokens": 1857476.0,
      "step": 810
    },
    {
      "entropy": 6.710385322570801,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4346,
      "mean_token_accuracy": 0.13412694856524468,
      "num_tokens": 1869073.0,
      "step": 815
    },
    {
      "entropy": 6.6248006343841555,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004095,
      "loss": 6.3166,
      "mean_token_accuracy": 0.1385165847837925,
      "num_tokens": 1880439.0,
      "step": 820
    },
    {
      "entropy": 6.801689147949219,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.296875,
      "learning_rate": 0.000412,
      "loss": 6.499,
      "mean_token_accuracy": 0.128297408670187,
      "num_tokens": 1891600.0,
      "step": 825
    },
    {
      "entropy": 6.755054044723511,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004145,
      "loss": 6.4488,
      "mean_token_accuracy": 0.1346999153494835,
      "num_tokens": 1903126.0,
      "step": 830
    },
    {
      "entropy": 6.787124109268189,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000417,
      "loss": 6.417,
      "mean_token_accuracy": 0.1378346562385559,
      "num_tokens": 1913913.0,
      "step": 835
    },
    {
      "entropy": 6.613849306106568,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004195,
      "loss": 6.3337,
      "mean_token_accuracy": 0.14059634506702423,
      "num_tokens": 1924961.0,
      "step": 840
    },
    {
      "entropy": 6.824811744689941,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.421875,
      "learning_rate": 0.000422,
      "loss": 6.4201,
      "mean_token_accuracy": 0.12982920035719872,
      "num_tokens": 1936773.0,
      "step": 845
    },
    {
      "entropy": 6.610549211502075,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004245,
      "loss": 6.3343,
      "mean_token_accuracy": 0.13738347217440605,
      "num_tokens": 1948190.0,
      "step": 850
    },
    {
      "entropy": 6.6286153316497805,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000427,
      "loss": 6.3262,
      "mean_token_accuracy": 0.138392174243927,
      "num_tokens": 1960038.0,
      "step": 855
    },
    {
      "entropy": 6.762907218933106,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004295,
      "loss": 6.4021,
      "mean_token_accuracy": 0.13606049939990045,
      "num_tokens": 1970535.0,
      "step": 860
    },
    {
      "entropy": 6.655164575576782,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.59375,
      "learning_rate": 0.000432,
      "loss": 6.3764,
      "mean_token_accuracy": 0.13180122673511505,
      "num_tokens": 1981386.0,
      "step": 865
    },
    {
      "entropy": 6.615013980865479,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004345,
      "loss": 6.2949,
      "mean_token_accuracy": 0.1441991925239563,
      "num_tokens": 1993196.0,
      "step": 870
    },
    {
      "entropy": 6.65190896987915,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000437,
      "loss": 6.3309,
      "mean_token_accuracy": 0.14546860083937646,
      "num_tokens": 2004756.0,
      "step": 875
    },
    {
      "entropy": 6.595526599884034,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004395,
      "loss": 6.2971,
      "mean_token_accuracy": 0.1399446427822113,
      "num_tokens": 2016020.0,
      "step": 880
    },
    {
      "entropy": 6.638343858718872,
      "epoch": 0.08501440922190202,
      "grad_norm": 1.203125,
      "learning_rate": 0.000442,
      "loss": 6.3364,
      "mean_token_accuracy": 0.14163062497973442,
      "num_tokens": 2027747.0,
      "step": 885
    },
    {
      "entropy": 6.649935054779053,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004445,
      "loss": 6.3262,
      "mean_token_accuracy": 0.13975787982344628,
      "num_tokens": 2038841.0,
      "step": 890
    },
    {
      "entropy": 6.640672445297241,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.21875,
      "learning_rate": 0.000447,
      "loss": 6.3547,
      "mean_token_accuracy": 0.13744548857212066,
      "num_tokens": 2049905.0,
      "step": 895
    },
    {
      "entropy": 6.644440841674805,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4266,
      "mean_token_accuracy": 0.13376614451408386,
      "num_tokens": 2062492.0,
      "step": 900
    },
    {
      "entropy": 6.674169445037842,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3634,
      "mean_token_accuracy": 0.13451617136597632,
      "num_tokens": 2073840.0,
      "step": 905
    },
    {
      "entropy": 6.615630722045898,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2784,
      "mean_token_accuracy": 0.13790287226438522,
      "num_tokens": 2085720.0,
      "step": 910
    },
    {
      "entropy": 6.676623725891114,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3559,
      "mean_token_accuracy": 0.14061954542994498,
      "num_tokens": 2096649.0,
      "step": 915
    },
    {
      "entropy": 6.63926134109497,
      "epoch": 0.08837656099903939,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3586,
      "mean_token_accuracy": 0.1393554113805294,
      "num_tokens": 2109030.0,
      "step": 920
    },
    {
      "entropy": 6.6048328399658205,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.40625,
      "learning_rate": 0.000462,
      "loss": 6.2757,
      "mean_token_accuracy": 0.142194165289402,
      "num_tokens": 2121384.0,
      "step": 925
    },
    {
      "entropy": 6.563203191757202,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004645,
      "loss": 6.2225,
      "mean_token_accuracy": 0.14719335213303567,
      "num_tokens": 2131377.0,
      "step": 930
    },
    {
      "entropy": 6.577233839035034,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000467,
      "loss": 6.4123,
      "mean_token_accuracy": 0.13454320505261422,
      "num_tokens": 2142364.0,
      "step": 935
    },
    {
      "entropy": 6.619031524658203,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004695,
      "loss": 6.2849,
      "mean_token_accuracy": 0.13945741057395936,
      "num_tokens": 2153040.0,
      "step": 940
    },
    {
      "entropy": 6.542799091339111,
      "epoch": 0.09077809798270893,
      "grad_norm": 1.3125,
      "learning_rate": 0.000472,
      "loss": 6.3295,
      "mean_token_accuracy": 0.14165178686380386,
      "num_tokens": 2165571.0,
      "step": 945
    },
    {
      "entropy": 6.620811080932617,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.375,
      "learning_rate": 0.0004745,
      "loss": 6.3003,
      "mean_token_accuracy": 0.13918614387512207,
      "num_tokens": 2177241.0,
      "step": 950
    },
    {
      "entropy": 6.539478206634522,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.25,
      "learning_rate": 0.000477,
      "loss": 6.3211,
      "mean_token_accuracy": 0.135619805008173,
      "num_tokens": 2187475.0,
      "step": 955
    },
    {
      "entropy": 6.6213781356811525,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795,
      "loss": 6.3492,
      "mean_token_accuracy": 0.1400366671383381,
      "num_tokens": 2198185.0,
      "step": 960
    },
    {
      "entropy": 6.5712048530578615,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000482,
      "loss": 6.3041,
      "mean_token_accuracy": 0.1407790169119835,
      "num_tokens": 2210404.0,
      "step": 965
    },
    {
      "entropy": 6.612986421585083,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004845,
      "loss": 6.2737,
      "mean_token_accuracy": 0.13978574126958848,
      "num_tokens": 2222188.0,
      "step": 970
    },
    {
      "entropy": 6.56772894859314,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.390625,
      "learning_rate": 0.000487,
      "loss": 6.333,
      "mean_token_accuracy": 0.13907585740089418,
      "num_tokens": 2233418.0,
      "step": 975
    },
    {
      "entropy": 6.512521982192993,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004895,
      "loss": 6.2849,
      "mean_token_accuracy": 0.1478159360587597,
      "num_tokens": 2245053.0,
      "step": 980
    },
    {
      "entropy": 6.605170869827271,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492,
      "loss": 6.3158,
      "mean_token_accuracy": 0.14485108256340026,
      "num_tokens": 2256375.0,
      "step": 985
    },
    {
      "entropy": 6.405674791336059,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004945,
      "loss": 6.1641,
      "mean_token_accuracy": 0.1475372150540352,
      "num_tokens": 2267074.0,
      "step": 990
    },
    {
      "entropy": 6.520063352584839,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.234375,
      "learning_rate": 0.000497,
      "loss": 6.1987,
      "mean_token_accuracy": 0.1514969453215599,
      "num_tokens": 2277168.0,
      "step": 995
    },
    {
      "entropy": 6.435000085830689,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995,
      "loss": 6.1878,
      "mean_token_accuracy": 0.15069199353456497,
      "num_tokens": 2288178.0,
      "step": 1000
    },
    {
      "entropy": 6.503354835510254,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.1929,
      "mean_token_accuracy": 0.14548032358288765,
      "num_tokens": 2299765.0,
      "step": 1005
    },
    {
      "entropy": 6.603171157836914,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.3207,
      "mean_token_accuracy": 0.1360747143626213,
      "num_tokens": 2312047.0,
      "step": 1010
    },
    {
      "entropy": 6.503301382064819,
      "epoch": 0.09750240153698367,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.2995,
      "mean_token_accuracy": 0.1352042004466057,
      "num_tokens": 2324118.0,
      "step": 1015
    },
    {
      "entropy": 6.585948276519775,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.2985,
      "mean_token_accuracy": 0.14105767011642456,
      "num_tokens": 2335171.0,
      "step": 1020
    },
    {
      "entropy": 6.513099575042725,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.1997,
      "mean_token_accuracy": 0.14366835802793504,
      "num_tokens": 2346338.0,
      "step": 1025
    },
    {
      "entropy": 6.442260217666626,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.2248,
      "mean_token_accuracy": 0.15347937047481536,
      "num_tokens": 2357185.0,
      "step": 1030
    },
    {
      "entropy": 6.52047381401062,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.1793,
      "mean_token_accuracy": 0.15213955715298652,
      "num_tokens": 2368494.0,
      "step": 1035
    },
    {
      "entropy": 6.535188245773315,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.3719,
      "mean_token_accuracy": 0.13562044501304626,
      "num_tokens": 2380800.0,
      "step": 1040
    },
    {
      "entropy": 6.575578689575195,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.3159,
      "mean_token_accuracy": 0.14009395092725754,
      "num_tokens": 2391753.0,
      "step": 1045
    },
    {
      "entropy": 6.376717662811279,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.1859,
      "mean_token_accuracy": 0.14269104301929475,
      "num_tokens": 2403324.0,
      "step": 1050
    },
    {
      "entropy": 6.512421941757202,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.2805,
      "mean_token_accuracy": 0.1382668823003769,
      "num_tokens": 2414873.0,
      "step": 1055
    },
    {
      "entropy": 6.387165594100952,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.0851,
      "mean_token_accuracy": 0.15157609730958937,
      "num_tokens": 2425308.0,
      "step": 1060
    },
    {
      "entropy": 6.382681798934937,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.1202,
      "mean_token_accuracy": 0.15165834426879882,
      "num_tokens": 2436835.0,
      "step": 1065
    },
    {
      "entropy": 6.456397867202758,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.21875,
      "learning_rate": 0.00049999950258691,
      "loss": 6.2256,
      "mean_token_accuracy": 0.14640265703201294,
      "num_tokens": 2446798.0,
      "step": 1070
    },
    {
      "entropy": 6.332650756835937,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999994278861731,
      "loss": 6.1221,
      "mean_token_accuracy": 0.14805126786231995,
      "num_tokens": 2457308.0,
      "step": 1075
    },
    {
      "entropy": 6.4620646953582765,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.1985,
      "mean_token_accuracy": 0.13535723462700844,
      "num_tokens": 2468917.0,
      "step": 1080
    },
    {
      "entropy": 6.465546655654907,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.2148,
      "mean_token_accuracy": 0.1429847039282322,
      "num_tokens": 2481363.0,
      "step": 1085
    },
    {
      "entropy": 6.456140518188477,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.2043,
      "mean_token_accuracy": 0.14445993900299073,
      "num_tokens": 2493082.0,
      "step": 1090
    },
    {
      "entropy": 6.3651855945587155,
      "epoch": 0.10518731988472622,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.1666,
      "mean_token_accuracy": 0.14558956474065782,
      "num_tokens": 2503849.0,
      "step": 1095
    },
    {
      "entropy": 6.470709943771363,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.2132,
      "mean_token_accuracy": 0.1486193262040615,
      "num_tokens": 2514528.0,
      "step": 1100
    },
    {
      "entropy": 6.319680213928223,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004999988699816299,
      "loss": 6.1932,
      "mean_token_accuracy": 0.1492076650261879,
      "num_tokens": 2524971.0,
      "step": 1105
    },
    {
      "entropy": 6.440478706359864,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999987587142058,
      "loss": 6.0986,
      "mean_token_accuracy": 0.14526561200618743,
      "num_tokens": 2535674.0,
      "step": 1110
    },
    {
      "entropy": 6.500247812271118,
      "epoch": 0.10710854947166186,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.3347,
      "mean_token_accuracy": 0.13726388663053513,
      "num_tokens": 2547108.0,
      "step": 1115
    },
    {
      "entropy": 6.384337902069092,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.1486,
      "mean_token_accuracy": 0.14365847408771515,
      "num_tokens": 2559474.0,
      "step": 1120
    },
    {
      "entropy": 6.2849139213562015,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999983935691265,
      "loss": 5.9875,
      "mean_token_accuracy": 0.16158309876918792,
      "num_tokens": 2571264.0,
      "step": 1125
    },
    {
      "entropy": 6.373571968078613,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.125,
      "learning_rate": 0.000499998261406507,
      "loss": 6.1607,
      "mean_token_accuracy": 0.14907533824443817,
      "num_tokens": 2583731.0,
      "step": 1130
    },
    {
      "entropy": 6.570598459243774,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.209,
      "mean_token_accuracy": 0.14517664313316345,
      "num_tokens": 2595497.0,
      "step": 1135
    },
    {
      "entropy": 6.279880571365356,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.1441,
      "mean_token_accuracy": 0.14655442014336587,
      "num_tokens": 2607358.0,
      "step": 1140
    },
    {
      "entropy": 6.362181329727173,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.087,
      "mean_token_accuracy": 0.15378076806664467,
      "num_tokens": 2618936.0,
      "step": 1145
    },
    {
      "entropy": 6.368446207046508,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.2243,
      "mean_token_accuracy": 0.14323752224445344,
      "num_tokens": 2631840.0,
      "step": 1150
    },
    {
      "entropy": 6.41338152885437,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.25,
      "learning_rate": 0.000499997522236602,
      "loss": 6.1738,
      "mean_token_accuracy": 0.1520377591252327,
      "num_tokens": 2642412.0,
      "step": 1155
    },
    {
      "entropy": 6.376598930358886,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.1544,
      "mean_token_accuracy": 0.14735933318734168,
      "num_tokens": 2653890.0,
      "step": 1160
    },
    {
      "entropy": 6.446410703659057,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.2058,
      "mean_token_accuracy": 0.15076753124594688,
      "num_tokens": 2664888.0,
      "step": 1165
    },
    {
      "entropy": 6.336441469192505,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999970160493391,
      "loss": 6.1032,
      "mean_token_accuracy": 0.14862595051527022,
      "num_tokens": 2675550.0,
      "step": 1170
    },
    {
      "entropy": 6.2902812480926515,
      "epoch": 0.11287223823246878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999968368727209,
      "loss": 6.1139,
      "mean_token_accuracy": 0.15255775302648544,
      "num_tokens": 2688022.0,
      "step": 1175
    },
    {
      "entropy": 6.3844489574432375,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999966524723442,
      "loss": 6.116,
      "mean_token_accuracy": 0.1474123939871788,
      "num_tokens": 2698737.0,
      "step": 1180
    },
    {
      "entropy": 6.314852046966553,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.0924,
      "mean_token_accuracy": 0.15541501641273497,
      "num_tokens": 2709844.0,
      "step": 1185
    },
    {
      "entropy": 6.342870473861694,
      "epoch": 0.1143131604226705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.1408,
      "mean_token_accuracy": 0.1548864386975765,
      "num_tokens": 2720273.0,
      "step": 1190
    },
    {
      "entropy": 6.401251649856567,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499996067928707,
      "loss": 6.1552,
      "mean_token_accuracy": 0.15096053779125213,
      "num_tokens": 2731354.0,
      "step": 1195
    },
    {
      "entropy": 6.333895015716553,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.1386,
      "mean_token_accuracy": 0.15193476974964143,
      "num_tokens": 2742966.0,
      "step": 1200
    },
    {
      "entropy": 6.265989780426025,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.0592,
      "mean_token_accuracy": 0.15965109765529634,
      "num_tokens": 2755010.0,
      "step": 1205
    },
    {
      "entropy": 6.39248833656311,
      "epoch": 0.11623439000960614,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.0964,
      "mean_token_accuracy": 0.14827171862125396,
      "num_tokens": 2766176.0,
      "step": 1210
    },
    {
      "entropy": 6.379584693908692,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.2296,
      "mean_token_accuracy": 0.14675042033195496,
      "num_tokens": 2777861.0,
      "step": 1215
    },
    {
      "entropy": 6.266990661621094,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499994989214566,
      "loss": 6.0566,
      "mean_token_accuracy": 0.14940764307975768,
      "num_tokens": 2788725.0,
      "step": 1220
    },
    {
      "entropy": 6.2430566310882565,
      "epoch": 0.11767531219980788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.0743,
      "mean_token_accuracy": 0.15087774246931077,
      "num_tokens": 2801613.0,
      "step": 1225
    },
    {
      "entropy": 6.386273336410523,
      "epoch": 0.11815561959654179,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.1488,
      "mean_token_accuracy": 0.15173738598823547,
      "num_tokens": 2812474.0,
      "step": 1230
    },
    {
      "entropy": 6.247502326965332,
      "epoch": 0.1186359269932757,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.0647,
      "mean_token_accuracy": 0.15294726118445395,
      "num_tokens": 2823178.0,
      "step": 1235
    },
    {
      "entropy": 6.219358348846436,
      "epoch": 0.11911623439000961,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499994032216345,
      "loss": 6.0531,
      "mean_token_accuracy": 0.1527070477604866,
      "num_tokens": 2836486.0,
      "step": 1240
    },
    {
      "entropy": 6.325845193862915,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999937799075422,
      "loss": 6.1189,
      "mean_token_accuracy": 0.15587549209594725,
      "num_tokens": 2847902.0,
      "step": 1245
    },
    {
      "entropy": 6.120101833343506,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499993522375052,
      "loss": 5.9993,
      "mean_token_accuracy": 0.15293321758508682,
      "num_tokens": 2859991.0,
      "step": 1250
    },
    {
      "entropy": 6.44623122215271,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999932596188802,
      "loss": 6.1907,
      "mean_token_accuracy": 0.1484307639300823,
      "num_tokens": 2870269.0,
      "step": 1255
    },
    {
      "entropy": 6.26564474105835,
      "epoch": 0.12103746397694524,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999929916390331,
      "loss": 6.0708,
      "mean_token_accuracy": 0.1487228438258171,
      "num_tokens": 2882191.0,
      "step": 1260
    },
    {
      "entropy": 6.222062683105468,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999927184355169,
      "loss": 6.0777,
      "mean_token_accuracy": 0.14720139130949975,
      "num_tokens": 2892775.0,
      "step": 1265
    },
    {
      "entropy": 6.304709196090698,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.0619,
      "mean_token_accuracy": 0.15707618668675422,
      "num_tokens": 2904750.0,
      "step": 1270
    },
    {
      "entropy": 6.313355350494385,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.119,
      "mean_token_accuracy": 0.14971737340092658,
      "num_tokens": 2916150.0,
      "step": 1275
    },
    {
      "entropy": 6.3394777297973635,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.1047,
      "mean_token_accuracy": 0.14871277660131454,
      "num_tokens": 2928452.0,
      "step": 1280
    },
    {
      "entropy": 6.252659940719605,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999915733848886,
      "loss": 6.0908,
      "mean_token_accuracy": 0.14653694331645967,
      "num_tokens": 2940577.0,
      "step": 1285
    },
    {
      "entropy": 6.2738316535949705,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499991274063124,
      "loss": 6.0563,
      "mean_token_accuracy": 0.15000832378864287,
      "num_tokens": 2952302.0,
      "step": 1290
    },
    {
      "entropy": 6.277376508712768,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999909695177301,
      "loss": 6.1199,
      "mean_token_accuracy": 0.1537736400961876,
      "num_tokens": 2964611.0,
      "step": 1295
    },
    {
      "entropy": 6.316516351699829,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499990659748714,
      "loss": 6.1008,
      "mean_token_accuracy": 0.14772574454545975,
      "num_tokens": 2975668.0,
      "step": 1300
    },
    {
      "entropy": 6.379649114608765,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999903447560828,
      "loss": 6.163,
      "mean_token_accuracy": 0.15174921303987504,
      "num_tokens": 2987303.0,
      "step": 1305
    },
    {
      "entropy": 6.3311192989349365,
      "epoch": 0.12584053794428435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999900245398439,
      "loss": 6.06,
      "mean_token_accuracy": 0.155066779255867,
      "num_tokens": 3000400.0,
      "step": 1310
    },
    {
      "entropy": 6.25769944190979,
      "epoch": 0.12632084534101826,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.9963,
      "mean_token_accuracy": 0.15082165524363517,
      "num_tokens": 3012336.0,
      "step": 1315
    },
    {
      "entropy": 6.3025330066680905,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999893684365729,
      "loss": 6.0618,
      "mean_token_accuracy": 0.1516493648290634,
      "num_tokens": 3023004.0,
      "step": 1320
    },
    {
      "entropy": 6.239853048324585,
      "epoch": 0.12728146013448607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999890325495559,
      "loss": 6.1332,
      "mean_token_accuracy": 0.1481592871248722,
      "num_tokens": 3035147.0,
      "step": 1325
    },
    {
      "entropy": 6.269144773483276,
      "epoch": 0.12776176753121998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.964,
      "mean_token_accuracy": 0.15615130662918092,
      "num_tokens": 3045611.0,
      "step": 1330
    },
    {
      "entropy": 6.211176919937134,
      "epoch": 0.1282420749279539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999883451047981,
      "loss": 5.9794,
      "mean_token_accuracy": 0.15196620672941208,
      "num_tokens": 3056420.0,
      "step": 1335
    },
    {
      "entropy": 6.195685625076294,
      "epoch": 0.1287223823246878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.963,
      "mean_token_accuracy": 0.15559092015028,
      "num_tokens": 3068770.0,
      "step": 1340
    },
    {
      "entropy": 6.254584407806396,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999876367657954,
      "loss": 6.0985,
      "mean_token_accuracy": 0.1450921505689621,
      "num_tokens": 3080806.0,
      "step": 1345
    },
    {
      "entropy": 6.348188734054565,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999872747609725,
      "loss": 6.1052,
      "mean_token_accuracy": 0.14815450757741927,
      "num_tokens": 3091769.0,
      "step": 1350
    },
    {
      "entropy": 6.241403007507325,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999869075326132,
      "loss": 6.0277,
      "mean_token_accuracy": 0.15045573711395263,
      "num_tokens": 3103121.0,
      "step": 1355
    },
    {
      "entropy": 6.2066021919250485,
      "epoch": 0.13064361191162344,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499986535080726,
      "loss": 6.0157,
      "mean_token_accuracy": 0.15554290562868117,
      "num_tokens": 3115606.0,
      "step": 1360
    },
    {
      "entropy": 6.200456380844116,
      "epoch": 0.13112391930835735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999861574053196,
      "loss": 5.9027,
      "mean_token_accuracy": 0.16081953197717666,
      "num_tokens": 3127961.0,
      "step": 1365
    },
    {
      "entropy": 6.141843366622925,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999857745064027,
      "loss": 5.9219,
      "mean_token_accuracy": 0.16132236272096634,
      "num_tokens": 3138316.0,
      "step": 1370
    },
    {
      "entropy": 6.1357402324676515,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499985386383984,
      "loss": 5.9155,
      "mean_token_accuracy": 0.15786689072847365,
      "num_tokens": 3150818.0,
      "step": 1375
    },
    {
      "entropy": 6.233044052124024,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999849930380729,
      "loss": 6.0552,
      "mean_token_accuracy": 0.1527769923210144,
      "num_tokens": 3162066.0,
      "step": 1380
    },
    {
      "entropy": 6.123853874206543,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999845944686781,
      "loss": 6.0345,
      "mean_token_accuracy": 0.1494033455848694,
      "num_tokens": 3172209.0,
      "step": 1385
    },
    {
      "entropy": 6.176547431945801,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.8576,
      "mean_token_accuracy": 0.16600174009799956,
      "num_tokens": 3183248.0,
      "step": 1390
    },
    {
      "entropy": 6.142237281799316,
      "epoch": 0.1340057636887608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.9616,
      "mean_token_accuracy": 0.15896387100219728,
      "num_tokens": 3194748.0,
      "step": 1395
    },
    {
      "entropy": 6.1211505889892575,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.9232,
      "mean_token_accuracy": 0.16622720211744307,
      "num_tokens": 3205669.0,
      "step": 1400
    },
    {
      "entropy": 6.197529077529907,
      "epoch": 0.13496637848222862,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999829479564518,
      "loss": 6.0093,
      "mean_token_accuracy": 0.15054902583360671,
      "num_tokens": 3216035.0,
      "step": 1405
    },
    {
      "entropy": 6.228098821640015,
      "epoch": 0.13544668587896252,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499982523269781,
      "loss": 6.0043,
      "mean_token_accuracy": 0.15448517501354217,
      "num_tokens": 3227192.0,
      "step": 1410
    },
    {
      "entropy": 6.172352313995361,
      "epoch": 0.13592699327569643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999820933596842,
      "loss": 6.0283,
      "mean_token_accuracy": 0.1561071217060089,
      "num_tokens": 3240237.0,
      "step": 1415
    },
    {
      "entropy": 6.193251514434815,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499981658226171,
      "loss": 5.9033,
      "mean_token_accuracy": 0.16419047713279725,
      "num_tokens": 3251963.0,
      "step": 1420
    },
    {
      "entropy": 6.135825729370117,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499981217869252,
      "loss": 6.0291,
      "mean_token_accuracy": 0.16153036952018737,
      "num_tokens": 3263101.0,
      "step": 1425
    },
    {
      "entropy": 6.17513861656189,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9057,
      "mean_token_accuracy": 0.16388555020093917,
      "num_tokens": 3275100.0,
      "step": 1430
    },
    {
      "entropy": 6.167034149169922,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999803214852367,
      "loss": 6.0065,
      "mean_token_accuracy": 0.1535780042409897,
      "num_tokens": 3287025.0,
      "step": 1435
    },
    {
      "entropy": 6.230725336074829,
      "epoch": 0.138328530259366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999798654581613,
      "loss": 6.0077,
      "mean_token_accuracy": 0.15366143360733986,
      "num_tokens": 3299867.0,
      "step": 1440
    },
    {
      "entropy": 6.1359796047210695,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.9324,
      "mean_token_accuracy": 0.15622105449438095,
      "num_tokens": 3311183.0,
      "step": 1445
    },
    {
      "entropy": 6.170093536376953,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999789377339279,
      "loss": 6.0028,
      "mean_token_accuracy": 0.15496657863259317,
      "num_tokens": 3322247.0,
      "step": 1450
    },
    {
      "entropy": 6.195834493637085,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.9332,
      "mean_token_accuracy": 0.1554928407073021,
      "num_tokens": 3333369.0,
      "step": 1455
    },
    {
      "entropy": 6.079760551452637,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999779891163231,
      "loss": 5.9575,
      "mean_token_accuracy": 0.15849529206752777,
      "num_tokens": 3345876.0,
      "step": 1460
    },
    {
      "entropy": 6.154807806015015,
      "epoch": 0.14073006724303555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999775069725339,
      "loss": 5.8498,
      "mean_token_accuracy": 0.1605023294687271,
      "num_tokens": 3357323.0,
      "step": 1465
    },
    {
      "entropy": 6.096273374557495,
      "epoch": 0.14121037463976946,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499977019605435,
      "loss": 5.9311,
      "mean_token_accuracy": 0.15962084680795668,
      "num_tokens": 3367689.0,
      "step": 1470
    },
    {
      "entropy": 6.006128931045533,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.8952,
      "mean_token_accuracy": 0.16288901418447493,
      "num_tokens": 3379472.0,
      "step": 1475
    },
    {
      "entropy": 6.239159154891968,
      "epoch": 0.14217098943323728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.9533,
      "mean_token_accuracy": 0.15554243996739386,
      "num_tokens": 3390929.0,
      "step": 1480
    },
    {
      "entropy": 6.209478282928467,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.9401,
      "mean_token_accuracy": 0.16148580312728883,
      "num_tokens": 3401242.0,
      "step": 1485
    },
    {
      "entropy": 6.060628652572632,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.9322,
      "mean_token_accuracy": 0.1567712590098381,
      "num_tokens": 3411169.0,
      "step": 1490
    },
    {
      "entropy": 6.067643499374389,
      "epoch": 0.143611911623439,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.7653,
      "mean_token_accuracy": 0.1653980001807213,
      "num_tokens": 3423265.0,
      "step": 1495
    },
    {
      "entropy": 6.061836051940918,
      "epoch": 0.1440922190201729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.8723,
      "mean_token_accuracy": 0.1658122293651104,
      "num_tokens": 3434793.0,
      "step": 1500
    },
    {
      "entropy": 6.013027095794678,
      "epoch": 0.14457252641690682,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499973461784038,
      "loss": 5.8078,
      "mean_token_accuracy": 0.17317451983690263,
      "num_tokens": 3445732.0,
      "step": 1505
    },
    {
      "entropy": 6.127257823944092,
      "epoch": 0.14505283381364073,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9813,
      "mean_token_accuracy": 0.15522808730602264,
      "num_tokens": 3457090.0,
      "step": 1510
    },
    {
      "entropy": 6.194346618652344,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499972398254517,
      "loss": 5.9726,
      "mean_token_accuracy": 0.15402631014585494,
      "num_tokens": 3468087.0,
      "step": 1515
    },
    {
      "entropy": 6.157139253616333,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499971858654964,
      "loss": 5.9108,
      "mean_token_accuracy": 0.15916238874197006,
      "num_tokens": 3478820.0,
      "step": 1520
    },
    {
      "entropy": 6.0432126998901365,
      "epoch": 0.14649375600384246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.9387,
      "mean_token_accuracy": 0.15639192685484887,
      "num_tokens": 3489878.0,
      "step": 1525
    },
    {
      "entropy": 6.179038763046265,
      "epoch": 0.14697406340057637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.9227,
      "mean_token_accuracy": 0.15898001939058304,
      "num_tokens": 3500944.0,
      "step": 1530
    },
    {
      "entropy": 6.006661891937256,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.8949,
      "mean_token_accuracy": 0.166755573451519,
      "num_tokens": 3511383.0,
      "step": 1535
    },
    {
      "entropy": 6.193773031234741,
      "epoch": 0.14793467819404418,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999696480250929,
      "loss": 6.0009,
      "mean_token_accuracy": 0.1563890740275383,
      "num_tokens": 3523300.0,
      "step": 1540
    },
    {
      "entropy": 6.137639045715332,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.904,
      "mean_token_accuracy": 0.15553956106305122,
      "num_tokens": 3534371.0,
      "step": 1545
    },
    {
      "entropy": 6.11188440322876,
      "epoch": 0.148895292987512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.8848,
      "mean_token_accuracy": 0.1581243097782135,
      "num_tokens": 3544847.0,
      "step": 1550
    },
    {
      "entropy": 6.084521484375,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8683,
      "mean_token_accuracy": 0.1633513331413269,
      "num_tokens": 3555859.0,
      "step": 1555
    },
    {
      "entropy": 6.081562280654907,
      "epoch": 0.14985590778097982,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8724,
      "mean_token_accuracy": 0.16277971267700195,
      "num_tokens": 3568283.0,
      "step": 1560
    },
    {
      "entropy": 6.040332508087158,
      "epoch": 0.15033621517771373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.9535,
      "mean_token_accuracy": 0.15841858685016633,
      "num_tokens": 3581442.0,
      "step": 1565
    },
    {
      "entropy": 6.214689922332764,
      "epoch": 0.15081652257444764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049996617538674,
      "loss": 5.999,
      "mean_token_accuracy": 0.15400405079126359,
      "num_tokens": 3594055.0,
      "step": 1570
    },
    {
      "entropy": 6.139510774612427,
      "epoch": 0.15129682997118155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9564,
      "mean_token_accuracy": 0.1526302471756935,
      "num_tokens": 3605952.0,
      "step": 1575
    },
    {
      "entropy": 6.104340982437134,
      "epoch": 0.15177713736791545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.9762,
      "mean_token_accuracy": 0.15697605460882186,
      "num_tokens": 3618544.0,
      "step": 1580
    },
    {
      "entropy": 6.139195299148559,
      "epoch": 0.15225744476464936,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.928,
      "mean_token_accuracy": 0.16384828090667725,
      "num_tokens": 3630445.0,
      "step": 1585
    },
    {
      "entropy": 6.140196180343628,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.9277,
      "mean_token_accuracy": 0.162806536257267,
      "num_tokens": 3642516.0,
      "step": 1590
    },
    {
      "entropy": 6.032298421859741,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.8473,
      "mean_token_accuracy": 0.1700352057814598,
      "num_tokens": 3654425.0,
      "step": 1595
    },
    {
      "entropy": 6.052908039093017,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.8659,
      "mean_token_accuracy": 0.1607646018266678,
      "num_tokens": 3666521.0,
      "step": 1600
    },
    {
      "entropy": 6.144074392318726,
      "epoch": 0.15417867435158503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.9113,
      "mean_token_accuracy": 0.15713147521018983,
      "num_tokens": 3679121.0,
      "step": 1605
    },
    {
      "entropy": 6.08037257194519,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.9215,
      "mean_token_accuracy": 0.15551023483276366,
      "num_tokens": 3691095.0,
      "step": 1610
    },
    {
      "entropy": 6.048700380325317,
      "epoch": 0.15513928914505284,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8863,
      "mean_token_accuracy": 0.16148294806480407,
      "num_tokens": 3703426.0,
      "step": 1615
    },
    {
      "entropy": 6.160890388488769,
      "epoch": 0.15561959654178675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9506,
      "mean_token_accuracy": 0.16333511769771575,
      "num_tokens": 3715429.0,
      "step": 1620
    },
    {
      "entropy": 6.134197616577149,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.9302,
      "mean_token_accuracy": 0.1580351047217846,
      "num_tokens": 3726327.0,
      "step": 1625
    },
    {
      "entropy": 6.134684467315674,
      "epoch": 0.15658021133525457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9004,
      "mean_token_accuracy": 0.15458485931158067,
      "num_tokens": 3736818.0,
      "step": 1630
    },
    {
      "entropy": 6.102252435684204,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.9023,
      "mean_token_accuracy": 0.16163242906332015,
      "num_tokens": 3747776.0,
      "step": 1635
    },
    {
      "entropy": 5.938621568679809,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.7522,
      "mean_token_accuracy": 0.16386649161577224,
      "num_tokens": 3758990.0,
      "step": 1640
    },
    {
      "entropy": 6.073081064224243,
      "epoch": 0.1580211335254563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.8335,
      "mean_token_accuracy": 0.1618332728743553,
      "num_tokens": 3769686.0,
      "step": 1645
    },
    {
      "entropy": 6.034470748901367,
      "epoch": 0.1585014409221902,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8842,
      "mean_token_accuracy": 0.16474503725767137,
      "num_tokens": 3781815.0,
      "step": 1650
    },
    {
      "entropy": 6.107155132293701,
      "epoch": 0.15898174831892412,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9224,
      "mean_token_accuracy": 0.16111959964036943,
      "num_tokens": 3793392.0,
      "step": 1655
    },
    {
      "entropy": 6.020937395095825,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.8622,
      "mean_token_accuracy": 0.15857846289873123,
      "num_tokens": 3804974.0,
      "step": 1660
    },
    {
      "entropy": 5.980496025085449,
      "epoch": 0.15994236311239193,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049995393823994,
      "loss": 5.7399,
      "mean_token_accuracy": 0.1712213933467865,
      "num_tokens": 3817166.0,
      "step": 1665
    },
    {
      "entropy": 5.985702800750732,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.8584,
      "mean_token_accuracy": 0.16114549189805985,
      "num_tokens": 3828151.0,
      "step": 1670
    },
    {
      "entropy": 6.079834032058716,
      "epoch": 0.16090297790585975,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8566,
      "mean_token_accuracy": 0.16157617270946503,
      "num_tokens": 3839439.0,
      "step": 1675
    },
    {
      "entropy": 6.109479236602783,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.9183,
      "mean_token_accuracy": 0.15559976994991304,
      "num_tokens": 3851694.0,
      "step": 1680
    },
    {
      "entropy": 5.999596118927002,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049995112174787,
      "loss": 5.8943,
      "mean_token_accuracy": 0.15952107906341553,
      "num_tokens": 3863593.0,
      "step": 1685
    },
    {
      "entropy": 5.972264003753662,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16952102333307267,
      "num_tokens": 3874588.0,
      "step": 1690
    },
    {
      "entropy": 6.094362878799439,
      "epoch": 0.1628242074927954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.8981,
      "mean_token_accuracy": 0.15856973826885223,
      "num_tokens": 3884662.0,
      "step": 1695
    },
    {
      "entropy": 6.140773725509644,
      "epoch": 0.1633045148895293,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.9891,
      "mean_token_accuracy": 0.15623455494642258,
      "num_tokens": 3896569.0,
      "step": 1700
    },
    {
      "entropy": 6.140805101394653,
      "epoch": 0.1637848222862632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.8462,
      "mean_token_accuracy": 0.16176871806383133,
      "num_tokens": 3907691.0,
      "step": 1705
    },
    {
      "entropy": 6.034537076950073,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.8498,
      "mean_token_accuracy": 0.16770250350236893,
      "num_tokens": 3918794.0,
      "step": 1710
    },
    {
      "entropy": 6.016002893447876,
      "epoch": 0.16474543707973102,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8424,
      "mean_token_accuracy": 0.17401330322027206,
      "num_tokens": 3929773.0,
      "step": 1715
    },
    {
      "entropy": 6.039723968505859,
      "epoch": 0.16522574447646493,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.8626,
      "mean_token_accuracy": 0.16028911992907524,
      "num_tokens": 3940264.0,
      "step": 1720
    },
    {
      "entropy": 5.988501930236817,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7585,
      "mean_token_accuracy": 0.16902887523174287,
      "num_tokens": 3951500.0,
      "step": 1725
    },
    {
      "entropy": 6.034324693679809,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.8427,
      "mean_token_accuracy": 0.16535028964281082,
      "num_tokens": 3963580.0,
      "step": 1730
    },
    {
      "entropy": 5.986796712875366,
      "epoch": 0.16666666666666666,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.7757,
      "mean_token_accuracy": 0.17422386556863784,
      "num_tokens": 3975994.0,
      "step": 1735
    },
    {
      "entropy": 5.9908356189727785,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499942945525439,
      "loss": 5.7922,
      "mean_token_accuracy": 0.1610643297433853,
      "num_tokens": 3987897.0,
      "step": 1740
    },
    {
      "entropy": 6.048891305923462,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.9403,
      "mean_token_accuracy": 0.1578907661139965,
      "num_tokens": 3999829.0,
      "step": 1745
    },
    {
      "entropy": 6.001609849929809,
      "epoch": 0.16810758885686838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.7895,
      "mean_token_accuracy": 0.16723742038011552,
      "num_tokens": 4010882.0,
      "step": 1750
    },
    {
      "entropy": 5.978082466125488,
      "epoch": 0.1685878962536023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.7717,
      "mean_token_accuracy": 0.16948328018188477,
      "num_tokens": 4021423.0,
      "step": 1755
    },
    {
      "entropy": 6.093415307998657,
      "epoch": 0.1690682036503362,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.9183,
      "mean_token_accuracy": 0.15228435546159744,
      "num_tokens": 4033590.0,
      "step": 1760
    },
    {
      "entropy": 5.927422332763672,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00049993902015901,
      "loss": 5.7775,
      "mean_token_accuracy": 0.16497890055179595,
      "num_tokens": 4043978.0,
      "step": 1765
    },
    {
      "entropy": 6.080128145217896,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.9011,
      "mean_token_accuracy": 0.16162478625774385,
      "num_tokens": 4054513.0,
      "step": 1770
    },
    {
      "entropy": 6.0692665576934814,
      "epoch": 0.17050912584053796,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.8467,
      "mean_token_accuracy": 0.16818321645259857,
      "num_tokens": 4066019.0,
      "step": 1775
    },
    {
      "entropy": 5.963143253326416,
      "epoch": 0.17098943323727187,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.8214,
      "mean_token_accuracy": 0.16922373026609422,
      "num_tokens": 4077688.0,
      "step": 1780
    },
    {
      "entropy": 6.082214260101319,
      "epoch": 0.17146974063400577,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.8939,
      "mean_token_accuracy": 0.1704793967306614,
      "num_tokens": 4089803.0,
      "step": 1785
    },
    {
      "entropy": 6.019527435302734,
      "epoch": 0.17195004803073968,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.8323,
      "mean_token_accuracy": 0.16811123341321946,
      "num_tokens": 4101969.0,
      "step": 1790
    },
    {
      "entropy": 5.995560216903686,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8181,
      "mean_token_accuracy": 0.16215983778238297,
      "num_tokens": 4113567.0,
      "step": 1795
    },
    {
      "entropy": 5.987828254699707,
      "epoch": 0.1729106628242075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7861,
      "mean_token_accuracy": 0.1671761840581894,
      "num_tokens": 4125191.0,
      "step": 1800
    },
    {
      "entropy": 6.054584407806397,
      "epoch": 0.1733909702209414,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.8297,
      "mean_token_accuracy": 0.167048180103302,
      "num_tokens": 4135050.0,
      "step": 1805
    },
    {
      "entropy": 5.962006425857544,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.8864,
      "mean_token_accuracy": 0.16525756269693376,
      "num_tokens": 4146874.0,
      "step": 1810
    },
    {
      "entropy": 6.027948713302612,
      "epoch": 0.17435158501440923,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.8119,
      "mean_token_accuracy": 0.16480804234743118,
      "num_tokens": 4158118.0,
      "step": 1815
    },
    {
      "entropy": 5.980266284942627,
      "epoch": 0.17483189241114314,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7898,
      "mean_token_accuracy": 0.17785038501024247,
      "num_tokens": 4169035.0,
      "step": 1820
    },
    {
      "entropy": 5.911714839935303,
      "epoch": 0.17531219980787704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499929066597658,
      "loss": 5.767,
      "mean_token_accuracy": 0.17150768488645554,
      "num_tokens": 4180314.0,
      "step": 1825
    },
    {
      "entropy": 6.06360354423523,
      "epoch": 0.17579250720461095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1665028989315033,
      "num_tokens": 4192238.0,
      "step": 1830
    },
    {
      "entropy": 5.918634033203125,
      "epoch": 0.17627281460134486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.7864,
      "mean_token_accuracy": 0.16353099048137665,
      "num_tokens": 4203346.0,
      "step": 1835
    },
    {
      "entropy": 5.99358344078064,
      "epoch": 0.17675312199807877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.8248,
      "mean_token_accuracy": 0.16922380030155182,
      "num_tokens": 4213763.0,
      "step": 1840
    },
    {
      "entropy": 5.949887943267822,
      "epoch": 0.17723342939481268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.7987,
      "mean_token_accuracy": 0.1714692607522011,
      "num_tokens": 4225553.0,
      "step": 1845
    },
    {
      "entropy": 6.06374306678772,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499924697335629,
      "loss": 5.7171,
      "mean_token_accuracy": 0.17483123242855073,
      "num_tokens": 4236058.0,
      "step": 1850
    },
    {
      "entropy": 5.883856916427613,
      "epoch": 0.1781940441882805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.7983,
      "mean_token_accuracy": 0.17132625132799148,
      "num_tokens": 4248299.0,
      "step": 1855
    },
    {
      "entropy": 5.987727499008178,
      "epoch": 0.1786743515850144,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.7295,
      "mean_token_accuracy": 0.16837194561958313,
      "num_tokens": 4259704.0,
      "step": 1860
    },
    {
      "entropy": 5.883204078674316,
      "epoch": 0.17915465898174832,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499922013113246,
      "loss": 5.6159,
      "mean_token_accuracy": 0.183360692858696,
      "num_tokens": 4270480.0,
      "step": 1865
    },
    {
      "entropy": 5.932878494262695,
      "epoch": 0.17963496637848222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.8661,
      "mean_token_accuracy": 0.16950789242982864,
      "num_tokens": 4282104.0,
      "step": 1870
    },
    {
      "entropy": 6.039040184020996,
      "epoch": 0.18011527377521613,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.8454,
      "mean_token_accuracy": 0.16162874400615693,
      "num_tokens": 4294251.0,
      "step": 1875
    },
    {
      "entropy": 5.932565784454345,
      "epoch": 0.18059558117195004,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.7256,
      "mean_token_accuracy": 0.16642534136772155,
      "num_tokens": 4305569.0,
      "step": 1880
    },
    {
      "entropy": 6.010633945465088,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.856,
      "mean_token_accuracy": 0.16204970329999924,
      "num_tokens": 4317845.0,
      "step": 1885
    },
    {
      "entropy": 5.937364482879639,
      "epoch": 0.18155619596541786,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.7241,
      "mean_token_accuracy": 0.16895480155944825,
      "num_tokens": 4329987.0,
      "step": 1890
    },
    {
      "entropy": 5.916909265518188,
      "epoch": 0.18203650336215177,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7336,
      "mean_token_accuracy": 0.16701279431581498,
      "num_tokens": 4341628.0,
      "step": 1895
    },
    {
      "entropy": 5.97577314376831,
      "epoch": 0.18251681075888568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.8141,
      "mean_token_accuracy": 0.16247764229774475,
      "num_tokens": 4352379.0,
      "step": 1900
    },
    {
      "entropy": 5.888210439682007,
      "epoch": 0.1829971181556196,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7384,
      "mean_token_accuracy": 0.17271625846624375,
      "num_tokens": 4364800.0,
      "step": 1905
    },
    {
      "entropy": 5.934367513656616,
      "epoch": 0.1834774255523535,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.7544,
      "mean_token_accuracy": 0.16907702684402465,
      "num_tokens": 4376048.0,
      "step": 1910
    },
    {
      "entropy": 5.916653251647949,
      "epoch": 0.1839577329490874,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.7474,
      "mean_token_accuracy": 0.16713621020317077,
      "num_tokens": 4388072.0,
      "step": 1915
    },
    {
      "entropy": 6.049605655670166,
      "epoch": 0.1844380403458213,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.8408,
      "mean_token_accuracy": 0.16282255798578263,
      "num_tokens": 4399843.0,
      "step": 1920
    },
    {
      "entropy": 5.854936361312866,
      "epoch": 0.18491834774255522,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.692,
      "mean_token_accuracy": 0.17128440588712693,
      "num_tokens": 4411373.0,
      "step": 1925
    },
    {
      "entropy": 5.888366031646728,
      "epoch": 0.18539865513928913,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499909838400118,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1735598623752594,
      "num_tokens": 4421857.0,
      "step": 1930
    },
    {
      "entropy": 5.947378730773925,
      "epoch": 0.18587896253602307,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.7819,
      "mean_token_accuracy": 0.1680931806564331,
      "num_tokens": 4432728.0,
      "step": 1935
    },
    {
      "entropy": 5.886118745803833,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.714,
      "mean_token_accuracy": 0.1756661355495453,
      "num_tokens": 4444683.0,
      "step": 1940
    },
    {
      "entropy": 5.994152688980103,
      "epoch": 0.18683957732949089,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.8381,
      "mean_token_accuracy": 0.16196750849485397,
      "num_tokens": 4456961.0,
      "step": 1945
    },
    {
      "entropy": 5.912719821929931,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.6738,
      "mean_token_accuracy": 0.17798083573579787,
      "num_tokens": 4468424.0,
      "step": 1950
    },
    {
      "entropy": 5.929690361022949,
      "epoch": 0.1878001921229587,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.8701,
      "mean_token_accuracy": 0.15679999366402625,
      "num_tokens": 4479813.0,
      "step": 1955
    },
    {
      "entropy": 6.098432826995849,
      "epoch": 0.1882804995196926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.8826,
      "mean_token_accuracy": 0.16513065099716187,
      "num_tokens": 4491172.0,
      "step": 1960
    },
    {
      "entropy": 5.891456174850464,
      "epoch": 0.18876080691642652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.7655,
      "mean_token_accuracy": 0.17174434214830397,
      "num_tokens": 4503063.0,
      "step": 1965
    },
    {
      "entropy": 6.019789409637451,
      "epoch": 0.18924111431316043,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.8443,
      "mean_token_accuracy": 0.15457844734191895,
      "num_tokens": 4514188.0,
      "step": 1970
    },
    {
      "entropy": 5.98090615272522,
      "epoch": 0.18972142170989434,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7495,
      "mean_token_accuracy": 0.17291804850101472,
      "num_tokens": 4525293.0,
      "step": 1975
    },
    {
      "entropy": 5.9353631973266605,
      "epoch": 0.19020172910662825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.8315,
      "mean_token_accuracy": 0.16792338639497756,
      "num_tokens": 4536589.0,
      "step": 1980
    },
    {
      "entropy": 5.87034125328064,
      "epoch": 0.19068203650336216,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.5689,
      "mean_token_accuracy": 0.19097184985876084,
      "num_tokens": 4547594.0,
      "step": 1985
    },
    {
      "entropy": 5.888665723800659,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.7814,
      "mean_token_accuracy": 0.173173625767231,
      "num_tokens": 4559850.0,
      "step": 1990
    },
    {
      "entropy": 5.935269927978515,
      "epoch": 0.19164265129682997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.7343,
      "mean_token_accuracy": 0.16940659731626512,
      "num_tokens": 4570727.0,
      "step": 1995
    },
    {
      "entropy": 5.766289901733399,
      "epoch": 0.19212295869356388,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.6639,
      "mean_token_accuracy": 0.17158041000366211,
      "num_tokens": 4582248.0,
      "step": 2000
    },
    {
      "entropy": 5.963186931610108,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.7005,
      "mean_token_accuracy": 0.16607439517974854,
      "num_tokens": 4592604.0,
      "step": 2005
    },
    {
      "entropy": 5.888526773452758,
      "epoch": 0.1930835734870317,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7435,
      "mean_token_accuracy": 0.1702524498105049,
      "num_tokens": 4604398.0,
      "step": 2010
    },
    {
      "entropy": 5.91964282989502,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.8039,
      "mean_token_accuracy": 0.16930960565805436,
      "num_tokens": 4616434.0,
      "step": 2015
    },
    {
      "entropy": 5.981127023696899,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.7597,
      "mean_token_accuracy": 0.17877342998981477,
      "num_tokens": 4627577.0,
      "step": 2020
    },
    {
      "entropy": 5.960260248184204,
      "epoch": 0.19452449567723343,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.8227,
      "mean_token_accuracy": 0.16877581179141998,
      "num_tokens": 4639698.0,
      "step": 2025
    },
    {
      "entropy": 6.018915462493896,
      "epoch": 0.19500480307396734,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.8322,
      "mean_token_accuracy": 0.16535419821739197,
      "num_tokens": 4650643.0,
      "step": 2030
    },
    {
      "entropy": 5.94498839378357,
      "epoch": 0.19548511047070125,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7511,
      "mean_token_accuracy": 0.17331523150205613,
      "num_tokens": 4662897.0,
      "step": 2035
    },
    {
      "entropy": 5.939677953720093,
      "epoch": 0.19596541786743515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.8446,
      "mean_token_accuracy": 0.1710888773202896,
      "num_tokens": 4673701.0,
      "step": 2040
    },
    {
      "entropy": 6.051514387130737,
      "epoch": 0.19644572526416906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.7976,
      "mean_token_accuracy": 0.17028483599424363,
      "num_tokens": 4685873.0,
      "step": 2045
    },
    {
      "entropy": 5.855360126495361,
      "epoch": 0.19692603266090297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.6799,
      "mean_token_accuracy": 0.171894171833992,
      "num_tokens": 4697179.0,
      "step": 2050
    },
    {
      "entropy": 5.93500165939331,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17378782033920287,
      "num_tokens": 4707752.0,
      "step": 2055
    },
    {
      "entropy": 5.785550880432129,
      "epoch": 0.1978866474543708,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.5828,
      "mean_token_accuracy": 0.17932410091161727,
      "num_tokens": 4718413.0,
      "step": 2060
    },
    {
      "entropy": 5.901092576980591,
      "epoch": 0.1983669548511047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7755,
      "mean_token_accuracy": 0.167689011991024,
      "num_tokens": 4730033.0,
      "step": 2065
    },
    {
      "entropy": 5.849337959289551,
      "epoch": 0.1988472622478386,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499880618731189,
      "loss": 5.632,
      "mean_token_accuracy": 0.17377811521291733,
      "num_tokens": 4742084.0,
      "step": 2070
    },
    {
      "entropy": 5.942147588729858,
      "epoch": 0.19932756964457252,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.8568,
      "mean_token_accuracy": 0.16341482400894164,
      "num_tokens": 4753885.0,
      "step": 2075
    },
    {
      "entropy": 5.894834518432617,
      "epoch": 0.19980787704130643,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.7098,
      "mean_token_accuracy": 0.1734362706542015,
      "num_tokens": 4765686.0,
      "step": 2080
    },
    {
      "entropy": 5.933993625640869,
      "epoch": 0.20028818443804033,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7384,
      "mean_token_accuracy": 0.16806679219007492,
      "num_tokens": 4777157.0,
      "step": 2085
    },
    {
      "entropy": 5.881355714797974,
      "epoch": 0.20076849183477424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6823,
      "mean_token_accuracy": 0.1787089616060257,
      "num_tokens": 4788583.0,
      "step": 2090
    },
    {
      "entropy": 5.965854835510254,
      "epoch": 0.20124879923150815,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1650993473827839,
      "num_tokens": 4800749.0,
      "step": 2095
    },
    {
      "entropy": 5.834138011932373,
      "epoch": 0.2017291066282421,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.7169,
      "mean_token_accuracy": 0.17363064736127853,
      "num_tokens": 4812488.0,
      "step": 2100
    },
    {
      "entropy": 5.952000665664673,
      "epoch": 0.202209414024976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6942,
      "mean_token_accuracy": 0.16981008946895598,
      "num_tokens": 4823495.0,
      "step": 2105
    },
    {
      "entropy": 5.821967363357544,
      "epoch": 0.2026897214217099,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.7474,
      "mean_token_accuracy": 0.17119863927364348,
      "num_tokens": 4834450.0,
      "step": 2110
    },
    {
      "entropy": 5.896064186096192,
      "epoch": 0.20317002881844382,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.653,
      "mean_token_accuracy": 0.18205936402082443,
      "num_tokens": 4846826.0,
      "step": 2115
    },
    {
      "entropy": 5.961601400375367,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.8431,
      "mean_token_accuracy": 0.16039666682481765,
      "num_tokens": 4859668.0,
      "step": 2120
    },
    {
      "entropy": 5.8840587615966795,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.7134,
      "mean_token_accuracy": 0.17416077703237534,
      "num_tokens": 4871727.0,
      "step": 2125
    },
    {
      "entropy": 5.968330240249633,
      "epoch": 0.20461095100864554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.78,
      "mean_token_accuracy": 0.1742082878947258,
      "num_tokens": 4882191.0,
      "step": 2130
    },
    {
      "entropy": 5.897537279129028,
      "epoch": 0.20509125840537945,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.7018,
      "mean_token_accuracy": 0.1743673115968704,
      "num_tokens": 4892416.0,
      "step": 2135
    },
    {
      "entropy": 5.846418809890747,
      "epoch": 0.20557156580211336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1706405311822891,
      "num_tokens": 4903572.0,
      "step": 2140
    },
    {
      "entropy": 5.968627214431763,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.7787,
      "mean_token_accuracy": 0.16432214826345443,
      "num_tokens": 4915417.0,
      "step": 2145
    },
    {
      "entropy": 5.875264406204224,
      "epoch": 0.20653218059558118,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.6596,
      "mean_token_accuracy": 0.17557333707809447,
      "num_tokens": 4926943.0,
      "step": 2150
    },
    {
      "entropy": 5.82595853805542,
      "epoch": 0.2070124879923151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6216,
      "mean_token_accuracy": 0.1786582961678505,
      "num_tokens": 4937724.0,
      "step": 2155
    },
    {
      "entropy": 5.967111301422119,
      "epoch": 0.207492795389049,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.7773,
      "mean_token_accuracy": 0.16382010281085968,
      "num_tokens": 4949970.0,
      "step": 2160
    },
    {
      "entropy": 5.899398708343506,
      "epoch": 0.2079731027857829,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.7756,
      "mean_token_accuracy": 0.1687404468655586,
      "num_tokens": 4961389.0,
      "step": 2165
    },
    {
      "entropy": 5.956924343109131,
      "epoch": 0.2084534101825168,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.8212,
      "mean_token_accuracy": 0.16519832164049147,
      "num_tokens": 4973888.0,
      "step": 2170
    },
    {
      "entropy": 5.873337078094482,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.7713,
      "mean_token_accuracy": 0.1622347429394722,
      "num_tokens": 4985423.0,
      "step": 2175
    },
    {
      "entropy": 5.875820159912109,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.7277,
      "mean_token_accuracy": 0.16388818621635437,
      "num_tokens": 4997711.0,
      "step": 2180
    },
    {
      "entropy": 5.944964838027954,
      "epoch": 0.20989433237271854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.7587,
      "mean_token_accuracy": 0.17094310671091079,
      "num_tokens": 5009633.0,
      "step": 2185
    },
    {
      "entropy": 5.89435224533081,
      "epoch": 0.21037463976945245,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.6898,
      "mean_token_accuracy": 0.17721298336982727,
      "num_tokens": 5021523.0,
      "step": 2190
    },
    {
      "entropy": 5.954970121383667,
      "epoch": 0.21085494716618636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7963,
      "mean_token_accuracy": 0.1734875962138176,
      "num_tokens": 5033513.0,
      "step": 2195
    },
    {
      "entropy": 6.01756477355957,
      "epoch": 0.21133525456292027,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7897,
      "mean_token_accuracy": 0.16785951554775239,
      "num_tokens": 5047055.0,
      "step": 2200
    },
    {
      "entropy": 5.842679691314697,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.6612,
      "mean_token_accuracy": 0.17763629406690598,
      "num_tokens": 5057613.0,
      "step": 2205
    },
    {
      "entropy": 5.941239261627198,
      "epoch": 0.21229586935638808,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.7791,
      "mean_token_accuracy": 0.170269151031971,
      "num_tokens": 5068886.0,
      "step": 2210
    },
    {
      "entropy": 5.8616218090057375,
      "epoch": 0.212776176753122,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.7561,
      "mean_token_accuracy": 0.16782673597335815,
      "num_tokens": 5079978.0,
      "step": 2215
    },
    {
      "entropy": 5.881478404998779,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1720557138323784,
      "num_tokens": 5091021.0,
      "step": 2220
    },
    {
      "entropy": 5.84704737663269,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.6118,
      "mean_token_accuracy": 0.18117671012878417,
      "num_tokens": 5101483.0,
      "step": 2225
    },
    {
      "entropy": 5.8953385829925535,
      "epoch": 0.21421709894332372,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.7518,
      "mean_token_accuracy": 0.16402507573366165,
      "num_tokens": 5113116.0,
      "step": 2230
    },
    {
      "entropy": 5.954845809936524,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.6943,
      "mean_token_accuracy": 0.17064779102802277,
      "num_tokens": 5124824.0,
      "step": 2235
    },
    {
      "entropy": 5.77202959060669,
      "epoch": 0.21517771373679154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.6247,
      "mean_token_accuracy": 0.1712438568472862,
      "num_tokens": 5137567.0,
      "step": 2240
    },
    {
      "entropy": 5.833844184875488,
      "epoch": 0.21565802113352545,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6923,
      "mean_token_accuracy": 0.17277579456567765,
      "num_tokens": 5149016.0,
      "step": 2245
    },
    {
      "entropy": 5.808039903640747,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.5749,
      "mean_token_accuracy": 0.1784022033214569,
      "num_tokens": 5160356.0,
      "step": 2250
    },
    {
      "entropy": 5.881575202941894,
      "epoch": 0.21661863592699326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7716,
      "mean_token_accuracy": 0.16418982148170472,
      "num_tokens": 5172100.0,
      "step": 2255
    },
    {
      "entropy": 5.9312032699584964,
      "epoch": 0.21709894332372717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.7397,
      "mean_token_accuracy": 0.16633592247962953,
      "num_tokens": 5183984.0,
      "step": 2260
    },
    {
      "entropy": 5.862771272659302,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1685562714934349,
      "num_tokens": 5195853.0,
      "step": 2265
    },
    {
      "entropy": 5.7872209548950195,
      "epoch": 0.21805955811719502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.564,
      "mean_token_accuracy": 0.17655707597732545,
      "num_tokens": 5206995.0,
      "step": 2270
    },
    {
      "entropy": 5.662486600875854,
      "epoch": 0.21853986551392893,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.601,
      "mean_token_accuracy": 0.17818109691143036,
      "num_tokens": 5219291.0,
      "step": 2275
    },
    {
      "entropy": 5.872530078887939,
      "epoch": 0.21902017291066284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.6762,
      "mean_token_accuracy": 0.17558231055736542,
      "num_tokens": 5230856.0,
      "step": 2280
    },
    {
      "entropy": 5.743492794036865,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.5404,
      "mean_token_accuracy": 0.18575960099697114,
      "num_tokens": 5242871.0,
      "step": 2285
    },
    {
      "entropy": 5.753171300888061,
      "epoch": 0.21998078770413065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.645,
      "mean_token_accuracy": 0.17795871943235397,
      "num_tokens": 5253835.0,
      "step": 2290
    },
    {
      "entropy": 5.834173107147217,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.6285,
      "mean_token_accuracy": 0.17905001789331437,
      "num_tokens": 5266195.0,
      "step": 2295
    },
    {
      "entropy": 5.795336961746216,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.6088,
      "mean_token_accuracy": 0.18106043636798858,
      "num_tokens": 5277499.0,
      "step": 2300
    },
    {
      "entropy": 5.926272821426392,
      "epoch": 0.22142170989433238,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499822369630386,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16407669484615325,
      "num_tokens": 5288622.0,
      "step": 2305
    },
    {
      "entropy": 5.853435325622558,
      "epoch": 0.2219020172910663,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.7626,
      "mean_token_accuracy": 0.1640668675303459,
      "num_tokens": 5299664.0,
      "step": 2310
    },
    {
      "entropy": 5.848914766311646,
      "epoch": 0.2223823246878002,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6597,
      "mean_token_accuracy": 0.17648861855268477,
      "num_tokens": 5311627.0,
      "step": 2315
    },
    {
      "entropy": 5.746734476089477,
      "epoch": 0.2228626320845341,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5634,
      "mean_token_accuracy": 0.18326989710330963,
      "num_tokens": 5323000.0,
      "step": 2320
    },
    {
      "entropy": 5.806483554840088,
      "epoch": 0.22334293948126802,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.601,
      "mean_token_accuracy": 0.17859914302825927,
      "num_tokens": 5333811.0,
      "step": 2325
    },
    {
      "entropy": 5.761362361907959,
      "epoch": 0.22382324687800192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499815494300587,
      "loss": 5.6037,
      "mean_token_accuracy": 0.17953155487775802,
      "num_tokens": 5344762.0,
      "step": 2330
    },
    {
      "entropy": 5.87151312828064,
      "epoch": 0.22430355427473583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17849526405334473,
      "num_tokens": 5356112.0,
      "step": 2335
    },
    {
      "entropy": 5.816241693496704,
      "epoch": 0.22478386167146974,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.7338,
      "mean_token_accuracy": 0.17089829891920089,
      "num_tokens": 5367339.0,
      "step": 2340
    },
    {
      "entropy": 5.803972244262695,
      "epoch": 0.22526416906820365,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.6492,
      "mean_token_accuracy": 0.1800484448671341,
      "num_tokens": 5378627.0,
      "step": 2345
    },
    {
      "entropy": 5.852473402023316,
      "epoch": 0.22574447646493756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.6988,
      "mean_token_accuracy": 0.170039651542902,
      "num_tokens": 5390209.0,
      "step": 2350
    },
    {
      "entropy": 5.897865295410156,
      "epoch": 0.22622478386167147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6853,
      "mean_token_accuracy": 0.17408402562141417,
      "num_tokens": 5401578.0,
      "step": 2355
    },
    {
      "entropy": 5.756231594085693,
      "epoch": 0.22670509125840538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.604,
      "mean_token_accuracy": 0.17672595530748367,
      "num_tokens": 5413831.0,
      "step": 2360
    },
    {
      "entropy": 5.865112972259522,
      "epoch": 0.2271853986551393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6552,
      "mean_token_accuracy": 0.17331201881170272,
      "num_tokens": 5425430.0,
      "step": 2365
    },
    {
      "entropy": 5.671746349334716,
      "epoch": 0.2276657060518732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.5077,
      "mean_token_accuracy": 0.18612135499715804,
      "num_tokens": 5435353.0,
      "step": 2370
    },
    {
      "entropy": 5.847486209869385,
      "epoch": 0.2281460134486071,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7082,
      "mean_token_accuracy": 0.17188700735569,
      "num_tokens": 5446925.0,
      "step": 2375
    },
    {
      "entropy": 5.763326358795166,
      "epoch": 0.228626320845341,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.5835,
      "mean_token_accuracy": 0.18231132328510286,
      "num_tokens": 5456613.0,
      "step": 2380
    },
    {
      "entropy": 5.823438310623169,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.6236,
      "mean_token_accuracy": 0.17775536924600602,
      "num_tokens": 5467790.0,
      "step": 2385
    },
    {
      "entropy": 5.813823795318603,
      "epoch": 0.22958693563880883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6084,
      "mean_token_accuracy": 0.17633272111415862,
      "num_tokens": 5479166.0,
      "step": 2390
    },
    {
      "entropy": 5.752776956558227,
      "epoch": 0.23006724303554274,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.592,
      "mean_token_accuracy": 0.18635759204626084,
      "num_tokens": 5490186.0,
      "step": 2395
    },
    {
      "entropy": 5.799315357208252,
      "epoch": 0.23054755043227665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.581,
      "mean_token_accuracy": 0.173208612203598,
      "num_tokens": 5500416.0,
      "step": 2400
    },
    {
      "entropy": 5.78081750869751,
      "epoch": 0.23102785782901056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.6455,
      "mean_token_accuracy": 0.17610323131084443,
      "num_tokens": 5512189.0,
      "step": 2405
    },
    {
      "entropy": 5.750943851470947,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499792616381322,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18103354275226594,
      "num_tokens": 5523631.0,
      "step": 2410
    },
    {
      "entropy": 5.769753503799438,
      "epoch": 0.23198847262247838,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5793,
      "mean_token_accuracy": 0.18183256834745407,
      "num_tokens": 5535637.0,
      "step": 2415
    },
    {
      "entropy": 5.7926108837127686,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499789662696278,
      "loss": 5.5809,
      "mean_token_accuracy": 0.18048014342784882,
      "num_tokens": 5546470.0,
      "step": 2420
    },
    {
      "entropy": 5.866393899917602,
      "epoch": 0.2329490874159462,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.7472,
      "mean_token_accuracy": 0.1731916069984436,
      "num_tokens": 5558633.0,
      "step": 2425
    },
    {
      "entropy": 5.785652208328247,
      "epoch": 0.2334293948126801,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.6529,
      "mean_token_accuracy": 0.1741258069872856,
      "num_tokens": 5570427.0,
      "step": 2430
    },
    {
      "entropy": 5.814468336105347,
      "epoch": 0.23390970220941404,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499785193026685,
      "loss": 5.6383,
      "mean_token_accuracy": 0.17221823185682297,
      "num_tokens": 5580991.0,
      "step": 2435
    },
    {
      "entropy": 5.831542491912842,
      "epoch": 0.23439000960614795,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.688,
      "mean_token_accuracy": 0.1702160969376564,
      "num_tokens": 5592777.0,
      "step": 2440
    },
    {
      "entropy": 5.743597841262817,
      "epoch": 0.23487031700288186,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.5844,
      "mean_token_accuracy": 0.18186513036489488,
      "num_tokens": 5603326.0,
      "step": 2445
    },
    {
      "entropy": 5.7442912578582765,
      "epoch": 0.23535062439961577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.5642,
      "mean_token_accuracy": 0.17944303005933762,
      "num_tokens": 5614504.0,
      "step": 2450
    },
    {
      "entropy": 5.771292018890381,
      "epoch": 0.23583093179634967,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.6286,
      "mean_token_accuracy": 0.18041289001703262,
      "num_tokens": 5625150.0,
      "step": 2455
    },
    {
      "entropy": 5.772236967086792,
      "epoch": 0.23631123919308358,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6376,
      "mean_token_accuracy": 0.17271037846803666,
      "num_tokens": 5636815.0,
      "step": 2460
    },
    {
      "entropy": 5.824061107635498,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17515492290258408,
      "num_tokens": 5648272.0,
      "step": 2465
    },
    {
      "entropy": 5.761957883834839,
      "epoch": 0.2372718539865514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.5481,
      "mean_token_accuracy": 0.17955229729413985,
      "num_tokens": 5659227.0,
      "step": 2470
    },
    {
      "entropy": 5.7706011772155765,
      "epoch": 0.2377521613832853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.668,
      "mean_token_accuracy": 0.1772313669323921,
      "num_tokens": 5670411.0,
      "step": 2475
    },
    {
      "entropy": 5.764390659332276,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1818181738257408,
      "num_tokens": 5681570.0,
      "step": 2480
    },
    {
      "entropy": 5.693533134460449,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499769954908461,
      "loss": 5.5075,
      "mean_token_accuracy": 0.18814098387956618,
      "num_tokens": 5693021.0,
      "step": 2485
    },
    {
      "entropy": 5.732751131057739,
      "epoch": 0.23919308357348704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.5841,
      "mean_token_accuracy": 0.17531975656747817,
      "num_tokens": 5704043.0,
      "step": 2490
    },
    {
      "entropy": 5.747956085205078,
      "epoch": 0.23967339097022095,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.6703,
      "mean_token_accuracy": 0.17353179901838303,
      "num_tokens": 5715735.0,
      "step": 2495
    },
    {
      "entropy": 5.906291723251343,
      "epoch": 0.24015369836695485,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1773464038968086,
      "num_tokens": 5725778.0,
      "step": 2500
    },
    {
      "entropy": 5.7474383354187015,
      "epoch": 0.24063400576368876,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6558,
      "mean_token_accuracy": 0.17596883922815323,
      "num_tokens": 5737224.0,
      "step": 2505
    },
    {
      "entropy": 5.883420705795288,
      "epoch": 0.24111431316042267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.6777,
      "mean_token_accuracy": 0.17386398762464522,
      "num_tokens": 5749226.0,
      "step": 2510
    },
    {
      "entropy": 5.84810619354248,
      "epoch": 0.24159462055715658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.6523,
      "mean_token_accuracy": 0.17698420137166976,
      "num_tokens": 5760282.0,
      "step": 2515
    },
    {
      "entropy": 5.712239980697632,
      "epoch": 0.2420749279538905,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997589777234,
      "loss": 5.5534,
      "mean_token_accuracy": 0.18517332524061203,
      "num_tokens": 5771756.0,
      "step": 2520
    },
    {
      "entropy": 5.766052198410034,
      "epoch": 0.2425552353506244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.5824,
      "mean_token_accuracy": 0.1730706885457039,
      "num_tokens": 5783237.0,
      "step": 2525
    },
    {
      "entropy": 5.8274911403656,
      "epoch": 0.2430355427473583,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.6739,
      "mean_token_accuracy": 0.17468079626560212,
      "num_tokens": 5794825.0,
      "step": 2530
    },
    {
      "entropy": 5.793378686904907,
      "epoch": 0.24351585014409222,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.5834,
      "mean_token_accuracy": 0.18117998987436296,
      "num_tokens": 5805578.0,
      "step": 2535
    },
    {
      "entropy": 5.833753156661987,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6623,
      "mean_token_accuracy": 0.16987032443284988,
      "num_tokens": 5818201.0,
      "step": 2540
    },
    {
      "entropy": 5.754938840866089,
      "epoch": 0.24447646493756003,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.5809,
      "mean_token_accuracy": 0.17766104489564896,
      "num_tokens": 5830319.0,
      "step": 2545
    },
    {
      "entropy": 5.7188770294189455,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5319,
      "mean_token_accuracy": 0.18243667036294936,
      "num_tokens": 5840638.0,
      "step": 2550
    },
    {
      "entropy": 5.779051208496094,
      "epoch": 0.24543707973102785,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.5674,
      "mean_token_accuracy": 0.17579597234725952,
      "num_tokens": 5852472.0,
      "step": 2555
    },
    {
      "entropy": 5.779905652999878,
      "epoch": 0.24591738712776176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.6071,
      "mean_token_accuracy": 0.17549769282341005,
      "num_tokens": 5863455.0,
      "step": 2560
    },
    {
      "entropy": 5.7139819145202635,
      "epoch": 0.24639769452449567,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5082,
      "mean_token_accuracy": 0.1821136549115181,
      "num_tokens": 5873141.0,
      "step": 2565
    },
    {
      "entropy": 5.714353847503662,
      "epoch": 0.24687800192122958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6137,
      "mean_token_accuracy": 0.177882120013237,
      "num_tokens": 5884363.0,
      "step": 2570
    },
    {
      "entropy": 5.8017199516296385,
      "epoch": 0.2473583093179635,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.6563,
      "mean_token_accuracy": 0.17355215102434157,
      "num_tokens": 5895856.0,
      "step": 2575
    },
    {
      "entropy": 5.7232537269592285,
      "epoch": 0.2478386167146974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.4999,
      "mean_token_accuracy": 0.18378839939832686,
      "num_tokens": 5906657.0,
      "step": 2580
    },
    {
      "entropy": 5.733435249328613,
      "epoch": 0.2483189241114313,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17644155025482178,
      "num_tokens": 5919496.0,
      "step": 2585
    },
    {
      "entropy": 5.843182802200317,
      "epoch": 0.24879923150816521,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16554286181926728,
      "num_tokens": 5932593.0,
      "step": 2590
    },
    {
      "entropy": 5.776597499847412,
      "epoch": 0.24927953890489912,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5987,
      "mean_token_accuracy": 0.1750054731965065,
      "num_tokens": 5944080.0,
      "step": 2595
    },
    {
      "entropy": 5.734813213348389,
      "epoch": 0.24975984630163303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.5622,
      "mean_token_accuracy": 0.18719092905521392,
      "num_tokens": 5955703.0,
      "step": 2600
    },
    {
      "entropy": 5.736793804168701,
      "epoch": 0.25024015369836694,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6578,
      "mean_token_accuracy": 0.1708198830485344,
      "num_tokens": 5966979.0,
      "step": 2605
    },
    {
      "entropy": 5.811160707473755,
      "epoch": 0.2507204610951009,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.5433,
      "mean_token_accuracy": 0.17897313684225083,
      "num_tokens": 5978808.0,
      "step": 2610
    },
    {
      "entropy": 5.785584354400635,
      "epoch": 0.25120076849183476,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.7051,
      "mean_token_accuracy": 0.16843380331993102,
      "num_tokens": 5992145.0,
      "step": 2615
    },
    {
      "entropy": 5.7584997653961185,
      "epoch": 0.2516810758885687,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17471926659345627,
      "num_tokens": 6003723.0,
      "step": 2620
    },
    {
      "entropy": 5.771317625045777,
      "epoch": 0.2521613832853026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6058,
      "mean_token_accuracy": 0.18242801427841188,
      "num_tokens": 6014064.0,
      "step": 2625
    },
    {
      "entropy": 5.698479509353637,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18279289156198503,
      "num_tokens": 6025455.0,
      "step": 2630
    },
    {
      "entropy": 5.782511425018311,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.6313,
      "mean_token_accuracy": 0.1700494483113289,
      "num_tokens": 6037347.0,
      "step": 2635
    },
    {
      "entropy": 5.6879799365997314,
      "epoch": 0.25360230547550433,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.5155,
      "mean_token_accuracy": 0.18077107816934584,
      "num_tokens": 6049236.0,
      "step": 2640
    },
    {
      "entropy": 5.664460229873657,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1818702682852745,
      "num_tokens": 6060982.0,
      "step": 2645
    },
    {
      "entropy": 5.795065593719483,
      "epoch": 0.25456292026897215,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.5409,
      "mean_token_accuracy": 0.18014705032110215,
      "num_tokens": 6070925.0,
      "step": 2650
    },
    {
      "entropy": 5.672208833694458,
      "epoch": 0.25504322766570603,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.505,
      "mean_token_accuracy": 0.18191429376602172,
      "num_tokens": 6081279.0,
      "step": 2655
    },
    {
      "entropy": 5.707340812683105,
      "epoch": 0.25552353506243997,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.5966,
      "mean_token_accuracy": 0.17941437065601348,
      "num_tokens": 6091797.0,
      "step": 2660
    },
    {
      "entropy": 5.816579580307007,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.6217,
      "mean_token_accuracy": 0.18266446739435196,
      "num_tokens": 6103435.0,
      "step": 2665
    },
    {
      "entropy": 5.8208917617797855,
      "epoch": 0.2564841498559078,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499709036684407,
      "loss": 5.6735,
      "mean_token_accuracy": 0.1784491568803787,
      "num_tokens": 6114531.0,
      "step": 2670
    },
    {
      "entropy": 5.659421300888061,
      "epoch": 0.25696445725264166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.5132,
      "mean_token_accuracy": 0.1838804692029953,
      "num_tokens": 6126110.0,
      "step": 2675
    },
    {
      "entropy": 5.79349536895752,
      "epoch": 0.2574447646493756,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.6132,
      "mean_token_accuracy": 0.17701641321182252,
      "num_tokens": 6137114.0,
      "step": 2680
    },
    {
      "entropy": 5.69478325843811,
      "epoch": 0.2579250720461095,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.5683,
      "mean_token_accuracy": 0.17944201529026033,
      "num_tokens": 6148696.0,
      "step": 2685
    },
    {
      "entropy": 5.735519981384277,
      "epoch": 0.2584053794428434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.5677,
      "mean_token_accuracy": 0.18062195777893067,
      "num_tokens": 6160235.0,
      "step": 2690
    },
    {
      "entropy": 5.747908687591552,
      "epoch": 0.25888568683957736,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.5407,
      "mean_token_accuracy": 0.17935636788606643,
      "num_tokens": 6172031.0,
      "step": 2695
    },
    {
      "entropy": 5.596827173233033,
      "epoch": 0.25936599423631124,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18828413337469102,
      "num_tokens": 6183547.0,
      "step": 2700
    },
    {
      "entropy": 5.798890924453735,
      "epoch": 0.2598463016330452,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.6161,
      "mean_token_accuracy": 0.17421343475580214,
      "num_tokens": 6194768.0,
      "step": 2705
    },
    {
      "entropy": 5.752058315277099,
      "epoch": 0.26032660902977905,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17473251819610597,
      "num_tokens": 6206099.0,
      "step": 2710
    },
    {
      "entropy": 5.768714952468872,
      "epoch": 0.260806916426513,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6049,
      "mean_token_accuracy": 0.18249890506267546,
      "num_tokens": 6217530.0,
      "step": 2715
    },
    {
      "entropy": 5.824987697601318,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.6404,
      "mean_token_accuracy": 0.1773010641336441,
      "num_tokens": 6228564.0,
      "step": 2720
    },
    {
      "entropy": 5.682008457183838,
      "epoch": 0.2617675312199808,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499689548463118,
      "loss": 5.5072,
      "mean_token_accuracy": 0.18143168538808824,
      "num_tokens": 6239945.0,
      "step": 2725
    },
    {
      "entropy": 5.719662475585937,
      "epoch": 0.2622478386167147,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6148,
      "mean_token_accuracy": 0.17328692674636842,
      "num_tokens": 6251829.0,
      "step": 2730
    },
    {
      "entropy": 5.861980676651001,
      "epoch": 0.2627281460134486,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.7102,
      "mean_token_accuracy": 0.1663394570350647,
      "num_tokens": 6264823.0,
      "step": 2735
    },
    {
      "entropy": 5.684702014923095,
      "epoch": 0.2632084534101825,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.4084,
      "mean_token_accuracy": 0.18651311248540878,
      "num_tokens": 6276247.0,
      "step": 2740
    },
    {
      "entropy": 5.626027154922485,
      "epoch": 0.26368876080691644,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.5308,
      "mean_token_accuracy": 0.18354606330394746,
      "num_tokens": 6287593.0,
      "step": 2745
    },
    {
      "entropy": 5.799522447586059,
      "epoch": 0.2641690682036503,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.676,
      "mean_token_accuracy": 0.17569343596696854,
      "num_tokens": 6299918.0,
      "step": 2750
    },
    {
      "entropy": 5.681373929977417,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5314,
      "mean_token_accuracy": 0.18340865522623062,
      "num_tokens": 6311147.0,
      "step": 2755
    },
    {
      "entropy": 5.615675544738769,
      "epoch": 0.26512968299711814,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.4432,
      "mean_token_accuracy": 0.18849314153194427,
      "num_tokens": 6323239.0,
      "step": 2760
    },
    {
      "entropy": 5.773361492156982,
      "epoch": 0.2656099903938521,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5348,
      "mean_token_accuracy": 0.1791366621851921,
      "num_tokens": 6334625.0,
      "step": 2765
    },
    {
      "entropy": 5.70220251083374,
      "epoch": 0.26609029779058596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499673134056204,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17963075488805771,
      "num_tokens": 6346068.0,
      "step": 2770
    },
    {
      "entropy": 5.713240051269532,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.5531,
      "mean_token_accuracy": 0.18010566383600235,
      "num_tokens": 6357097.0,
      "step": 2775
    },
    {
      "entropy": 5.7573648452758786,
      "epoch": 0.2670509125840538,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.6084,
      "mean_token_accuracy": 0.17204165309667588,
      "num_tokens": 6367481.0,
      "step": 2780
    },
    {
      "entropy": 5.685487174987793,
      "epoch": 0.2675312199807877,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.5207,
      "mean_token_accuracy": 0.18194421827793122,
      "num_tokens": 6378873.0,
      "step": 2785
    },
    {
      "entropy": 5.747344875335694,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.5625,
      "mean_token_accuracy": 0.1808866173028946,
      "num_tokens": 6390651.0,
      "step": 2790
    },
    {
      "entropy": 5.640884160995483,
      "epoch": 0.26849183477425553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.4351,
      "mean_token_accuracy": 0.18793859630823134,
      "num_tokens": 6401631.0,
      "step": 2795
    },
    {
      "entropy": 5.6249631404876705,
      "epoch": 0.2689721421709894,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.5219,
      "mean_token_accuracy": 0.186911478638649,
      "num_tokens": 6413875.0,
      "step": 2800
    },
    {
      "entropy": 5.782164573669434,
      "epoch": 0.26945244956772335,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.5488,
      "mean_token_accuracy": 0.17567767947912216,
      "num_tokens": 6425764.0,
      "step": 2805
    },
    {
      "entropy": 5.648550701141358,
      "epoch": 0.26993275696445723,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1833810329437256,
      "num_tokens": 6437911.0,
      "step": 2810
    },
    {
      "entropy": 5.6579090595245365,
      "epoch": 0.27041306436119117,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049965629713367,
      "loss": 5.4874,
      "mean_token_accuracy": 0.1852079749107361,
      "num_tokens": 6449942.0,
      "step": 2815
    },
    {
      "entropy": 5.709984302520752,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.6016,
      "mean_token_accuracy": 0.18187092244625092,
      "num_tokens": 6461729.0,
      "step": 2820
    },
    {
      "entropy": 5.747480726242065,
      "epoch": 0.271373679154659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.5839,
      "mean_token_accuracy": 0.18295794278383254,
      "num_tokens": 6472046.0,
      "step": 2825
    },
    {
      "entropy": 5.714561557769775,
      "epoch": 0.27185398655139287,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1794055625796318,
      "num_tokens": 6483308.0,
      "step": 2830
    },
    {
      "entropy": 5.5759443759918215,
      "epoch": 0.2723342939481268,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.4213,
      "mean_token_accuracy": 0.18884292393922805,
      "num_tokens": 6495093.0,
      "step": 2835
    },
    {
      "entropy": 5.70222053527832,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.5213,
      "mean_token_accuracy": 0.18563241511583328,
      "num_tokens": 6505933.0,
      "step": 2840
    },
    {
      "entropy": 5.663483715057373,
      "epoch": 0.2732949087415946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.56,
      "mean_token_accuracy": 0.17418374568223954,
      "num_tokens": 6517280.0,
      "step": 2845
    },
    {
      "entropy": 5.739472961425781,
      "epoch": 0.2737752161383285,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.62,
      "mean_token_accuracy": 0.1724501520395279,
      "num_tokens": 6528980.0,
      "step": 2850
    },
    {
      "entropy": 5.694519329071045,
      "epoch": 0.27425552353506244,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.6063,
      "mean_token_accuracy": 0.18012285381555557,
      "num_tokens": 6540670.0,
      "step": 2855
    },
    {
      "entropy": 5.6691272258758545,
      "epoch": 0.2747358309317964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.5641,
      "mean_token_accuracy": 0.18185704350471496,
      "num_tokens": 6551302.0,
      "step": 2860
    },
    {
      "entropy": 5.717091369628906,
      "epoch": 0.27521613832853026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5206,
      "mean_token_accuracy": 0.17356996536254882,
      "num_tokens": 6563177.0,
      "step": 2865
    },
    {
      "entropy": 5.848795604705811,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.6331,
      "mean_token_accuracy": 0.18121621906757354,
      "num_tokens": 6573323.0,
      "step": 2870
    },
    {
      "entropy": 5.6664739608764645,
      "epoch": 0.2761767531219981,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5584,
      "mean_token_accuracy": 0.17100237160921097,
      "num_tokens": 6585382.0,
      "step": 2875
    },
    {
      "entropy": 5.612715768814087,
      "epoch": 0.276657060518732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.4473,
      "mean_token_accuracy": 0.18974230736494063,
      "num_tokens": 6596629.0,
      "step": 2880
    },
    {
      "entropy": 5.721270322799683,
      "epoch": 0.2771373679154659,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499629266625136,
      "loss": 5.5591,
      "mean_token_accuracy": 0.17991002202033995,
      "num_tokens": 6608408.0,
      "step": 2885
    },
    {
      "entropy": 5.779607629776001,
      "epoch": 0.27761767531219983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5857,
      "mean_token_accuracy": 0.17561327517032624,
      "num_tokens": 6619375.0,
      "step": 2890
    },
    {
      "entropy": 5.758464813232422,
      "epoch": 0.2780979827089337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.6217,
      "mean_token_accuracy": 0.17862538397312164,
      "num_tokens": 6631317.0,
      "step": 2895
    },
    {
      "entropy": 5.706511449813843,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4881,
      "mean_token_accuracy": 0.18149750381708146,
      "num_tokens": 6642009.0,
      "step": 2900
    },
    {
      "entropy": 5.686898136138916,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.6358,
      "mean_token_accuracy": 0.17196063697338104,
      "num_tokens": 6654713.0,
      "step": 2905
    },
    {
      "entropy": 5.74081768989563,
      "epoch": 0.27953890489913547,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.5878,
      "mean_token_accuracy": 0.1804736316204071,
      "num_tokens": 6667157.0,
      "step": 2910
    },
    {
      "entropy": 5.703048181533814,
      "epoch": 0.28001921229586935,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5329,
      "mean_token_accuracy": 0.1874094307422638,
      "num_tokens": 6679573.0,
      "step": 2915
    },
    {
      "entropy": 5.73606538772583,
      "epoch": 0.2804995196926033,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.6154,
      "mean_token_accuracy": 0.1785501942038536,
      "num_tokens": 6691768.0,
      "step": 2920
    },
    {
      "entropy": 5.75312385559082,
      "epoch": 0.28097982708933716,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049961336167259,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1757212147116661,
      "num_tokens": 6701964.0,
      "step": 2925
    },
    {
      "entropy": 5.748111152648926,
      "epoch": 0.2814601344860711,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5561,
      "mean_token_accuracy": 0.18022931665182113,
      "num_tokens": 6713506.0,
      "step": 2930
    },
    {
      "entropy": 5.749008369445801,
      "epoch": 0.281940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5821,
      "mean_token_accuracy": 0.17544898986816407,
      "num_tokens": 6724616.0,
      "step": 2935
    },
    {
      "entropy": 5.6963791847229,
      "epoch": 0.2824207492795389,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5629,
      "mean_token_accuracy": 0.18043071925640106,
      "num_tokens": 6735054.0,
      "step": 2940
    },
    {
      "entropy": 5.741255426406861,
      "epoch": 0.2829010566762728,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1793173149228096,
      "num_tokens": 6746756.0,
      "step": 2945
    },
    {
      "entropy": 5.704806613922119,
      "epoch": 0.28338136407300674,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.4832,
      "mean_token_accuracy": 0.1823109433054924,
      "num_tokens": 6759566.0,
      "step": 2950
    },
    {
      "entropy": 5.69014163017273,
      "epoch": 0.2838616714697406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.5125,
      "mean_token_accuracy": 0.18087273538112641,
      "num_tokens": 6771375.0,
      "step": 2955
    },
    {
      "entropy": 5.748940420150757,
      "epoch": 0.28434197886647455,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.6176,
      "mean_token_accuracy": 0.1698983356356621,
      "num_tokens": 6783252.0,
      "step": 2960
    },
    {
      "entropy": 5.6928369998931885,
      "epoch": 0.28482228626320844,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.5816,
      "mean_token_accuracy": 0.1742567628622055,
      "num_tokens": 6795525.0,
      "step": 2965
    },
    {
      "entropy": 5.742653036117554,
      "epoch": 0.28530259365994237,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.5129,
      "mean_token_accuracy": 0.18347676545381547,
      "num_tokens": 6807212.0,
      "step": 2970
    },
    {
      "entropy": 5.64953875541687,
      "epoch": 0.28578290105667625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.458,
      "mean_token_accuracy": 0.18245918452739715,
      "num_tokens": 6819367.0,
      "step": 2975
    },
    {
      "entropy": 5.665671300888062,
      "epoch": 0.2862632084534102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.5454,
      "mean_token_accuracy": 0.18168749809265136,
      "num_tokens": 6830405.0,
      "step": 2980
    },
    {
      "entropy": 5.6374242305755615,
      "epoch": 0.28674351585014407,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.513,
      "mean_token_accuracy": 0.1782350167632103,
      "num_tokens": 6841479.0,
      "step": 2985
    },
    {
      "entropy": 5.671642160415649,
      "epoch": 0.287223823246878,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5612,
      "mean_token_accuracy": 0.17904469966888428,
      "num_tokens": 6851585.0,
      "step": 2990
    },
    {
      "entropy": 5.694643163681031,
      "epoch": 0.2877041306436119,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.521,
      "mean_token_accuracy": 0.18657984137535094,
      "num_tokens": 6863176.0,
      "step": 2995
    },
    {
      "entropy": 5.726582813262939,
      "epoch": 0.2881844380403458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.5769,
      "mean_token_accuracy": 0.1821230873465538,
      "num_tokens": 6874021.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.499006296238743,
      "eval_loss": 5.545695781707764,
      "eval_mean_token_accuracy": 0.1887321266854706,
      "eval_num_tokens": 6874021.0,
      "eval_runtime": 20.7487,
      "eval_samples_per_second": 1581.543,
      "eval_steps_per_second": 197.699,
      "step": 3000
    },
    {
      "entropy": 5.730997228622437,
      "epoch": 0.2886647454370797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.5433,
      "mean_token_accuracy": 0.17902631163597107,
      "num_tokens": 6884999.0,
      "step": 3005
    },
    {
      "entropy": 5.664561748504639,
      "epoch": 0.28914505283381364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.5182,
      "mean_token_accuracy": 0.1828203856945038,
      "num_tokens": 6897021.0,
      "step": 3010
    },
    {
      "entropy": 5.648147296905518,
      "epoch": 0.2896253602305475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.5212,
      "mean_token_accuracy": 0.18067535609006882,
      "num_tokens": 6908320.0,
      "step": 3015
    },
    {
      "entropy": 5.756967782974243,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.5074,
      "mean_token_accuracy": 0.1846700832247734,
      "num_tokens": 6919957.0,
      "step": 3020
    },
    {
      "entropy": 5.698904705047608,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.4824,
      "mean_token_accuracy": 0.17627989798784255,
      "num_tokens": 6930531.0,
      "step": 3025
    },
    {
      "entropy": 5.720267581939697,
      "epoch": 0.2910662824207493,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.6076,
      "mean_token_accuracy": 0.17746972441673278,
      "num_tokens": 6942556.0,
      "step": 3030
    },
    {
      "entropy": 5.70178074836731,
      "epoch": 0.2915465898174832,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5621,
      "mean_token_accuracy": 0.18203302770853041,
      "num_tokens": 6954410.0,
      "step": 3035
    },
    {
      "entropy": 5.717909717559815,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.5576,
      "mean_token_accuracy": 0.1767730161547661,
      "num_tokens": 6964970.0,
      "step": 3040
    },
    {
      "entropy": 5.675133419036865,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18839906603097917,
      "num_tokens": 6976016.0,
      "step": 3045
    },
    {
      "entropy": 5.634436845779419,
      "epoch": 0.2929875120076849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.4858,
      "mean_token_accuracy": 0.18708995282649993,
      "num_tokens": 6987199.0,
      "step": 3050
    },
    {
      "entropy": 5.65491418838501,
      "epoch": 0.29346781940441885,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.4941,
      "mean_token_accuracy": 0.19084091037511824,
      "num_tokens": 6999174.0,
      "step": 3055
    },
    {
      "entropy": 5.698542308807373,
      "epoch": 0.29394812680115273,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499557218580945,
      "loss": 5.5702,
      "mean_token_accuracy": 0.18034818172454833,
      "num_tokens": 7012148.0,
      "step": 3060
    },
    {
      "entropy": 5.570310926437378,
      "epoch": 0.29442843419788667,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.4448,
      "mean_token_accuracy": 0.18498128652572632,
      "num_tokens": 7023238.0,
      "step": 3065
    },
    {
      "entropy": 5.627264976501465,
      "epoch": 0.29490874159462055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1867678791284561,
      "num_tokens": 7034944.0,
      "step": 3070
    },
    {
      "entropy": 5.691465091705322,
      "epoch": 0.2953890489913545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499550745832149,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19256407022476196,
      "num_tokens": 7046880.0,
      "step": 3075
    },
    {
      "entropy": 5.630908727645874,
      "epoch": 0.29586935638808837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.4752,
      "mean_token_accuracy": 0.18294412046670913,
      "num_tokens": 7057678.0,
      "step": 3080
    },
    {
      "entropy": 5.569836091995239,
      "epoch": 0.2963496637848223,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.443,
      "mean_token_accuracy": 0.1914595991373062,
      "num_tokens": 7068336.0,
      "step": 3085
    },
    {
      "entropy": 5.677788400650025,
      "epoch": 0.2968299711815562,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5156,
      "mean_token_accuracy": 0.17752356827259064,
      "num_tokens": 7080341.0,
      "step": 3090
    },
    {
      "entropy": 5.829857444763183,
      "epoch": 0.2973102785782901,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.6642,
      "mean_token_accuracy": 0.17347445338964462,
      "num_tokens": 7090618.0,
      "step": 3095
    },
    {
      "entropy": 5.578703689575195,
      "epoch": 0.297790585975024,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19119216352701188,
      "num_tokens": 7101843.0,
      "step": 3100
    },
    {
      "entropy": 5.6363883972167965,
      "epoch": 0.29827089337175794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.4874,
      "mean_token_accuracy": 0.1814226746559143,
      "num_tokens": 7112014.0,
      "step": 3105
    },
    {
      "entropy": 5.710436534881592,
      "epoch": 0.2987512007684918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.5624,
      "mean_token_accuracy": 0.17562751471996307,
      "num_tokens": 7123860.0,
      "step": 3110
    },
    {
      "entropy": 5.6648670673370365,
      "epoch": 0.29923150816522576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.5357,
      "mean_token_accuracy": 0.18224934190511705,
      "num_tokens": 7135901.0,
      "step": 3115
    },
    {
      "entropy": 5.694220733642578,
      "epoch": 0.29971181556195964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.4932,
      "mean_token_accuracy": 0.19112819135189058,
      "num_tokens": 7148743.0,
      "step": 3120
    },
    {
      "entropy": 5.702276515960693,
      "epoch": 0.3001921229586936,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.5516,
      "mean_token_accuracy": 0.18374135941267014,
      "num_tokens": 7160731.0,
      "step": 3125
    },
    {
      "entropy": 5.703460311889648,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5461,
      "mean_token_accuracy": 0.181298191845417,
      "num_tokens": 7172861.0,
      "step": 3130
    },
    {
      "entropy": 5.636017751693726,
      "epoch": 0.3011527377521614,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.4796,
      "mean_token_accuracy": 0.18060984313488007,
      "num_tokens": 7183954.0,
      "step": 3135
    },
    {
      "entropy": 5.591547632217408,
      "epoch": 0.3016330451488953,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4632,
      "mean_token_accuracy": 0.18435155898332595,
      "num_tokens": 7195466.0,
      "step": 3140
    },
    {
      "entropy": 5.671104192733765,
      "epoch": 0.3021133525456292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.45,
      "mean_token_accuracy": 0.18866306990385057,
      "num_tokens": 7207173.0,
      "step": 3145
    },
    {
      "entropy": 5.707768821716309,
      "epoch": 0.3025936599423631,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.5867,
      "mean_token_accuracy": 0.18185509890317916,
      "num_tokens": 7220095.0,
      "step": 3150
    },
    {
      "entropy": 5.608586359024048,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18203232288360596,
      "num_tokens": 7230664.0,
      "step": 3155
    },
    {
      "entropy": 5.678262376785279,
      "epoch": 0.3035542747358309,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.4528,
      "mean_token_accuracy": 0.19190701693296433,
      "num_tokens": 7241278.0,
      "step": 3160
    },
    {
      "entropy": 5.556720972061157,
      "epoch": 0.30403458213256485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.6049,
      "mean_token_accuracy": 0.17547910213470458,
      "num_tokens": 7252999.0,
      "step": 3165
    },
    {
      "entropy": 5.738193988800049,
      "epoch": 0.3045148895292987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18523052483797073,
      "num_tokens": 7263949.0,
      "step": 3170
    },
    {
      "entropy": 5.610001945495606,
      "epoch": 0.30499519692603266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.4864,
      "mean_token_accuracy": 0.18810611367225646,
      "num_tokens": 7275467.0,
      "step": 3175
    },
    {
      "entropy": 5.6769002914428714,
      "epoch": 0.30547550432276654,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4697,
      "mean_token_accuracy": 0.1874818757176399,
      "num_tokens": 7285534.0,
      "step": 3180
    },
    {
      "entropy": 5.706569576263428,
      "epoch": 0.3059558117195005,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499501845410965,
      "loss": 5.5873,
      "mean_token_accuracy": 0.18355770856142045,
      "num_tokens": 7297252.0,
      "step": 3185
    },
    {
      "entropy": 5.64103512763977,
      "epoch": 0.30643611911623436,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.443,
      "mean_token_accuracy": 0.1906636580824852,
      "num_tokens": 7308492.0,
      "step": 3190
    },
    {
      "entropy": 5.620066118240357,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.4835,
      "mean_token_accuracy": 0.1803540736436844,
      "num_tokens": 7319703.0,
      "step": 3195
    },
    {
      "entropy": 5.6844628810882565,
      "epoch": 0.30739673390970224,
      "grad_norm": 1.0,
      "learning_rate": 0.000499494981699318,
      "loss": 5.4619,
      "mean_token_accuracy": 0.18943321704864502,
      "num_tokens": 7331022.0,
      "step": 3200
    },
    {
      "entropy": 5.6341009616851805,
      "epoch": 0.3078770413064361,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499492683371151,
      "loss": 5.505,
      "mean_token_accuracy": 0.18882815837860106,
      "num_tokens": 7342977.0,
      "step": 3205
    },
    {
      "entropy": 5.703154134750366,
      "epoch": 0.30835734870317005,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.4818,
      "mean_token_accuracy": 0.18102553188800813,
      "num_tokens": 7353227.0,
      "step": 3210
    },
    {
      "entropy": 5.651052856445313,
      "epoch": 0.30883765609990393,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5372,
      "mean_token_accuracy": 0.16935457289218903,
      "num_tokens": 7364165.0,
      "step": 3215
    },
    {
      "entropy": 5.601374197006225,
      "epoch": 0.30931796349663787,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18606537580490112,
      "num_tokens": 7374800.0,
      "step": 3220
    },
    {
      "entropy": 5.573496866226196,
      "epoch": 0.30979827089337175,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.4187,
      "mean_token_accuracy": 0.18962224274873735,
      "num_tokens": 7386360.0,
      "step": 3225
    },
    {
      "entropy": 5.631180143356323,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4459,
      "mean_token_accuracy": 0.1859535649418831,
      "num_tokens": 7397066.0,
      "step": 3230
    },
    {
      "entropy": 5.6822919845581055,
      "epoch": 0.31075888568683957,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.5174,
      "mean_token_accuracy": 0.17699058800935746,
      "num_tokens": 7408349.0,
      "step": 3235
    },
    {
      "entropy": 5.69844274520874,
      "epoch": 0.3112391930835735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.5627,
      "mean_token_accuracy": 0.17459277361631392,
      "num_tokens": 7420343.0,
      "step": 3240
    },
    {
      "entropy": 5.702674007415771,
      "epoch": 0.3117195004803074,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.4782,
      "mean_token_accuracy": 0.1868354335427284,
      "num_tokens": 7431683.0,
      "step": 3245
    },
    {
      "entropy": 5.497278642654419,
      "epoch": 0.3121998078770413,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499471763881684,
      "loss": 5.3854,
      "mean_token_accuracy": 0.18908552974462509,
      "num_tokens": 7443327.0,
      "step": 3250
    },
    {
      "entropy": 5.6632318019866945,
      "epoch": 0.3126801152737752,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.4845,
      "mean_token_accuracy": 0.18859678357839585,
      "num_tokens": 7454002.0,
      "step": 3255
    },
    {
      "entropy": 5.57216854095459,
      "epoch": 0.31316042267050914,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499467057777609,
      "loss": 5.4814,
      "mean_token_accuracy": 0.1825019970536232,
      "num_tokens": 7464074.0,
      "step": 3260
    },
    {
      "entropy": 5.594534063339234,
      "epoch": 0.313640730067243,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4098,
      "mean_token_accuracy": 0.19650202989578247,
      "num_tokens": 7475543.0,
      "step": 3265
    },
    {
      "entropy": 5.698621129989624,
      "epoch": 0.31412103746397696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.468,
      "mean_token_accuracy": 0.1819482773542404,
      "num_tokens": 7486818.0,
      "step": 3270
    },
    {
      "entropy": 5.642549324035644,
      "epoch": 0.31460134486071084,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499459959535695,
      "loss": 5.5132,
      "mean_token_accuracy": 0.18159783780574798,
      "num_tokens": 7499046.0,
      "step": 3275
    },
    {
      "entropy": 5.616050767898559,
      "epoch": 0.3150816522574448,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.4514,
      "mean_token_accuracy": 0.17918917685747146,
      "num_tokens": 7509853.0,
      "step": 3280
    },
    {
      "entropy": 5.551430940628052,
      "epoch": 0.31556195965417866,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.3068,
      "mean_token_accuracy": 0.18969039022922515,
      "num_tokens": 7521091.0,
      "step": 3285
    },
    {
      "entropy": 5.587959337234497,
      "epoch": 0.3160422670509126,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4236,
      "mean_token_accuracy": 0.1940646082162857,
      "num_tokens": 7531317.0,
      "step": 3290
    },
    {
      "entropy": 5.627415990829467,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.5672,
      "mean_token_accuracy": 0.1813266508281231,
      "num_tokens": 7542822.0,
      "step": 3295
    },
    {
      "entropy": 5.725085687637329,
      "epoch": 0.3170028818443804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.5085,
      "mean_token_accuracy": 0.17990946918725967,
      "num_tokens": 7552434.0,
      "step": 3300
    },
    {
      "entropy": 5.583597040176391,
      "epoch": 0.3174831892411143,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.3771,
      "mean_token_accuracy": 0.19330264627933502,
      "num_tokens": 7563895.0,
      "step": 3305
    },
    {
      "entropy": 5.4530237197875975,
      "epoch": 0.31796349663784823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18980606347322465,
      "num_tokens": 7575391.0,
      "step": 3310
    },
    {
      "entropy": 5.681956958770752,
      "epoch": 0.3184438040345821,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.5493,
      "mean_token_accuracy": 0.19029652178287507,
      "num_tokens": 7587119.0,
      "step": 3315
    },
    {
      "entropy": 5.655573844909668,
      "epoch": 0.31892411143131605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.5209,
      "mean_token_accuracy": 0.17941325157880783,
      "num_tokens": 7598615.0,
      "step": 3320
    },
    {
      "entropy": 5.637982654571533,
      "epoch": 0.31940441882804993,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.4826,
      "mean_token_accuracy": 0.19178740084171295,
      "num_tokens": 7610159.0,
      "step": 3325
    },
    {
      "entropy": 5.707050704956055,
      "epoch": 0.31988472622478387,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.4699,
      "mean_token_accuracy": 0.18814541697502135,
      "num_tokens": 7620922.0,
      "step": 3330
    },
    {
      "entropy": 5.567379283905029,
      "epoch": 0.32036503362151775,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.386,
      "mean_token_accuracy": 0.19375266283750534,
      "num_tokens": 7632343.0,
      "step": 3335
    },
    {
      "entropy": 5.705099582672119,
      "epoch": 0.3208453410182517,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.5929,
      "mean_token_accuracy": 0.1719297081232071,
      "num_tokens": 7644748.0,
      "step": 3340
    },
    {
      "entropy": 5.70050573348999,
      "epoch": 0.32132564841498557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499426214287049,
      "loss": 5.5325,
      "mean_token_accuracy": 0.19119921773672105,
      "num_tokens": 7655449.0,
      "step": 3345
    },
    {
      "entropy": 5.643219518661499,
      "epoch": 0.3218059558117195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.506,
      "mean_token_accuracy": 0.1850568026304245,
      "num_tokens": 7665623.0,
      "step": 3350
    },
    {
      "entropy": 5.587236118316651,
      "epoch": 0.3222862632084534,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.3992,
      "mean_token_accuracy": 0.18939037919044494,
      "num_tokens": 7676860.0,
      "step": 3355
    },
    {
      "entropy": 5.712094449996949,
      "epoch": 0.3227665706051873,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499418850289057,
      "loss": 5.5746,
      "mean_token_accuracy": 0.18135148137807847,
      "num_tokens": 7687778.0,
      "step": 3360
    },
    {
      "entropy": 5.674486255645752,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.4749,
      "mean_token_accuracy": 0.18185528069734574,
      "num_tokens": 7699154.0,
      "step": 3365
    },
    {
      "entropy": 5.571884107589722,
      "epoch": 0.32372718539865514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.481,
      "mean_token_accuracy": 0.1896562933921814,
      "num_tokens": 7711010.0,
      "step": 3370
    },
    {
      "entropy": 5.700831890106201,
      "epoch": 0.3242074927953891,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.5075,
      "mean_token_accuracy": 0.18210598528385163,
      "num_tokens": 7721969.0,
      "step": 3375
    },
    {
      "entropy": 5.675892448425293,
      "epoch": 0.32468780019212296,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.4505,
      "mean_token_accuracy": 0.189385986328125,
      "num_tokens": 7733762.0,
      "step": 3380
    },
    {
      "entropy": 5.599011373519898,
      "epoch": 0.3251681075888569,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.4692,
      "mean_token_accuracy": 0.18583348244428635,
      "num_tokens": 7744614.0,
      "step": 3385
    },
    {
      "entropy": 5.595316648483276,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.4031,
      "mean_token_accuracy": 0.19141158014535903,
      "num_tokens": 7755799.0,
      "step": 3390
    },
    {
      "entropy": 5.563256788253784,
      "epoch": 0.3261287223823247,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049940148526788,
      "loss": 5.4626,
      "mean_token_accuracy": 0.1855825215578079,
      "num_tokens": 7768140.0,
      "step": 3395
    },
    {
      "entropy": 5.60364294052124,
      "epoch": 0.3266090297790586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4585,
      "mean_token_accuracy": 0.19093261510133744,
      "num_tokens": 7778494.0,
      "step": 3400
    },
    {
      "entropy": 5.65981650352478,
      "epoch": 0.3270893371757925,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.5256,
      "mean_token_accuracy": 0.18769457191228867,
      "num_tokens": 7789234.0,
      "step": 3405
    },
    {
      "entropy": 5.611981296539307,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.5214,
      "mean_token_accuracy": 0.18589239120483397,
      "num_tokens": 7800602.0,
      "step": 3410
    },
    {
      "entropy": 5.560681581497192,
      "epoch": 0.32804995196926034,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1929910495877266,
      "num_tokens": 7812803.0,
      "step": 3415
    },
    {
      "entropy": 5.772423648834229,
      "epoch": 0.3285302593659942,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.6264,
      "mean_token_accuracy": 0.17467853426933289,
      "num_tokens": 7825432.0,
      "step": 3420
    },
    {
      "entropy": 5.597874021530151,
      "epoch": 0.32901056676272816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.4575,
      "mean_token_accuracy": 0.18343233615159987,
      "num_tokens": 7837258.0,
      "step": 3425
    },
    {
      "entropy": 5.565313911437988,
      "epoch": 0.32949087415946204,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.3389,
      "mean_token_accuracy": 0.19879245907068252,
      "num_tokens": 7847573.0,
      "step": 3430
    },
    {
      "entropy": 5.640517377853394,
      "epoch": 0.329971181556196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1822481170296669,
      "num_tokens": 7857957.0,
      "step": 3435
    },
    {
      "entropy": 5.587132692337036,
      "epoch": 0.33045148895292986,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.3933,
      "mean_token_accuracy": 0.19951517283916473,
      "num_tokens": 7867996.0,
      "step": 3440
    },
    {
      "entropy": 5.546762084960937,
      "epoch": 0.3309317963496638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.4531,
      "mean_token_accuracy": 0.18115595430135728,
      "num_tokens": 7879245.0,
      "step": 3445
    },
    {
      "entropy": 5.667378091812134,
      "epoch": 0.3314121037463977,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.531,
      "mean_token_accuracy": 0.17801956832408905,
      "num_tokens": 7891186.0,
      "step": 3450
    },
    {
      "entropy": 5.718769073486328,
      "epoch": 0.3318924111431316,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.5526,
      "mean_token_accuracy": 0.18821173906326294,
      "num_tokens": 7902231.0,
      "step": 3455
    },
    {
      "entropy": 5.631023216247558,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.5036,
      "mean_token_accuracy": 0.187532676756382,
      "num_tokens": 7913364.0,
      "step": 3460
    },
    {
      "entropy": 5.843285274505615,
      "epoch": 0.33285302593659943,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.6988,
      "mean_token_accuracy": 0.1718384563922882,
      "num_tokens": 7925217.0,
      "step": 3465
    },
    {
      "entropy": 5.5377867221832275,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.4338,
      "mean_token_accuracy": 0.18306865096092223,
      "num_tokens": 7937636.0,
      "step": 3470
    },
    {
      "entropy": 5.622658777236938,
      "epoch": 0.33381364073006725,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4672,
      "mean_token_accuracy": 0.17787594795227052,
      "num_tokens": 7948958.0,
      "step": 3475
    },
    {
      "entropy": 5.690193700790405,
      "epoch": 0.33429394812680113,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18434032946825027,
      "num_tokens": 7960013.0,
      "step": 3480
    },
    {
      "entropy": 5.594874286651612,
      "epoch": 0.33477425552353507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.5073,
      "mean_token_accuracy": 0.18652689158916474,
      "num_tokens": 7971041.0,
      "step": 3485
    },
    {
      "entropy": 5.7146190166473385,
      "epoch": 0.33525456292026895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.5267,
      "mean_token_accuracy": 0.1811326265335083,
      "num_tokens": 7982752.0,
      "step": 3490
    },
    {
      "entropy": 5.6629688262939455,
      "epoch": 0.3357348703170029,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4039,
      "mean_token_accuracy": 0.18894438147544862,
      "num_tokens": 7993681.0,
      "step": 3495
    },
    {
      "entropy": 5.592116165161133,
      "epoch": 0.33621517771373677,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.4337,
      "mean_token_accuracy": 0.1875965714454651,
      "num_tokens": 8004727.0,
      "step": 3500
    },
    {
      "entropy": 5.604901647567749,
      "epoch": 0.3366954851104707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.4503,
      "mean_token_accuracy": 0.19188206791877746,
      "num_tokens": 8015423.0,
      "step": 3505
    },
    {
      "entropy": 5.629399013519287,
      "epoch": 0.3371757925072046,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.4163,
      "mean_token_accuracy": 0.19524227529764177,
      "num_tokens": 8027911.0,
      "step": 3510
    },
    {
      "entropy": 5.630470561981201,
      "epoch": 0.3376560999039385,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4435,
      "mean_token_accuracy": 0.19002441614866256,
      "num_tokens": 8038831.0,
      "step": 3515
    },
    {
      "entropy": 5.485448122024536,
      "epoch": 0.3381364073006724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.333,
      "mean_token_accuracy": 0.19800836890935897,
      "num_tokens": 8049906.0,
      "step": 3520
    },
    {
      "entropy": 5.534181308746338,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.4262,
      "mean_token_accuracy": 0.18656017929315566,
      "num_tokens": 8061158.0,
      "step": 3525
    },
    {
      "entropy": 5.679272985458374,
      "epoch": 0.3390970220941403,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.4049,
      "mean_token_accuracy": 0.18500697761774063,
      "num_tokens": 8071958.0,
      "step": 3530
    },
    {
      "entropy": 5.590602445602417,
      "epoch": 0.33957732949087416,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499329472516601,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1830559030175209,
      "num_tokens": 8084068.0,
      "step": 3535
    },
    {
      "entropy": 5.501382446289062,
      "epoch": 0.3400576368876081,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.3107,
      "mean_token_accuracy": 0.2009879857301712,
      "num_tokens": 8096132.0,
      "step": 3540
    },
    {
      "entropy": 5.591581535339356,
      "epoch": 0.340537944284342,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.4852,
      "mean_token_accuracy": 0.18972329199314117,
      "num_tokens": 8107327.0,
      "step": 3545
    },
    {
      "entropy": 5.586108875274658,
      "epoch": 0.3410182516810759,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499321514656929,
      "loss": 5.451,
      "mean_token_accuracy": 0.18751484006643296,
      "num_tokens": 8118584.0,
      "step": 3550
    },
    {
      "entropy": 5.632966232299805,
      "epoch": 0.3414985590778098,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.4978,
      "mean_token_accuracy": 0.18059134632349014,
      "num_tokens": 8130081.0,
      "step": 3555
    },
    {
      "entropy": 5.603939819335937,
      "epoch": 0.34197886647454373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4545,
      "mean_token_accuracy": 0.18668154925107955,
      "num_tokens": 8141463.0,
      "step": 3560
    },
    {
      "entropy": 5.662025403976441,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18695254474878312,
      "num_tokens": 8153863.0,
      "step": 3565
    },
    {
      "entropy": 5.524808263778686,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.423,
      "mean_token_accuracy": 0.18549051284790039,
      "num_tokens": 8165695.0,
      "step": 3570
    },
    {
      "entropy": 5.607334613800049,
      "epoch": 0.34341978866474543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.4845,
      "mean_token_accuracy": 0.1852959141135216,
      "num_tokens": 8176456.0,
      "step": 3575
    },
    {
      "entropy": 5.65879111289978,
      "epoch": 0.34390009606147937,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.4549,
      "mean_token_accuracy": 0.19000289142131804,
      "num_tokens": 8189050.0,
      "step": 3580
    },
    {
      "entropy": 5.562312746047974,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.3878,
      "mean_token_accuracy": 0.18902230262756348,
      "num_tokens": 8200011.0,
      "step": 3585
    },
    {
      "entropy": 5.573532772064209,
      "epoch": 0.3448607108549472,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499300064546858,
      "loss": 5.3829,
      "mean_token_accuracy": 0.19188244193792342,
      "num_tokens": 8211770.0,
      "step": 3590
    },
    {
      "entropy": 5.589154958724976,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.4252,
      "mean_token_accuracy": 0.19337030947208406,
      "num_tokens": 8223582.0,
      "step": 3595
    },
    {
      "entropy": 5.698397827148438,
      "epoch": 0.345821325648415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499294649942276,
      "loss": 5.553,
      "mean_token_accuracy": 0.18515172749757766,
      "num_tokens": 8234336.0,
      "step": 3600
    },
    {
      "entropy": 5.645198345184326,
      "epoch": 0.3463016330451489,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4546,
      "mean_token_accuracy": 0.187334606051445,
      "num_tokens": 8244605.0,
      "step": 3605
    },
    {
      "entropy": 5.615684843063354,
      "epoch": 0.3467819404418828,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.5042,
      "mean_token_accuracy": 0.18244509696960448,
      "num_tokens": 8255876.0,
      "step": 3610
    },
    {
      "entropy": 5.569511413574219,
      "epoch": 0.3472622478386167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.3968,
      "mean_token_accuracy": 0.19007992446422578,
      "num_tokens": 8266602.0,
      "step": 3615
    },
    {
      "entropy": 5.599522733688355,
      "epoch": 0.34774255523535064,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.38,
      "mean_token_accuracy": 0.1920741856098175,
      "num_tokens": 8279804.0,
      "step": 3620
    },
    {
      "entropy": 5.669801664352417,
      "epoch": 0.3482228626320845,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.4916,
      "mean_token_accuracy": 0.18763836473226547,
      "num_tokens": 8291020.0,
      "step": 3625
    },
    {
      "entropy": 5.6229675769805905,
      "epoch": 0.34870317002881845,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4266,
      "mean_token_accuracy": 0.19148036688566208,
      "num_tokens": 8302192.0,
      "step": 3630
    },
    {
      "entropy": 5.604986190795898,
      "epoch": 0.34918347742555234,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.4054,
      "mean_token_accuracy": 0.1876965194940567,
      "num_tokens": 8313649.0,
      "step": 3635
    },
    {
      "entropy": 5.560658979415893,
      "epoch": 0.34966378482228627,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.4003,
      "mean_token_accuracy": 0.19191381931304932,
      "num_tokens": 8324694.0,
      "step": 3640
    },
    {
      "entropy": 5.679537630081176,
      "epoch": 0.35014409221902015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.5466,
      "mean_token_accuracy": 0.17840450406074523,
      "num_tokens": 8336517.0,
      "step": 3645
    },
    {
      "entropy": 5.643055534362793,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.4974,
      "mean_token_accuracy": 0.18391885459423066,
      "num_tokens": 8349001.0,
      "step": 3650
    },
    {
      "entropy": 5.640076971054077,
      "epoch": 0.35110470701248797,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.4248,
      "mean_token_accuracy": 0.18910549581050873,
      "num_tokens": 8361230.0,
      "step": 3655
    },
    {
      "entropy": 5.5414430618286135,
      "epoch": 0.3515850144092219,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.3548,
      "mean_token_accuracy": 0.19453334510326387,
      "num_tokens": 8372159.0,
      "step": 3660
    },
    {
      "entropy": 5.589181232452392,
      "epoch": 0.3520653218059558,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.4684,
      "mean_token_accuracy": 0.18958093523979186,
      "num_tokens": 8383228.0,
      "step": 3665
    },
    {
      "entropy": 5.648845243453979,
      "epoch": 0.3525456292026897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.4072,
      "mean_token_accuracy": 0.18761504739522933,
      "num_tokens": 8394582.0,
      "step": 3670
    },
    {
      "entropy": 5.605154609680175,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.4197,
      "mean_token_accuracy": 0.18835971355438233,
      "num_tokens": 8406784.0,
      "step": 3675
    },
    {
      "entropy": 5.585361003875732,
      "epoch": 0.35350624399615754,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.4038,
      "mean_token_accuracy": 0.19672708660364152,
      "num_tokens": 8418405.0,
      "step": 3680
    },
    {
      "entropy": 5.5960245609283445,
      "epoch": 0.3539865513928914,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3541,
      "mean_token_accuracy": 0.18893332481384278,
      "num_tokens": 8430432.0,
      "step": 3685
    },
    {
      "entropy": 5.607988166809082,
      "epoch": 0.35446685878962536,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.5037,
      "mean_token_accuracy": 0.18349802643060684,
      "num_tokens": 8442423.0,
      "step": 3690
    },
    {
      "entropy": 5.622403383255005,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.5182,
      "mean_token_accuracy": 0.18986132144927978,
      "num_tokens": 8454951.0,
      "step": 3695
    },
    {
      "entropy": 5.564925622940064,
      "epoch": 0.3554274735830932,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.3553,
      "mean_token_accuracy": 0.19411883056163787,
      "num_tokens": 8467758.0,
      "step": 3700
    },
    {
      "entropy": 5.60244460105896,
      "epoch": 0.3559077809798271,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18677409291267394,
      "num_tokens": 8479728.0,
      "step": 3705
    },
    {
      "entropy": 5.604581356048584,
      "epoch": 0.356388088376561,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.5095,
      "mean_token_accuracy": 0.18320033997297286,
      "num_tokens": 8492099.0,
      "step": 3710
    },
    {
      "entropy": 5.657941150665283,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.4536,
      "mean_token_accuracy": 0.17795021831989288,
      "num_tokens": 8504657.0,
      "step": 3715
    },
    {
      "entropy": 5.625857305526734,
      "epoch": 0.3573487031700288,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4227,
      "mean_token_accuracy": 0.18991409093141556,
      "num_tokens": 8514728.0,
      "step": 3720
    },
    {
      "entropy": 5.62817611694336,
      "epoch": 0.35782901056676275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.5363,
      "mean_token_accuracy": 0.18928872793912888,
      "num_tokens": 8525588.0,
      "step": 3725
    },
    {
      "entropy": 5.618461799621582,
      "epoch": 0.35830931796349663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4997,
      "mean_token_accuracy": 0.18794921934604644,
      "num_tokens": 8537468.0,
      "step": 3730
    },
    {
      "entropy": 5.642933511734009,
      "epoch": 0.35878962536023057,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.4099,
      "mean_token_accuracy": 0.1922216534614563,
      "num_tokens": 8548598.0,
      "step": 3735
    },
    {
      "entropy": 5.412218189239502,
      "epoch": 0.35926993275696445,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.2967,
      "mean_token_accuracy": 0.20350995361804963,
      "num_tokens": 8560301.0,
      "step": 3740
    },
    {
      "entropy": 5.571508169174194,
      "epoch": 0.3597502401536984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.3904,
      "mean_token_accuracy": 0.19346483945846557,
      "num_tokens": 8571186.0,
      "step": 3745
    },
    {
      "entropy": 5.607302665710449,
      "epoch": 0.36023054755043227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19522017240524292,
      "num_tokens": 8580725.0,
      "step": 3750
    },
    {
      "entropy": 5.5793187618255615,
      "epoch": 0.3607108549471662,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18237872421741486,
      "num_tokens": 8594598.0,
      "step": 3755
    },
    {
      "entropy": 5.718835163116455,
      "epoch": 0.3611911623439001,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.5138,
      "mean_token_accuracy": 0.18184855878353118,
      "num_tokens": 8606733.0,
      "step": 3760
    },
    {
      "entropy": 5.649420595169067,
      "epoch": 0.361671469740634,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1857554316520691,
      "num_tokens": 8618863.0,
      "step": 3765
    },
    {
      "entropy": 5.453961610794067,
      "epoch": 0.3621517771373679,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.2943,
      "mean_token_accuracy": 0.1941133290529251,
      "num_tokens": 8629270.0,
      "step": 3770
    },
    {
      "entropy": 5.610463285446167,
      "epoch": 0.36263208453410184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.4909,
      "mean_token_accuracy": 0.19132499098777772,
      "num_tokens": 8641363.0,
      "step": 3775
    },
    {
      "entropy": 5.66168212890625,
      "epoch": 0.3631123919308357,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.4499,
      "mean_token_accuracy": 0.1907256633043289,
      "num_tokens": 8653233.0,
      "step": 3780
    },
    {
      "entropy": 5.571015691757202,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499190723117722,
      "loss": 5.4484,
      "mean_token_accuracy": 0.18239522725343704,
      "num_tokens": 8665192.0,
      "step": 3785
    },
    {
      "entropy": 5.669944858551025,
      "epoch": 0.36407300672430354,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.527,
      "mean_token_accuracy": 0.181233774125576,
      "num_tokens": 8677669.0,
      "step": 3790
    },
    {
      "entropy": 5.529652452468872,
      "epoch": 0.3645533141210375,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19549092054367065,
      "num_tokens": 8688002.0,
      "step": 3795
    },
    {
      "entropy": 5.552938508987427,
      "epoch": 0.36503362151777136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.3809,
      "mean_token_accuracy": 0.19497113823890685,
      "num_tokens": 8698396.0,
      "step": 3800
    },
    {
      "entropy": 5.656788778305054,
      "epoch": 0.3655139289145053,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.3855,
      "mean_token_accuracy": 0.19022500067949294,
      "num_tokens": 8711135.0,
      "step": 3805
    },
    {
      "entropy": 5.545717477798462,
      "epoch": 0.3659942363112392,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499176132348474,
      "loss": 5.4329,
      "mean_token_accuracy": 0.19408364295959474,
      "num_tokens": 8723707.0,
      "step": 3810
    },
    {
      "entropy": 5.6224829196929935,
      "epoch": 0.3664745437079731,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.3788,
      "mean_token_accuracy": 0.19512805342674255,
      "num_tokens": 8734193.0,
      "step": 3815
    },
    {
      "entropy": 5.575182294845581,
      "epoch": 0.366954851104707,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.4498,
      "mean_token_accuracy": 0.18577555567026138,
      "num_tokens": 8745619.0,
      "step": 3820
    },
    {
      "entropy": 5.521268081665039,
      "epoch": 0.36743515850144093,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18944036811590195,
      "num_tokens": 8757331.0,
      "step": 3825
    },
    {
      "entropy": 5.584135866165161,
      "epoch": 0.3679154658981748,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.3848,
      "mean_token_accuracy": 0.19181567877531053,
      "num_tokens": 8768840.0,
      "step": 3830
    },
    {
      "entropy": 5.572615623474121,
      "epoch": 0.36839577329490875,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19323140680789946,
      "num_tokens": 8781214.0,
      "step": 3835
    },
    {
      "entropy": 5.598087167739868,
      "epoch": 0.3688760806916426,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.407,
      "mean_token_accuracy": 0.18879638165235518,
      "num_tokens": 8791645.0,
      "step": 3840
    },
    {
      "entropy": 5.666203022003174,
      "epoch": 0.36935638808837656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.58,
      "mean_token_accuracy": 0.17914286777377128,
      "num_tokens": 8803286.0,
      "step": 3845
    },
    {
      "entropy": 5.6028059959411625,
      "epoch": 0.36983669548511044,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3582,
      "mean_token_accuracy": 0.19065531343221664,
      "num_tokens": 8814207.0,
      "step": 3850
    },
    {
      "entropy": 5.574209070205688,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.4283,
      "mean_token_accuracy": 0.18504399955272674,
      "num_tokens": 8825540.0,
      "step": 3855
    },
    {
      "entropy": 5.650593423843384,
      "epoch": 0.37079731027857826,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.5504,
      "mean_token_accuracy": 0.17896147668361664,
      "num_tokens": 8838182.0,
      "step": 3860
    },
    {
      "entropy": 5.579078006744385,
      "epoch": 0.3712776176753122,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.4175,
      "mean_token_accuracy": 0.1951797842979431,
      "num_tokens": 8850211.0,
      "step": 3865
    },
    {
      "entropy": 5.548032665252686,
      "epoch": 0.37175792507204614,
      "grad_norm": 0.875,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.4724,
      "mean_token_accuracy": 0.18146964982151986,
      "num_tokens": 8862804.0,
      "step": 3870
    },
    {
      "entropy": 5.588915157318115,
      "epoch": 0.37223823246878,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1890851303935051,
      "num_tokens": 8874112.0,
      "step": 3875
    },
    {
      "entropy": 5.566713428497314,
      "epoch": 0.37271853986551395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4356,
      "mean_token_accuracy": 0.18930822014808654,
      "num_tokens": 8885114.0,
      "step": 3880
    },
    {
      "entropy": 5.597771644592285,
      "epoch": 0.37319884726224783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.329,
      "mean_token_accuracy": 0.1946409046649933,
      "num_tokens": 8895555.0,
      "step": 3885
    },
    {
      "entropy": 5.520039701461792,
      "epoch": 0.37367915465898177,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.3281,
      "mean_token_accuracy": 0.20043965727090834,
      "num_tokens": 8906073.0,
      "step": 3890
    },
    {
      "entropy": 5.59633960723877,
      "epoch": 0.37415946205571565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.4848,
      "mean_token_accuracy": 0.1856019526720047,
      "num_tokens": 8916587.0,
      "step": 3895
    },
    {
      "entropy": 5.615598583221436,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3741,
      "mean_token_accuracy": 0.1854495272040367,
      "num_tokens": 8927923.0,
      "step": 3900
    },
    {
      "entropy": 5.522138500213623,
      "epoch": 0.37512007684918347,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499119500825865,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19048261940479277,
      "num_tokens": 8939939.0,
      "step": 3905
    },
    {
      "entropy": 5.522910451889038,
      "epoch": 0.3756003842459174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.4,
      "mean_token_accuracy": 0.19586212038993836,
      "num_tokens": 8951748.0,
      "step": 3910
    },
    {
      "entropy": 5.554461240768433,
      "epoch": 0.3760806916426513,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.32,
      "mean_token_accuracy": 0.19885835498571397,
      "num_tokens": 8962922.0,
      "step": 3915
    },
    {
      "entropy": 5.472480392456054,
      "epoch": 0.3765609990393852,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.2987,
      "mean_token_accuracy": 0.2007770374417305,
      "num_tokens": 8974013.0,
      "step": 3920
    },
    {
      "entropy": 5.536104154586792,
      "epoch": 0.3770413064361191,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499107339028103,
      "loss": 5.377,
      "mean_token_accuracy": 0.18797796219587326,
      "num_tokens": 8986032.0,
      "step": 3925
    },
    {
      "entropy": 5.624913358688355,
      "epoch": 0.37752161383285304,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.3747,
      "mean_token_accuracy": 0.18829465806484222,
      "num_tokens": 8996889.0,
      "step": 3930
    },
    {
      "entropy": 5.5282268047332765,
      "epoch": 0.3780019212295869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3596,
      "mean_token_accuracy": 0.19118607342243193,
      "num_tokens": 9007594.0,
      "step": 3935
    },
    {
      "entropy": 5.473355913162232,
      "epoch": 0.37848222862632086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.3455,
      "mean_token_accuracy": 0.18814336955547334,
      "num_tokens": 9018097.0,
      "step": 3940
    },
    {
      "entropy": 5.59659333229065,
      "epoch": 0.37896253602305474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.4724,
      "mean_token_accuracy": 0.18152033388614655,
      "num_tokens": 9029554.0,
      "step": 3945
    },
    {
      "entropy": 5.562332487106323,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3375,
      "mean_token_accuracy": 0.19086694419384004,
      "num_tokens": 9042518.0,
      "step": 3950
    },
    {
      "entropy": 5.520054674148559,
      "epoch": 0.37992315081652256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3912,
      "mean_token_accuracy": 0.19118970036506652,
      "num_tokens": 9054524.0,
      "step": 3955
    },
    {
      "entropy": 5.604040002822876,
      "epoch": 0.3804034582132565,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19073721766471863,
      "num_tokens": 9065375.0,
      "step": 3960
    },
    {
      "entropy": 5.519604921340942,
      "epoch": 0.3808837656099904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.3067,
      "mean_token_accuracy": 0.19381832629442214,
      "num_tokens": 9076338.0,
      "step": 3965
    },
    {
      "entropy": 5.574223232269287,
      "epoch": 0.3813640730067243,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.3524,
      "mean_token_accuracy": 0.19108670055866242,
      "num_tokens": 9088407.0,
      "step": 3970
    },
    {
      "entropy": 5.517913103103638,
      "epoch": 0.3818443804034582,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.3182,
      "mean_token_accuracy": 0.19173867255449295,
      "num_tokens": 9099833.0,
      "step": 3975
    },
    {
      "entropy": 5.397404193878174,
      "epoch": 0.38232468780019213,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.2948,
      "mean_token_accuracy": 0.19632792323827744,
      "num_tokens": 9111126.0,
      "step": 3980
    },
    {
      "entropy": 5.605161476135254,
      "epoch": 0.382804995196926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.3929,
      "mean_token_accuracy": 0.19436297714710235,
      "num_tokens": 9121979.0,
      "step": 3985
    },
    {
      "entropy": 5.5903404712677,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.4058,
      "mean_token_accuracy": 0.19064545035362243,
      "num_tokens": 9132929.0,
      "step": 3990
    },
    {
      "entropy": 5.596808624267578,
      "epoch": 0.38376560999039383,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3732,
      "mean_token_accuracy": 0.1942216232419014,
      "num_tokens": 9143903.0,
      "step": 3995
    },
    {
      "entropy": 5.588313817977905,
      "epoch": 0.38424591738712777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.4502,
      "mean_token_accuracy": 0.18742062598466874,
      "num_tokens": 9154416.0,
      "step": 4000
    },
    {
      "entropy": 5.582321405410767,
      "epoch": 0.38472622478386165,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.4266,
      "mean_token_accuracy": 0.18962429016828536,
      "num_tokens": 9165613.0,
      "step": 4005
    },
    {
      "entropy": 5.491159725189209,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19434286206960677,
      "num_tokens": 9177884.0,
      "step": 4010
    },
    {
      "entropy": 5.613833093643189,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.5091,
      "mean_token_accuracy": 0.18056989163160325,
      "num_tokens": 9190041.0,
      "step": 4015
    },
    {
      "entropy": 5.628996181488037,
      "epoch": 0.3861671469740634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.3944,
      "mean_token_accuracy": 0.19238094687461854,
      "num_tokens": 9202390.0,
      "step": 4020
    },
    {
      "entropy": 5.512400102615357,
      "epoch": 0.3866474543707973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.3972,
      "mean_token_accuracy": 0.1906796544790268,
      "num_tokens": 9213437.0,
      "step": 4025
    },
    {
      "entropy": 5.599765205383301,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.4618,
      "mean_token_accuracy": 0.18272736817598342,
      "num_tokens": 9226310.0,
      "step": 4030
    },
    {
      "entropy": 5.603873920440674,
      "epoch": 0.38760806916426516,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.4077,
      "mean_token_accuracy": 0.19082859754562378,
      "num_tokens": 9236271.0,
      "step": 4035
    },
    {
      "entropy": 5.613268089294434,
      "epoch": 0.38808837656099904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.4624,
      "mean_token_accuracy": 0.19046298116445542,
      "num_tokens": 9247755.0,
      "step": 4040
    },
    {
      "entropy": 5.542822551727295,
      "epoch": 0.388568683957733,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.3169,
      "mean_token_accuracy": 0.20392679125070573,
      "num_tokens": 9259821.0,
      "step": 4045
    },
    {
      "entropy": 5.564394140243531,
      "epoch": 0.38904899135446686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18946493715047835,
      "num_tokens": 9270663.0,
      "step": 4050
    },
    {
      "entropy": 5.552659273147583,
      "epoch": 0.3895292987512008,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1895113468170166,
      "num_tokens": 9282588.0,
      "step": 4055
    },
    {
      "entropy": 5.5941455364227295,
      "epoch": 0.3900096061479347,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.4071,
      "mean_token_accuracy": 0.17811299264431,
      "num_tokens": 9293368.0,
      "step": 4060
    },
    {
      "entropy": 5.548636865615845,
      "epoch": 0.3904899135446686,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.468,
      "mean_token_accuracy": 0.1825200393795967,
      "num_tokens": 9305962.0,
      "step": 4065
    },
    {
      "entropy": 5.558157014846802,
      "epoch": 0.3909702209414025,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1954346388578415,
      "num_tokens": 9317955.0,
      "step": 4070
    },
    {
      "entropy": 5.549290418624878,
      "epoch": 0.3914505283381364,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.3593,
      "mean_token_accuracy": 0.19245800226926804,
      "num_tokens": 9329491.0,
      "step": 4075
    },
    {
      "entropy": 5.498097085952759,
      "epoch": 0.3919308357348703,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.4817,
      "mean_token_accuracy": 0.19573085010051727,
      "num_tokens": 9341612.0,
      "step": 4080
    },
    {
      "entropy": 5.5749123096466064,
      "epoch": 0.39241114313160425,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.4291,
      "mean_token_accuracy": 0.18815626055002213,
      "num_tokens": 9352302.0,
      "step": 4085
    },
    {
      "entropy": 5.589935445785523,
      "epoch": 0.3928914505283381,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.3586,
      "mean_token_accuracy": 0.19239330440759658,
      "num_tokens": 9363303.0,
      "step": 4090
    },
    {
      "entropy": 5.502812719345092,
      "epoch": 0.39337175792507206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.3286,
      "mean_token_accuracy": 0.18687490820884706,
      "num_tokens": 9375878.0,
      "step": 4095
    },
    {
      "entropy": 5.534584331512451,
      "epoch": 0.39385206532180594,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.3734,
      "mean_token_accuracy": 0.19118351340293885,
      "num_tokens": 9388342.0,
      "step": 4100
    },
    {
      "entropy": 5.500079584121704,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.2949,
      "mean_token_accuracy": 0.1912856638431549,
      "num_tokens": 9399047.0,
      "step": 4105
    },
    {
      "entropy": 5.587457847595215,
      "epoch": 0.39481268011527376,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.3811,
      "mean_token_accuracy": 0.19598662108182907,
      "num_tokens": 9410172.0,
      "step": 4110
    },
    {
      "entropy": 5.603793287277222,
      "epoch": 0.3952929875120077,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.5287,
      "mean_token_accuracy": 0.18236763179302215,
      "num_tokens": 9422287.0,
      "step": 4115
    },
    {
      "entropy": 5.639399194717408,
      "epoch": 0.3957732949087416,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.382,
      "mean_token_accuracy": 0.19403408318758011,
      "num_tokens": 9433709.0,
      "step": 4120
    },
    {
      "entropy": 5.453795385360718,
      "epoch": 0.3962536023054755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.2754,
      "mean_token_accuracy": 0.19953108578920364,
      "num_tokens": 9445138.0,
      "step": 4125
    },
    {
      "entropy": 5.477317523956299,
      "epoch": 0.3967339097022094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18944294303655623,
      "num_tokens": 9455263.0,
      "step": 4130
    },
    {
      "entropy": 5.614648723602295,
      "epoch": 0.39721421709894333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.3727,
      "mean_token_accuracy": 0.18981866538524628,
      "num_tokens": 9465491.0,
      "step": 4135
    },
    {
      "entropy": 5.482787322998047,
      "epoch": 0.3976945244956772,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.339,
      "mean_token_accuracy": 0.19033195674419404,
      "num_tokens": 9477601.0,
      "step": 4140
    },
    {
      "entropy": 5.545722436904907,
      "epoch": 0.39817483189241115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.3988,
      "mean_token_accuracy": 0.19300916343927382,
      "num_tokens": 9489385.0,
      "step": 4145
    },
    {
      "entropy": 5.52602744102478,
      "epoch": 0.39865513928914503,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.2736,
      "mean_token_accuracy": 0.1982199728488922,
      "num_tokens": 9501131.0,
      "step": 4150
    },
    {
      "entropy": 5.4960432052612305,
      "epoch": 0.39913544668587897,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498961495943777,
      "loss": 5.4284,
      "mean_token_accuracy": 0.19547935873270034,
      "num_tokens": 9513094.0,
      "step": 4155
    },
    {
      "entropy": 5.630834293365479,
      "epoch": 0.39961575408261285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.5256,
      "mean_token_accuracy": 0.1863490089774132,
      "num_tokens": 9524538.0,
      "step": 4160
    },
    {
      "entropy": 5.569158506393433,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.3489,
      "mean_token_accuracy": 0.20113594233989715,
      "num_tokens": 9535284.0,
      "step": 4165
    },
    {
      "entropy": 5.550960111618042,
      "epoch": 0.40057636887608067,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.4361,
      "mean_token_accuracy": 0.1874231815338135,
      "num_tokens": 9546472.0,
      "step": 4170
    },
    {
      "entropy": 5.565934753417968,
      "epoch": 0.4010566762728146,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18724920451641083,
      "num_tokens": 9558358.0,
      "step": 4175
    },
    {
      "entropy": 5.600048398971557,
      "epoch": 0.4015369836695485,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.4499,
      "mean_token_accuracy": 0.19195451587438583,
      "num_tokens": 9570653.0,
      "step": 4180
    },
    {
      "entropy": 5.564519691467285,
      "epoch": 0.4020172910662824,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.3972,
      "mean_token_accuracy": 0.19142527729272843,
      "num_tokens": 9582150.0,
      "step": 4185
    },
    {
      "entropy": 5.590189504623413,
      "epoch": 0.4024975984630163,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19532157182693483,
      "num_tokens": 9592462.0,
      "step": 4190
    },
    {
      "entropy": 5.4742389678955075,
      "epoch": 0.40297790585975024,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.3327,
      "mean_token_accuracy": 0.19161231219768524,
      "num_tokens": 9604027.0,
      "step": 4195
    },
    {
      "entropy": 5.4748578548431395,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19610172063112258,
      "num_tokens": 9616325.0,
      "step": 4200
    },
    {
      "entropy": 5.458920621871949,
      "epoch": 0.40393852065321806,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1975201338529587,
      "num_tokens": 9628133.0,
      "step": 4205
    },
    {
      "entropy": 5.664987993240357,
      "epoch": 0.404418828049952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.5003,
      "mean_token_accuracy": 0.18330966383218766,
      "num_tokens": 9639686.0,
      "step": 4210
    },
    {
      "entropy": 5.555954265594482,
      "epoch": 0.4048991354466859,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1899016246199608,
      "num_tokens": 9650222.0,
      "step": 4215
    },
    {
      "entropy": 5.458903455734253,
      "epoch": 0.4053794428434198,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.3048,
      "mean_token_accuracy": 0.20585627257823944,
      "num_tokens": 9661792.0,
      "step": 4220
    },
    {
      "entropy": 5.544176006317139,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.4053,
      "mean_token_accuracy": 0.18887956887483598,
      "num_tokens": 9673000.0,
      "step": 4225
    },
    {
      "entropy": 5.611917209625244,
      "epoch": 0.40634005763688763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.3492,
      "mean_token_accuracy": 0.19597309231758117,
      "num_tokens": 9685253.0,
      "step": 4230
    },
    {
      "entropy": 5.489803409576416,
      "epoch": 0.4068203650336215,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1868518963456154,
      "num_tokens": 9697245.0,
      "step": 4235
    },
    {
      "entropy": 5.496517181396484,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.3624,
      "mean_token_accuracy": 0.19268081486225128,
      "num_tokens": 9710095.0,
      "step": 4240
    },
    {
      "entropy": 5.51132984161377,
      "epoch": 0.40778097982708933,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498901431470218,
      "loss": 5.2435,
      "mean_token_accuracy": 0.20017803013324736,
      "num_tokens": 9721488.0,
      "step": 4245
    },
    {
      "entropy": 5.369005727767944,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1994133025407791,
      "num_tokens": 9733280.0,
      "step": 4250
    },
    {
      "entropy": 5.505516052246094,
      "epoch": 0.40874159462055715,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19074197262525558,
      "num_tokens": 9744514.0,
      "step": 4255
    },
    {
      "entropy": 5.514144372940064,
      "epoch": 0.4092219020172911,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.349,
      "mean_token_accuracy": 0.19237666875123977,
      "num_tokens": 9754931.0,
      "step": 4260
    },
    {
      "entropy": 5.476861238479614,
      "epoch": 0.40970220941402496,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19797009825706482,
      "num_tokens": 9765635.0,
      "step": 4265
    },
    {
      "entropy": 5.564598083496094,
      "epoch": 0.4101825168107589,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.4176,
      "mean_token_accuracy": 0.1858292818069458,
      "num_tokens": 9777512.0,
      "step": 4270
    },
    {
      "entropy": 5.581813335418701,
      "epoch": 0.4106628242074928,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.4259,
      "mean_token_accuracy": 0.182652647793293,
      "num_tokens": 9790128.0,
      "step": 4275
    },
    {
      "entropy": 5.5544164180755615,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498877618185287,
      "loss": 5.3703,
      "mean_token_accuracy": 0.19317726492881776,
      "num_tokens": 9802549.0,
      "step": 4280
    },
    {
      "entropy": 5.475317859649659,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19570045918226242,
      "num_tokens": 9813736.0,
      "step": 4285
    },
    {
      "entropy": 5.428925800323486,
      "epoch": 0.41210374639769454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.3157,
      "mean_token_accuracy": 0.19985667020082473,
      "num_tokens": 9823785.0,
      "step": 4290
    },
    {
      "entropy": 5.524678373336792,
      "epoch": 0.4125840537944284,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.269,
      "mean_token_accuracy": 0.1980641022324562,
      "num_tokens": 9834910.0,
      "step": 4295
    },
    {
      "entropy": 5.533667707443238,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1934407815337181,
      "num_tokens": 9845593.0,
      "step": 4300
    },
    {
      "entropy": 5.438078308105469,
      "epoch": 0.41354466858789624,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.2094,
      "mean_token_accuracy": 0.207307867705822,
      "num_tokens": 9856763.0,
      "step": 4305
    },
    {
      "entropy": 5.557668399810791,
      "epoch": 0.4140249759846302,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.4396,
      "mean_token_accuracy": 0.18502080589532852,
      "num_tokens": 9869528.0,
      "step": 4310
    },
    {
      "entropy": 5.517323637008667,
      "epoch": 0.41450528338136405,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.3325,
      "mean_token_accuracy": 0.19549246728420258,
      "num_tokens": 9881170.0,
      "step": 4315
    },
    {
      "entropy": 5.527880764007568,
      "epoch": 0.414985590778098,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.4255,
      "mean_token_accuracy": 0.18764246851205826,
      "num_tokens": 9893119.0,
      "step": 4320
    },
    {
      "entropy": 5.423438262939453,
      "epoch": 0.41546589817483187,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.2627,
      "mean_token_accuracy": 0.20306207984685898,
      "num_tokens": 9905339.0,
      "step": 4325
    },
    {
      "entropy": 5.555384874343872,
      "epoch": 0.4159462055715658,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.3633,
      "mean_token_accuracy": 0.19620103389024734,
      "num_tokens": 9917500.0,
      "step": 4330
    },
    {
      "entropy": 5.523201417922974,
      "epoch": 0.4164265129682997,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19124320894479752,
      "num_tokens": 9929667.0,
      "step": 4335
    },
    {
      "entropy": 5.538033056259155,
      "epoch": 0.4169068203650336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.3526,
      "mean_token_accuracy": 0.19107616543769837,
      "num_tokens": 9941102.0,
      "step": 4340
    },
    {
      "entropy": 5.46150393486023,
      "epoch": 0.4173871277617675,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2557,
      "mean_token_accuracy": 0.20461782515048982,
      "num_tokens": 9951986.0,
      "step": 4345
    },
    {
      "entropy": 5.464718580245972,
      "epoch": 0.41786743515850144,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2626,
      "mean_token_accuracy": 0.2002312421798706,
      "num_tokens": 9964486.0,
      "step": 4350
    },
    {
      "entropy": 5.49985671043396,
      "epoch": 0.4183477425552353,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.3613,
      "mean_token_accuracy": 0.1955868422985077,
      "num_tokens": 9975909.0,
      "step": 4355
    },
    {
      "entropy": 5.55556583404541,
      "epoch": 0.41882804995196926,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19315043240785598,
      "num_tokens": 9986951.0,
      "step": 4360
    },
    {
      "entropy": 5.574152660369873,
      "epoch": 0.41930835734870314,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.4691,
      "mean_token_accuracy": 0.18375526517629623,
      "num_tokens": 9999234.0,
      "step": 4365
    },
    {
      "entropy": 5.496746015548706,
      "epoch": 0.4197886647454371,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.3249,
      "mean_token_accuracy": 0.19960517138242723,
      "num_tokens": 10010178.0,
      "step": 4370
    },
    {
      "entropy": 5.469662141799927,
      "epoch": 0.420268972142171,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.2854,
      "mean_token_accuracy": 0.19531275182962418,
      "num_tokens": 10021183.0,
      "step": 4375
    },
    {
      "entropy": 5.641615295410157,
      "epoch": 0.4207492795389049,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17796630561351776,
      "num_tokens": 10033686.0,
      "step": 4380
    },
    {
      "entropy": 5.461948680877685,
      "epoch": 0.42122958693563883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.2841,
      "mean_token_accuracy": 0.19567082226276397,
      "num_tokens": 10045923.0,
      "step": 4385
    },
    {
      "entropy": 5.335393571853638,
      "epoch": 0.4217098943323727,
      "grad_norm": 0.859375,
      "learning_rate": 0.000498801118470758,
      "loss": 5.1912,
      "mean_token_accuracy": 0.20154503285884856,
      "num_tokens": 10057196.0,
      "step": 4390
    },
    {
      "entropy": 5.4671718120574955,
      "epoch": 0.42219020172910665,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498797581450484,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1907932788133621,
      "num_tokens": 10069655.0,
      "step": 4395
    },
    {
      "entropy": 5.550377655029297,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1976744294166565,
      "num_tokens": 10080876.0,
      "step": 4400
    },
    {
      "entropy": 5.534447097778321,
      "epoch": 0.42315081652257447,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.3552,
      "mean_token_accuracy": 0.19047154486179352,
      "num_tokens": 10091986.0,
      "step": 4405
    },
    {
      "entropy": 5.534992027282715,
      "epoch": 0.42363112391930835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.4398,
      "mean_token_accuracy": 0.19162497520446778,
      "num_tokens": 10104027.0,
      "step": 4410
    },
    {
      "entropy": 5.52318263053894,
      "epoch": 0.4241114313160423,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19698891639709473,
      "num_tokens": 10114665.0,
      "step": 4415
    },
    {
      "entropy": 5.497555065155029,
      "epoch": 0.42459173871277617,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.3206,
      "mean_token_accuracy": 0.19676625728607178,
      "num_tokens": 10126032.0,
      "step": 4420
    },
    {
      "entropy": 5.534703922271729,
      "epoch": 0.4250720461095101,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.3858,
      "mean_token_accuracy": 0.19109898060560226,
      "num_tokens": 10137256.0,
      "step": 4425
    },
    {
      "entropy": 5.452436637878418,
      "epoch": 0.425552353506244,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000498772676827552,
      "loss": 5.2682,
      "mean_token_accuracy": 0.19692862033843994,
      "num_tokens": 10149445.0,
      "step": 4430
    },
    {
      "entropy": 5.571733474731445,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.4037,
      "mean_token_accuracy": 0.18802779018878937,
      "num_tokens": 10161607.0,
      "step": 4435
    },
    {
      "entropy": 5.538624715805054,
      "epoch": 0.4265129682997118,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19259897768497466,
      "num_tokens": 10173184.0,
      "step": 4440
    },
    {
      "entropy": 5.47358283996582,
      "epoch": 0.42699327569644574,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19967445135116577,
      "num_tokens": 10184617.0,
      "step": 4445
    },
    {
      "entropy": 5.516195583343506,
      "epoch": 0.4274735830931796,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.3133,
      "mean_token_accuracy": 0.19398918002843857,
      "num_tokens": 10195608.0,
      "step": 4450
    },
    {
      "entropy": 5.499154186248779,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3918,
      "mean_token_accuracy": 0.19446780085563659,
      "num_tokens": 10206763.0,
      "step": 4455
    },
    {
      "entropy": 5.52622652053833,
      "epoch": 0.42843419788664744,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19333059340715408,
      "num_tokens": 10218714.0,
      "step": 4460
    },
    {
      "entropy": 5.5156656265258786,
      "epoch": 0.4289145052833814,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.3669,
      "mean_token_accuracy": 0.1908699557185173,
      "num_tokens": 10229902.0,
      "step": 4465
    },
    {
      "entropy": 5.496946096420288,
      "epoch": 0.42939481268011526,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.2409,
      "mean_token_accuracy": 0.19724130481481553,
      "num_tokens": 10241578.0,
      "step": 4470
    },
    {
      "entropy": 5.42222957611084,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.2035,
      "mean_token_accuracy": 0.20090758353471755,
      "num_tokens": 10252109.0,
      "step": 4475
    },
    {
      "entropy": 5.5068761825561525,
      "epoch": 0.4303554274735831,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.38,
      "mean_token_accuracy": 0.1923118144273758,
      "num_tokens": 10263357.0,
      "step": 4480
    },
    {
      "entropy": 5.550309705734253,
      "epoch": 0.430835734870317,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19260043650865555,
      "num_tokens": 10275536.0,
      "step": 4485
    },
    {
      "entropy": 5.575230693817138,
      "epoch": 0.4313160422670509,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.3461,
      "mean_token_accuracy": 0.20100529342889786,
      "num_tokens": 10286901.0,
      "step": 4490
    },
    {
      "entropy": 5.437625217437744,
      "epoch": 0.43179634966378483,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1957649528980255,
      "num_tokens": 10298961.0,
      "step": 4495
    },
    {
      "entropy": 5.524259567260742,
      "epoch": 0.4322766570605187,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.403,
      "mean_token_accuracy": 0.19641273021697997,
      "num_tokens": 10310911.0,
      "step": 4500
    },
    {
      "entropy": 5.567202568054199,
      "epoch": 0.43275696445725265,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.3775,
      "mean_token_accuracy": 0.1914454460144043,
      "num_tokens": 10322267.0,
      "step": 4505
    },
    {
      "entropy": 5.494447660446167,
      "epoch": 0.4332372718539865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.4018,
      "mean_token_accuracy": 0.1945730999112129,
      "num_tokens": 10335111.0,
      "step": 4510
    },
    {
      "entropy": 5.585152101516724,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.3466,
      "mean_token_accuracy": 0.19717937111854553,
      "num_tokens": 10345672.0,
      "step": 4515
    },
    {
      "entropy": 5.579340267181396,
      "epoch": 0.43419788664745435,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049870746748679,
      "loss": 5.3364,
      "mean_token_accuracy": 0.19359557032585145,
      "num_tokens": 10357369.0,
      "step": 4520
    },
    {
      "entropy": 5.46253399848938,
      "epoch": 0.4346781940441883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.3542,
      "mean_token_accuracy": 0.19162115156650544,
      "num_tokens": 10368842.0,
      "step": 4525
    },
    {
      "entropy": 5.506046867370605,
      "epoch": 0.43515850144092216,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19462748765945434,
      "num_tokens": 10380096.0,
      "step": 4530
    },
    {
      "entropy": 5.433768033981323,
      "epoch": 0.4356388088376561,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19543916881084442,
      "num_tokens": 10391996.0,
      "step": 4535
    },
    {
      "entropy": 5.506909990310669,
      "epoch": 0.43611911623439004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1950858324766159,
      "num_tokens": 10403607.0,
      "step": 4540
    },
    {
      "entropy": 5.500537729263305,
      "epoch": 0.4365994236311239,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.3027,
      "mean_token_accuracy": 0.19058295637369155,
      "num_tokens": 10413580.0,
      "step": 4545
    },
    {
      "entropy": 5.453525447845459,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19454682171344756,
      "num_tokens": 10426456.0,
      "step": 4550
    },
    {
      "entropy": 5.570721864700317,
      "epoch": 0.43756003842459174,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18605243414640427,
      "num_tokens": 10438869.0,
      "step": 4555
    },
    {
      "entropy": 5.467431879043579,
      "epoch": 0.43804034582132567,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.3617,
      "mean_token_accuracy": 0.18677616268396377,
      "num_tokens": 10450672.0,
      "step": 4560
    },
    {
      "entropy": 5.506100177764893,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.2719,
      "mean_token_accuracy": 0.19866655319929122,
      "num_tokens": 10461890.0,
      "step": 4565
    },
    {
      "entropy": 5.571147489547729,
      "epoch": 0.4390009606147935,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.4027,
      "mean_token_accuracy": 0.18525223433971405,
      "num_tokens": 10473866.0,
      "step": 4570
    },
    {
      "entropy": 5.531253671646118,
      "epoch": 0.43948126801152737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19253301173448562,
      "num_tokens": 10484889.0,
      "step": 4575
    },
    {
      "entropy": 5.478489875793457,
      "epoch": 0.4399615754082613,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.366,
      "mean_token_accuracy": 0.19603030383586884,
      "num_tokens": 10496108.0,
      "step": 4580
    },
    {
      "entropy": 5.481814002990722,
      "epoch": 0.4404418828049952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.3077,
      "mean_token_accuracy": 0.19383252114057542,
      "num_tokens": 10505472.0,
      "step": 4585
    },
    {
      "entropy": 5.359921264648437,
      "epoch": 0.4409221902017291,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2242,
      "mean_token_accuracy": 0.19945572316646576,
      "num_tokens": 10516954.0,
      "step": 4590
    },
    {
      "entropy": 5.584690189361572,
      "epoch": 0.441402497598463,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.341,
      "mean_token_accuracy": 0.19385933727025986,
      "num_tokens": 10528166.0,
      "step": 4595
    },
    {
      "entropy": 5.464456033706665,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.2583,
      "mean_token_accuracy": 0.198069229722023,
      "num_tokens": 10538112.0,
      "step": 4600
    },
    {
      "entropy": 5.505298757553101,
      "epoch": 0.4423631123919308,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.4021,
      "mean_token_accuracy": 0.19536336809396743,
      "num_tokens": 10549888.0,
      "step": 4605
    },
    {
      "entropy": 5.657861185073853,
      "epoch": 0.44284341978866476,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.4699,
      "mean_token_accuracy": 0.1838005244731903,
      "num_tokens": 10561710.0,
      "step": 4610
    },
    {
      "entropy": 5.473537921905518,
      "epoch": 0.44332372718539864,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.3241,
      "mean_token_accuracy": 0.19243624061346054,
      "num_tokens": 10574564.0,
      "step": 4615
    },
    {
      "entropy": 5.540151643753052,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.3188,
      "mean_token_accuracy": 0.19761135280132294,
      "num_tokens": 10586639.0,
      "step": 4620
    },
    {
      "entropy": 5.506497859954834,
      "epoch": 0.44428434197886646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.3144,
      "mean_token_accuracy": 0.1980729505419731,
      "num_tokens": 10598730.0,
      "step": 4625
    },
    {
      "entropy": 5.433584356307984,
      "epoch": 0.4447646493756004,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.2295,
      "mean_token_accuracy": 0.2066010907292366,
      "num_tokens": 10610352.0,
      "step": 4630
    },
    {
      "entropy": 5.467961835861206,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.3149,
      "mean_token_accuracy": 0.19723764508962632,
      "num_tokens": 10621951.0,
      "step": 4635
    },
    {
      "entropy": 5.48427186012268,
      "epoch": 0.4457252641690682,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498617904100496,
      "loss": 5.269,
      "mean_token_accuracy": 0.19730617105960846,
      "num_tokens": 10633207.0,
      "step": 4640
    },
    {
      "entropy": 5.5214195728302,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.3485,
      "mean_token_accuracy": 0.1888623684644699,
      "num_tokens": 10645853.0,
      "step": 4645
    },
    {
      "entropy": 5.466204071044922,
      "epoch": 0.44668587896253603,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.2731,
      "mean_token_accuracy": 0.197029510140419,
      "num_tokens": 10656892.0,
      "step": 4650
    },
    {
      "entropy": 5.430473184585571,
      "epoch": 0.4471661863592699,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19406193494796753,
      "num_tokens": 10670110.0,
      "step": 4655
    },
    {
      "entropy": 5.522889232635498,
      "epoch": 0.44764649375600385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.3824,
      "mean_token_accuracy": 0.19317276924848556,
      "num_tokens": 10681255.0,
      "step": 4660
    },
    {
      "entropy": 5.506927394866944,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.4456,
      "mean_token_accuracy": 0.19067862927913665,
      "num_tokens": 10692631.0,
      "step": 4665
    },
    {
      "entropy": 5.498530721664428,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.2764,
      "mean_token_accuracy": 0.19727545231580734,
      "num_tokens": 10704880.0,
      "step": 4670
    },
    {
      "entropy": 5.481644058227539,
      "epoch": 0.44908741594620555,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.267,
      "mean_token_accuracy": 0.19885830134153365,
      "num_tokens": 10715561.0,
      "step": 4675
    },
    {
      "entropy": 5.432435321807861,
      "epoch": 0.4495677233429395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.2767,
      "mean_token_accuracy": 0.19294027239084244,
      "num_tokens": 10727232.0,
      "step": 4680
    },
    {
      "entropy": 5.450063991546631,
      "epoch": 0.45004803073967337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1977899894118309,
      "num_tokens": 10739404.0,
      "step": 4685
    },
    {
      "entropy": 5.486606788635254,
      "epoch": 0.4505283381364073,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.3282,
      "mean_token_accuracy": 0.19158132523298263,
      "num_tokens": 10751948.0,
      "step": 4690
    },
    {
      "entropy": 5.4498594284057615,
      "epoch": 0.4510086455331412,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.2983,
      "mean_token_accuracy": 0.19393019676208495,
      "num_tokens": 10764611.0,
      "step": 4695
    },
    {
      "entropy": 5.533370399475098,
      "epoch": 0.4514889529298751,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.3189,
      "mean_token_accuracy": 0.19145917296409606,
      "num_tokens": 10775954.0,
      "step": 4700
    },
    {
      "entropy": 5.4836106300354,
      "epoch": 0.45196926032660906,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.3281,
      "mean_token_accuracy": 0.18578007370233535,
      "num_tokens": 10788723.0,
      "step": 4705
    },
    {
      "entropy": 5.504494524002075,
      "epoch": 0.45244956772334294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.3192,
      "mean_token_accuracy": 0.19870994240045547,
      "num_tokens": 10799879.0,
      "step": 4710
    },
    {
      "entropy": 5.444184827804565,
      "epoch": 0.4529298751200769,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.278,
      "mean_token_accuracy": 0.19549098014831542,
      "num_tokens": 10811838.0,
      "step": 4715
    },
    {
      "entropy": 5.476307773590088,
      "epoch": 0.45341018251681076,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.2976,
      "mean_token_accuracy": 0.19534050822257995,
      "num_tokens": 10822910.0,
      "step": 4720
    },
    {
      "entropy": 5.545818233489991,
      "epoch": 0.4538904899135447,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.3536,
      "mean_token_accuracy": 0.19093685001134872,
      "num_tokens": 10835344.0,
      "step": 4725
    },
    {
      "entropy": 5.473781394958496,
      "epoch": 0.4543707973102786,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19793201237916946,
      "num_tokens": 10846043.0,
      "step": 4730
    },
    {
      "entropy": 5.483222532272339,
      "epoch": 0.4548511047070125,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498544878867698,
      "loss": 5.2471,
      "mean_token_accuracy": 0.20124554485082627,
      "num_tokens": 10857783.0,
      "step": 4735
    },
    {
      "entropy": 5.502796602249146,
      "epoch": 0.4553314121037464,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19286340475082397,
      "num_tokens": 10870527.0,
      "step": 4740
    },
    {
      "entropy": 5.451869821548462,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1995761588215828,
      "num_tokens": 10882285.0,
      "step": 4745
    },
    {
      "entropy": 5.519151926040649,
      "epoch": 0.4562920268972142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.3397,
      "mean_token_accuracy": 0.19749111980199813,
      "num_tokens": 10895319.0,
      "step": 4750
    },
    {
      "entropy": 5.546896362304688,
      "epoch": 0.45677233429394815,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.3178,
      "mean_token_accuracy": 0.1997792139649391,
      "num_tokens": 10906253.0,
      "step": 4755
    },
    {
      "entropy": 5.323689603805542,
      "epoch": 0.457252641690682,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.2064,
      "mean_token_accuracy": 0.20633554905653,
      "num_tokens": 10918197.0,
      "step": 4760
    },
    {
      "entropy": 5.48868842124939,
      "epoch": 0.45773294908741596,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.4051,
      "mean_token_accuracy": 0.18659979701042176,
      "num_tokens": 10930771.0,
      "step": 4765
    },
    {
      "entropy": 5.524543571472168,
      "epoch": 0.45821325648414984,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19515040069818496,
      "num_tokens": 10942759.0,
      "step": 4770
    },
    {
      "entropy": 5.671936845779419,
      "epoch": 0.4586935638808838,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.5023,
      "mean_token_accuracy": 0.1835600718855858,
      "num_tokens": 10953557.0,
      "step": 4775
    },
    {
      "entropy": 5.4718035697937015,
      "epoch": 0.45917387127761766,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.2753,
      "mean_token_accuracy": 0.20337168574333192,
      "num_tokens": 10963142.0,
      "step": 4780
    },
    {
      "entropy": 5.392521715164184,
      "epoch": 0.4596541786743516,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.2352,
      "mean_token_accuracy": 0.2007603958249092,
      "num_tokens": 10974321.0,
      "step": 4785
    },
    {
      "entropy": 5.5639253616333,
      "epoch": 0.4601344860710855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.439,
      "mean_token_accuracy": 0.1905608966946602,
      "num_tokens": 10985524.0,
      "step": 4790
    },
    {
      "entropy": 5.547641324996948,
      "epoch": 0.4606147934678194,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.3173,
      "mean_token_accuracy": 0.19772396385669708,
      "num_tokens": 10997606.0,
      "step": 4795
    },
    {
      "entropy": 5.467640161514282,
      "epoch": 0.4610951008645533,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.3181,
      "mean_token_accuracy": 0.1880475327372551,
      "num_tokens": 11010692.0,
      "step": 4800
    },
    {
      "entropy": 5.430249834060669,
      "epoch": 0.46157540826128723,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.2577,
      "mean_token_accuracy": 0.2035054549574852,
      "num_tokens": 11022517.0,
      "step": 4805
    },
    {
      "entropy": 5.480600738525391,
      "epoch": 0.4620557156580211,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.2538,
      "mean_token_accuracy": 0.20069904923439025,
      "num_tokens": 11033141.0,
      "step": 4810
    },
    {
      "entropy": 5.440704441070556,
      "epoch": 0.46253602305475505,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.2587,
      "mean_token_accuracy": 0.1992591381072998,
      "num_tokens": 11044619.0,
      "step": 4815
    },
    {
      "entropy": 5.529746055603027,
      "epoch": 0.46301633045148893,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.3095,
      "mean_token_accuracy": 0.1996643826365471,
      "num_tokens": 11055934.0,
      "step": 4820
    },
    {
      "entropy": 5.456304121017456,
      "epoch": 0.46349663784822287,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.1932,
      "mean_token_accuracy": 0.20430234223604202,
      "num_tokens": 11066203.0,
      "step": 4825
    },
    {
      "entropy": 5.294573020935059,
      "epoch": 0.46397694524495675,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.1439,
      "mean_token_accuracy": 0.20539496690034867,
      "num_tokens": 11079664.0,
      "step": 4830
    },
    {
      "entropy": 5.391605186462402,
      "epoch": 0.4644572526416907,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498465986095997,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20457594841718674,
      "num_tokens": 11091186.0,
      "step": 4835
    },
    {
      "entropy": 5.495221424102783,
      "epoch": 0.46493756003842457,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.3313,
      "mean_token_accuracy": 0.19135818332433702,
      "num_tokens": 11102710.0,
      "step": 4840
    },
    {
      "entropy": 5.393061113357544,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498457982665893,
      "loss": 5.2222,
      "mean_token_accuracy": 0.20273594111204146,
      "num_tokens": 11114746.0,
      "step": 4845
    },
    {
      "entropy": 5.48639497756958,
      "epoch": 0.4658981748318924,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19519798308610917,
      "num_tokens": 11126572.0,
      "step": 4850
    },
    {
      "entropy": 5.427862167358398,
      "epoch": 0.4663784822286263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.1845,
      "mean_token_accuracy": 0.2038702443242073,
      "num_tokens": 11137830.0,
      "step": 4855
    },
    {
      "entropy": 5.402318859100342,
      "epoch": 0.4668587896253602,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498445938610739,
      "loss": 5.2247,
      "mean_token_accuracy": 0.1985754981637001,
      "num_tokens": 11148860.0,
      "step": 4860
    },
    {
      "entropy": 5.510347509384156,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.3408,
      "mean_token_accuracy": 0.18952501863241195,
      "num_tokens": 11161311.0,
      "step": 4865
    },
    {
      "entropy": 5.4862135887146,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.2453,
      "mean_token_accuracy": 0.19657314866781234,
      "num_tokens": 11173124.0,
      "step": 4870
    },
    {
      "entropy": 5.405213022232056,
      "epoch": 0.46829971181556196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1948263332247734,
      "num_tokens": 11184879.0,
      "step": 4875
    },
    {
      "entropy": 5.519789314270019,
      "epoch": 0.4687800192122959,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.3156,
      "mean_token_accuracy": 0.19586896896362305,
      "num_tokens": 11196243.0,
      "step": 4880
    },
    {
      "entropy": 5.473620462417602,
      "epoch": 0.4692603266090298,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498425761431707,
      "loss": 5.1839,
      "mean_token_accuracy": 0.2131851315498352,
      "num_tokens": 11207485.0,
      "step": 4885
    },
    {
      "entropy": 5.417775678634643,
      "epoch": 0.4697406340057637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498421710433677,
      "loss": 5.2255,
      "mean_token_accuracy": 0.1948077842593193,
      "num_tokens": 11219891.0,
      "step": 4890
    },
    {
      "entropy": 5.54314079284668,
      "epoch": 0.4702209414024976,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1956147536635399,
      "num_tokens": 11231329.0,
      "step": 4895
    },
    {
      "entropy": 5.435543823242187,
      "epoch": 0.47070124879923153,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.2117,
      "mean_token_accuracy": 0.20288991183042526,
      "num_tokens": 11241367.0,
      "step": 4900
    },
    {
      "entropy": 5.44714093208313,
      "epoch": 0.4711815561959654,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.2858,
      "mean_token_accuracy": 0.19364715069532396,
      "num_tokens": 11254532.0,
      "step": 4905
    },
    {
      "entropy": 5.544944953918457,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.3768,
      "mean_token_accuracy": 0.1902521938085556,
      "num_tokens": 11265223.0,
      "step": 4910
    },
    {
      "entropy": 5.443441390991211,
      "epoch": 0.47214217098943323,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3013,
      "mean_token_accuracy": 0.19836317002773285,
      "num_tokens": 11278228.0,
      "step": 4915
    },
    {
      "entropy": 5.577527093887329,
      "epoch": 0.47262247838616717,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498397295517835,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1912369191646576,
      "num_tokens": 11289654.0,
      "step": 4920
    },
    {
      "entropy": 5.449563312530517,
      "epoch": 0.47310278578290105,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20263250917196274,
      "num_tokens": 11301911.0,
      "step": 4925
    },
    {
      "entropy": 5.550692844390869,
      "epoch": 0.473583093179635,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19057601392269136,
      "num_tokens": 11312945.0,
      "step": 4930
    },
    {
      "entropy": 5.617296600341797,
      "epoch": 0.47406340057636887,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.4156,
      "mean_token_accuracy": 0.184731362760067,
      "num_tokens": 11324126.0,
      "step": 4935
    },
    {
      "entropy": 5.5141273021698,
      "epoch": 0.4745437079731028,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1906953737139702,
      "num_tokens": 11336395.0,
      "step": 4940
    },
    {
      "entropy": 5.499377822875976,
      "epoch": 0.4750240153698367,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19905748665332795,
      "num_tokens": 11347387.0,
      "step": 4945
    },
    {
      "entropy": 5.396799373626709,
      "epoch": 0.4755043227665706,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.2422,
      "mean_token_accuracy": 0.20498072654008864,
      "num_tokens": 11358467.0,
      "step": 4950
    },
    {
      "entropy": 5.49179253578186,
      "epoch": 0.4759846301633045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.4096,
      "mean_token_accuracy": 0.19101459980010987,
      "num_tokens": 11370322.0,
      "step": 4955
    },
    {
      "entropy": 5.400127029418945,
      "epoch": 0.47646493756003844,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.2078,
      "mean_token_accuracy": 0.20537745207548141,
      "num_tokens": 11381719.0,
      "step": 4960
    },
    {
      "entropy": 5.4417760372161865,
      "epoch": 0.4769452449567723,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.2271,
      "mean_token_accuracy": 0.20112286359071732,
      "num_tokens": 11393561.0,
      "step": 4965
    },
    {
      "entropy": 5.495945167541504,
      "epoch": 0.47742555235350626,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.3273,
      "mean_token_accuracy": 0.1953832298517227,
      "num_tokens": 11405411.0,
      "step": 4970
    },
    {
      "entropy": 5.537110948562622,
      "epoch": 0.47790585975024014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498352049904511,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1894658625125885,
      "num_tokens": 11417419.0,
      "step": 4975
    },
    {
      "entropy": 5.492123174667358,
      "epoch": 0.4783861671469741,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19503100365400314,
      "num_tokens": 11428145.0,
      "step": 4980
    },
    {
      "entropy": 5.382611560821533,
      "epoch": 0.47886647454370795,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.1931,
      "mean_token_accuracy": 0.20153277218341828,
      "num_tokens": 11439224.0,
      "step": 4985
    },
    {
      "entropy": 5.459662389755249,
      "epoch": 0.4793467819404419,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19697522670030593,
      "num_tokens": 11451731.0,
      "step": 4990
    },
    {
      "entropy": 5.489697217941284,
      "epoch": 0.47982708933717577,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.2723,
      "mean_token_accuracy": 0.20329660624265672,
      "num_tokens": 11463058.0,
      "step": 4995
    },
    {
      "entropy": 5.602862405776977,
      "epoch": 0.4803073967339097,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.3766,
      "mean_token_accuracy": 0.19309359192848205,
      "num_tokens": 11472618.0,
      "step": 5000
    },
    {
      "entropy": 5.436448240280152,
      "epoch": 0.4807877041306436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19574105590581894,
      "num_tokens": 11483114.0,
      "step": 5005
    },
    {
      "entropy": 5.403754663467407,
      "epoch": 0.4812680115273775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19807563424110414,
      "num_tokens": 11494281.0,
      "step": 5010
    },
    {
      "entropy": 5.481242561340332,
      "epoch": 0.4817483189241114,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.3145,
      "mean_token_accuracy": 0.19510846138000487,
      "num_tokens": 11505962.0,
      "step": 5015
    },
    {
      "entropy": 5.499330186843872,
      "epoch": 0.48222862632084534,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.3657,
      "mean_token_accuracy": 0.18856551945209504,
      "num_tokens": 11517558.0,
      "step": 5020
    },
    {
      "entropy": 5.483879709243775,
      "epoch": 0.4827089337175792,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19289948493242265,
      "num_tokens": 11529184.0,
      "step": 5025
    },
    {
      "entropy": 5.484569931030274,
      "epoch": 0.48318924111431316,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498306176837196,
      "loss": 5.2893,
      "mean_token_accuracy": 0.18928477317094802,
      "num_tokens": 11540727.0,
      "step": 5030
    },
    {
      "entropy": 5.440706968307495,
      "epoch": 0.48366954851104704,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.222,
      "mean_token_accuracy": 0.19865260273218155,
      "num_tokens": 11551510.0,
      "step": 5035
    },
    {
      "entropy": 5.502947568893433,
      "epoch": 0.484149855907781,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.353,
      "mean_token_accuracy": 0.19847064167261125,
      "num_tokens": 11563203.0,
      "step": 5040
    },
    {
      "entropy": 5.40673451423645,
      "epoch": 0.4846301633045149,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.215,
      "mean_token_accuracy": 0.2036955922842026,
      "num_tokens": 11576013.0,
      "step": 5045
    },
    {
      "entropy": 5.325626945495605,
      "epoch": 0.4851104707012488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.1837,
      "mean_token_accuracy": 0.21459532976150514,
      "num_tokens": 11587535.0,
      "step": 5050
    },
    {
      "entropy": 5.416988277435303,
      "epoch": 0.48559077809798273,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20377931892871856,
      "num_tokens": 11598487.0,
      "step": 5055
    },
    {
      "entropy": 5.380494689941406,
      "epoch": 0.4860710854947166,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.1061,
      "mean_token_accuracy": 0.21789742708206178,
      "num_tokens": 11609457.0,
      "step": 5060
    },
    {
      "entropy": 5.379251003265381,
      "epoch": 0.48655139289145055,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.1703,
      "mean_token_accuracy": 0.206756691634655,
      "num_tokens": 11620251.0,
      "step": 5065
    },
    {
      "entropy": 5.387517213821411,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20464167147874832,
      "num_tokens": 11630956.0,
      "step": 5070
    },
    {
      "entropy": 5.436505603790283,
      "epoch": 0.48751200768491837,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19775371849536896,
      "num_tokens": 11642560.0,
      "step": 5075
    },
    {
      "entropy": 5.353562116622925,
      "epoch": 0.48799231508165225,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.1914,
      "mean_token_accuracy": 0.2018374592065811,
      "num_tokens": 11654050.0,
      "step": 5080
    },
    {
      "entropy": 5.371949672698975,
      "epoch": 0.4884726224783862,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.3523,
      "mean_token_accuracy": 0.1960517406463623,
      "num_tokens": 11664947.0,
      "step": 5085
    },
    {
      "entropy": 5.506836891174316,
      "epoch": 0.48895292987512007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.2498,
      "mean_token_accuracy": 0.19737753570079802,
      "num_tokens": 11676927.0,
      "step": 5090
    },
    {
      "entropy": 5.56851544380188,
      "epoch": 0.489433237271854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3608,
      "mean_token_accuracy": 0.18899684697389602,
      "num_tokens": 11688573.0,
      "step": 5095
    },
    {
      "entropy": 5.375510406494141,
      "epoch": 0.4899135446685879,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.2578,
      "mean_token_accuracy": 0.19423613399267198,
      "num_tokens": 11698704.0,
      "step": 5100
    },
    {
      "entropy": 5.374054288864135,
      "epoch": 0.4903938520653218,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.1853,
      "mean_token_accuracy": 0.2054339423775673,
      "num_tokens": 11709466.0,
      "step": 5105
    },
    {
      "entropy": 5.459903144836426,
      "epoch": 0.4908741594620557,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.2851,
      "mean_token_accuracy": 0.19726966619491576,
      "num_tokens": 11721120.0,
      "step": 5110
    },
    {
      "entropy": 5.4925140857696535,
      "epoch": 0.49135446685878964,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.2708,
      "mean_token_accuracy": 0.1935970291495323,
      "num_tokens": 11732120.0,
      "step": 5115
    },
    {
      "entropy": 5.387935447692871,
      "epoch": 0.4918347742555235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.1585,
      "mean_token_accuracy": 0.204207181930542,
      "num_tokens": 11743501.0,
      "step": 5120
    },
    {
      "entropy": 5.4283873558044435,
      "epoch": 0.49231508165225746,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.2461,
      "mean_token_accuracy": 0.20177514851093292,
      "num_tokens": 11755051.0,
      "step": 5125
    },
    {
      "entropy": 5.521465635299682,
      "epoch": 0.49279538904899134,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.3923,
      "mean_token_accuracy": 0.18937383741140365,
      "num_tokens": 11767663.0,
      "step": 5130
    },
    {
      "entropy": 5.450193262100219,
      "epoch": 0.4932756964457253,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.256,
      "mean_token_accuracy": 0.20313976407051088,
      "num_tokens": 11778828.0,
      "step": 5135
    },
    {
      "entropy": 5.538891410827636,
      "epoch": 0.49375600384245916,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.4355,
      "mean_token_accuracy": 0.18950497061014177,
      "num_tokens": 11790654.0,
      "step": 5140
    },
    {
      "entropy": 5.489221286773682,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.2177,
      "mean_token_accuracy": 0.19754690080881118,
      "num_tokens": 11801489.0,
      "step": 5145
    },
    {
      "entropy": 5.370436906814575,
      "epoch": 0.494716618635927,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.2229,
      "mean_token_accuracy": 0.2016189604997635,
      "num_tokens": 11813818.0,
      "step": 5150
    },
    {
      "entropy": 5.386388111114502,
      "epoch": 0.4951969260326609,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.2316,
      "mean_token_accuracy": 0.19863317310810089,
      "num_tokens": 11824814.0,
      "step": 5155
    },
    {
      "entropy": 5.500951671600342,
      "epoch": 0.4956772334293948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.2708,
      "mean_token_accuracy": 0.19523157477378844,
      "num_tokens": 11836252.0,
      "step": 5160
    },
    {
      "entropy": 5.4199676513671875,
      "epoch": 0.49615754082612873,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19706585705280305,
      "num_tokens": 11847715.0,
      "step": 5165
    },
    {
      "entropy": 5.309402847290039,
      "epoch": 0.4966378482228626,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.2058,
      "mean_token_accuracy": 0.20216697901487352,
      "num_tokens": 11860309.0,
      "step": 5170
    },
    {
      "entropy": 5.3990058422088625,
      "epoch": 0.49711815561959655,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20591772198677064,
      "num_tokens": 11872569.0,
      "step": 5175
    },
    {
      "entropy": 5.411104297637939,
      "epoch": 0.49759846301633043,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.1837,
      "mean_token_accuracy": 0.2038358524441719,
      "num_tokens": 11884778.0,
      "step": 5180
    },
    {
      "entropy": 5.4573859691619875,
      "epoch": 0.49807877041306436,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.279,
      "mean_token_accuracy": 0.20060150176286698,
      "num_tokens": 11897324.0,
      "step": 5185
    },
    {
      "entropy": 5.510555553436279,
      "epoch": 0.49855907780979825,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.297,
      "mean_token_accuracy": 0.19576158821582795,
      "num_tokens": 11909145.0,
      "step": 5190
    },
    {
      "entropy": 5.4124188899993895,
      "epoch": 0.4990393852065322,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.1992,
      "mean_token_accuracy": 0.20338936448097228,
      "num_tokens": 11921021.0,
      "step": 5195
    },
    {
      "entropy": 5.549371528625488,
      "epoch": 0.49951969260326606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.3559,
      "mean_token_accuracy": 0.19443730115890503,
      "num_tokens": 11932565.0,
      "step": 5200
    },
    {
      "entropy": 5.3739917278289795,
      "epoch": 0.5,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498156044185496,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19840114265680314,
      "num_tokens": 11943225.0,
      "step": 5205
    },
    {
      "entropy": 5.454575777053833,
      "epoch": 0.5004803073967339,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.2702,
      "mean_token_accuracy": 0.2014799103140831,
      "num_tokens": 11954821.0,
      "step": 5210
    },
    {
      "entropy": 5.45289101600647,
      "epoch": 0.5009606147934679,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.2109,
      "mean_token_accuracy": 0.2056070551276207,
      "num_tokens": 11966090.0,
      "step": 5215
    },
    {
      "entropy": 5.432686042785645,
      "epoch": 0.5014409221902018,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19887453466653823,
      "num_tokens": 11977410.0,
      "step": 5220
    },
    {
      "entropy": 5.431569290161133,
      "epoch": 0.5019212295869356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19690183252096177,
      "num_tokens": 11989119.0,
      "step": 5225
    },
    {
      "entropy": 5.544884872436524,
      "epoch": 0.5024015369836695,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.3326,
      "mean_token_accuracy": 0.1929663226008415,
      "num_tokens": 12000165.0,
      "step": 5230
    },
    {
      "entropy": 5.433797931671142,
      "epoch": 0.5028818443804035,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1939511552453041,
      "num_tokens": 12012118.0,
      "step": 5235
    },
    {
      "entropy": 5.483579301834107,
      "epoch": 0.5033621517771374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.3412,
      "mean_token_accuracy": 0.19039848893880845,
      "num_tokens": 12023432.0,
      "step": 5240
    },
    {
      "entropy": 5.47205867767334,
      "epoch": 0.5038424591738713,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.24,
      "mean_token_accuracy": 0.20507996380329133,
      "num_tokens": 12034323.0,
      "step": 5245
    },
    {
      "entropy": 5.375516033172607,
      "epoch": 0.5043227665706052,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1988113269209862,
      "num_tokens": 12045532.0,
      "step": 5250
    },
    {
      "entropy": 5.480970144271851,
      "epoch": 0.5048030739673391,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.2675,
      "mean_token_accuracy": 0.1947346433997154,
      "num_tokens": 12057346.0,
      "step": 5255
    },
    {
      "entropy": 5.3928892612457275,
      "epoch": 0.505283381364073,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.2072,
      "mean_token_accuracy": 0.2055114433169365,
      "num_tokens": 12068670.0,
      "step": 5260
    },
    {
      "entropy": 5.462854099273682,
      "epoch": 0.5057636887608069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.3435,
      "mean_token_accuracy": 0.1926040083169937,
      "num_tokens": 12079664.0,
      "step": 5265
    },
    {
      "entropy": 5.479207324981689,
      "epoch": 0.5062439961575408,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20990796238183976,
      "num_tokens": 12089984.0,
      "step": 5270
    },
    {
      "entropy": 5.366336822509766,
      "epoch": 0.5067243035542748,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498094213793741,
      "loss": 5.2088,
      "mean_token_accuracy": 0.20466214418411255,
      "num_tokens": 12101182.0,
      "step": 5275
    },
    {
      "entropy": 5.498898315429687,
      "epoch": 0.5072046109510087,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498089758486952,
      "loss": 5.28,
      "mean_token_accuracy": 0.1944056808948517,
      "num_tokens": 12112002.0,
      "step": 5280
    },
    {
      "entropy": 5.509644031524658,
      "epoch": 0.5076849183477425,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.3716,
      "mean_token_accuracy": 0.18642902821302415,
      "num_tokens": 12124194.0,
      "step": 5285
    },
    {
      "entropy": 5.428500461578369,
      "epoch": 0.5081652257444764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.287,
      "mean_token_accuracy": 0.20239519625902175,
      "num_tokens": 12133966.0,
      "step": 5290
    },
    {
      "entropy": 5.440308284759522,
      "epoch": 0.5086455331412104,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.2319,
      "mean_token_accuracy": 0.20603344440460206,
      "num_tokens": 12145643.0,
      "step": 5295
    },
    {
      "entropy": 5.412059211730957,
      "epoch": 0.5091258405379443,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1935652792453766,
      "num_tokens": 12156804.0,
      "step": 5300
    },
    {
      "entropy": 5.4437727451324465,
      "epoch": 0.5096061479346782,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19508322924375535,
      "num_tokens": 12168700.0,
      "step": 5305
    },
    {
      "entropy": 5.47620415687561,
      "epoch": 0.5100864553314121,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049806291788073,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19422244131565095,
      "num_tokens": 12181050.0,
      "step": 5310
    },
    {
      "entropy": 5.502141284942627,
      "epoch": 0.510566762728146,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.3387,
      "mean_token_accuracy": 0.19137953519821166,
      "num_tokens": 12192001.0,
      "step": 5315
    },
    {
      "entropy": 5.551894998550415,
      "epoch": 0.5110470701248799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1882469743490219,
      "num_tokens": 12202436.0,
      "step": 5320
    },
    {
      "entropy": 5.434849834442138,
      "epoch": 0.5115273775216138,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.2412,
      "mean_token_accuracy": 0.2006586894392967,
      "num_tokens": 12214454.0,
      "step": 5325
    },
    {
      "entropy": 5.4280845642089846,
      "epoch": 0.5120076849183477,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.2545,
      "mean_token_accuracy": 0.20265544056892396,
      "num_tokens": 12225924.0,
      "step": 5330
    },
    {
      "entropy": 5.50420036315918,
      "epoch": 0.5124879923150817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049804040828932,
      "loss": 5.2618,
      "mean_token_accuracy": 0.2032623529434204,
      "num_tokens": 12236456.0,
      "step": 5335
    },
    {
      "entropy": 5.463288831710815,
      "epoch": 0.5129682997118156,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.2693,
      "mean_token_accuracy": 0.20118008106946944,
      "num_tokens": 12247719.0,
      "step": 5340
    },
    {
      "entropy": 5.4230491638183596,
      "epoch": 0.5134486071085494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.3235,
      "mean_token_accuracy": 0.19510107785463332,
      "num_tokens": 12259141.0,
      "step": 5345
    },
    {
      "entropy": 5.561901187896728,
      "epoch": 0.5139289145052833,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.3712,
      "mean_token_accuracy": 0.19242959171533586,
      "num_tokens": 12269748.0,
      "step": 5350
    },
    {
      "entropy": 5.478623867034912,
      "epoch": 0.5144092219020173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.3164,
      "mean_token_accuracy": 0.20066264569759368,
      "num_tokens": 12281456.0,
      "step": 5355
    },
    {
      "entropy": 5.458214664459229,
      "epoch": 0.5148895292987512,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.295,
      "mean_token_accuracy": 0.19240615367889405,
      "num_tokens": 12293763.0,
      "step": 5360
    },
    {
      "entropy": 5.428730916976929,
      "epoch": 0.5153698366954851,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20714261382818222,
      "num_tokens": 12305398.0,
      "step": 5365
    },
    {
      "entropy": 5.371615886688232,
      "epoch": 0.515850144092219,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.204,
      "mean_token_accuracy": 0.19764825254678725,
      "num_tokens": 12316582.0,
      "step": 5370
    },
    {
      "entropy": 5.510064649581909,
      "epoch": 0.516330451488953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19602434784173967,
      "num_tokens": 12328463.0,
      "step": 5375
    },
    {
      "entropy": 5.474542284011841,
      "epoch": 0.5168107588856868,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.3055,
      "mean_token_accuracy": 0.19989489912986755,
      "num_tokens": 12341040.0,
      "step": 5380
    },
    {
      "entropy": 5.3581665515899655,
      "epoch": 0.5172910662824207,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.1985,
      "mean_token_accuracy": 0.2028449460864067,
      "num_tokens": 12353024.0,
      "step": 5385
    },
    {
      "entropy": 5.40923433303833,
      "epoch": 0.5177713736791547,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.2095,
      "mean_token_accuracy": 0.2023212730884552,
      "num_tokens": 12366100.0,
      "step": 5390
    },
    {
      "entropy": 5.3701718807220455,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.2133,
      "mean_token_accuracy": 0.20770515501499176,
      "num_tokens": 12377829.0,
      "step": 5395
    },
    {
      "entropy": 5.512105274200439,
      "epoch": 0.5187319884726225,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19796691536903382,
      "num_tokens": 12390830.0,
      "step": 5400
    },
    {
      "entropy": 5.374357843399048,
      "epoch": 0.5192122958693564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.1835,
      "mean_token_accuracy": 0.20305409133434296,
      "num_tokens": 12401642.0,
      "step": 5405
    },
    {
      "entropy": 5.429293298721314,
      "epoch": 0.5196926032660903,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1978478580713272,
      "num_tokens": 12411653.0,
      "step": 5410
    },
    {
      "entropy": 5.46315484046936,
      "epoch": 0.5201729106628242,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497967507701269,
      "loss": 5.1935,
      "mean_token_accuracy": 0.2083654746413231,
      "num_tokens": 12422891.0,
      "step": 5415
    },
    {
      "entropy": 5.46509952545166,
      "epoch": 0.5206532180595581,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.2766,
      "mean_token_accuracy": 0.20137949436903,
      "num_tokens": 12434190.0,
      "step": 5420
    },
    {
      "entropy": 5.4995557308197025,
      "epoch": 0.521133525456292,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.3395,
      "mean_token_accuracy": 0.19757068753242493,
      "num_tokens": 12445796.0,
      "step": 5425
    },
    {
      "entropy": 5.561598253250122,
      "epoch": 0.521613832853026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1920577585697174,
      "num_tokens": 12456212.0,
      "step": 5430
    },
    {
      "entropy": 5.361889886856079,
      "epoch": 0.5220941402497599,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497949075471522,
      "loss": 5.1194,
      "mean_token_accuracy": 0.2050475671887398,
      "num_tokens": 12467871.0,
      "step": 5435
    },
    {
      "entropy": 5.452909755706787,
      "epoch": 0.5225744476464937,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.2174,
      "mean_token_accuracy": 0.1954452022910118,
      "num_tokens": 12478626.0,
      "step": 5440
    },
    {
      "entropy": 5.441848993301392,
      "epoch": 0.5230547550432276,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497939828298927,
      "loss": 5.303,
      "mean_token_accuracy": 0.19818771183490752,
      "num_tokens": 12491487.0,
      "step": 5445
    },
    {
      "entropy": 5.503297567367554,
      "epoch": 0.5235350624399616,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1951201155781746,
      "num_tokens": 12503240.0,
      "step": 5450
    },
    {
      "entropy": 5.464956569671631,
      "epoch": 0.5240153698366955,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19913747459650039,
      "num_tokens": 12513860.0,
      "step": 5455
    },
    {
      "entropy": 5.38545298576355,
      "epoch": 0.5244956772334294,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.2855,
      "mean_token_accuracy": 0.19753282517194748,
      "num_tokens": 12525884.0,
      "step": 5460
    },
    {
      "entropy": 5.465018606185913,
      "epoch": 0.5249759846301633,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.2025,
      "mean_token_accuracy": 0.19672263711690902,
      "num_tokens": 12536709.0,
      "step": 5465
    },
    {
      "entropy": 5.452314710617065,
      "epoch": 0.5254562920268973,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2059873014688492,
      "num_tokens": 12549727.0,
      "step": 5470
    },
    {
      "entropy": 5.409617853164673,
      "epoch": 0.5259365994236311,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.2703,
      "mean_token_accuracy": 0.1943661794066429,
      "num_tokens": 12562053.0,
      "step": 5475
    },
    {
      "entropy": 5.334145212173462,
      "epoch": 0.526416906820365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.1597,
      "mean_token_accuracy": 0.2090065971016884,
      "num_tokens": 12574096.0,
      "step": 5480
    },
    {
      "entropy": 5.446299457550049,
      "epoch": 0.5268972142170989,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.3383,
      "mean_token_accuracy": 0.19677683115005493,
      "num_tokens": 12585416.0,
      "step": 5485
    },
    {
      "entropy": 5.424954462051391,
      "epoch": 0.5273775216138329,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.1861,
      "mean_token_accuracy": 0.2083905816078186,
      "num_tokens": 12596861.0,
      "step": 5490
    },
    {
      "entropy": 5.364092206954956,
      "epoch": 0.5278578290105668,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.2039,
      "mean_token_accuracy": 0.20646013617515563,
      "num_tokens": 12607761.0,
      "step": 5495
    },
    {
      "entropy": 5.488156509399414,
      "epoch": 0.5283381364073007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19835564345121384,
      "num_tokens": 12619889.0,
      "step": 5500
    },
    {
      "entropy": 5.516491365432739,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19845593124628066,
      "num_tokens": 12630604.0,
      "step": 5505
    },
    {
      "entropy": 5.354294157028198,
      "epoch": 0.5292987512007685,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20391630083322526,
      "num_tokens": 12641172.0,
      "step": 5510
    },
    {
      "entropy": 5.373248291015625,
      "epoch": 0.5297790585975024,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.1094,
      "mean_token_accuracy": 0.2145618200302124,
      "num_tokens": 12651731.0,
      "step": 5515
    },
    {
      "entropy": 5.428430986404419,
      "epoch": 0.5302593659942363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.2315,
      "mean_token_accuracy": 0.19880771040916442,
      "num_tokens": 12663050.0,
      "step": 5520
    },
    {
      "entropy": 5.507349920272827,
      "epoch": 0.5307396733909702,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.336,
      "mean_token_accuracy": 0.19323476850986482,
      "num_tokens": 12674732.0,
      "step": 5525
    },
    {
      "entropy": 5.358083295822143,
      "epoch": 0.5312199807877042,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497860391494935,
      "loss": 5.1604,
      "mean_token_accuracy": 0.20278584212064743,
      "num_tokens": 12685981.0,
      "step": 5530
    },
    {
      "entropy": 5.474340152740479,
      "epoch": 0.531700288184438,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.291,
      "mean_token_accuracy": 0.19821719378232955,
      "num_tokens": 12697139.0,
      "step": 5535
    },
    {
      "entropy": 5.382588005065918,
      "epoch": 0.5321805955811719,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19628128707408904,
      "num_tokens": 12708268.0,
      "step": 5540
    },
    {
      "entropy": 5.412813377380371,
      "epoch": 0.5326609029779059,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.2185,
      "mean_token_accuracy": 0.20490823090076446,
      "num_tokens": 12720715.0,
      "step": 5545
    },
    {
      "entropy": 5.413670349121094,
      "epoch": 0.5331412103746398,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20597660094499587,
      "num_tokens": 12732409.0,
      "step": 5550
    },
    {
      "entropy": 5.36975474357605,
      "epoch": 0.5336215177713737,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.3292,
      "mean_token_accuracy": 0.19064888060092927,
      "num_tokens": 12745174.0,
      "step": 5555
    },
    {
      "entropy": 5.495306825637817,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.2716,
      "mean_token_accuracy": 0.19742067754268647,
      "num_tokens": 12756116.0,
      "step": 5560
    },
    {
      "entropy": 5.466085290908813,
      "epoch": 0.5345821325648416,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.2048,
      "mean_token_accuracy": 0.20205890983343125,
      "num_tokens": 12768183.0,
      "step": 5565
    },
    {
      "entropy": 5.321042108535766,
      "epoch": 0.5350624399615754,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.226,
      "mean_token_accuracy": 0.1981593042612076,
      "num_tokens": 12778537.0,
      "step": 5570
    },
    {
      "entropy": 5.383537244796753,
      "epoch": 0.5355427473583093,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.2597,
      "mean_token_accuracy": 0.19882985055446625,
      "num_tokens": 12789691.0,
      "step": 5575
    },
    {
      "entropy": 5.494786214828491,
      "epoch": 0.5360230547550432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.281,
      "mean_token_accuracy": 0.19676683098077774,
      "num_tokens": 12801438.0,
      "step": 5580
    },
    {
      "entropy": 5.452530431747436,
      "epoch": 0.5365033621517772,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19537653774023056,
      "num_tokens": 12812425.0,
      "step": 5585
    },
    {
      "entropy": 5.37575855255127,
      "epoch": 0.5369836695485111,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.2057,
      "mean_token_accuracy": 0.19847560077905654,
      "num_tokens": 12824735.0,
      "step": 5590
    },
    {
      "entropy": 5.488990831375122,
      "epoch": 0.537463976945245,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.2972,
      "mean_token_accuracy": 0.20113806277513505,
      "num_tokens": 12835454.0,
      "step": 5595
    },
    {
      "entropy": 5.53157114982605,
      "epoch": 0.5379442843419788,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.346,
      "mean_token_accuracy": 0.19288862645626068,
      "num_tokens": 12847086.0,
      "step": 5600
    },
    {
      "entropy": 5.461941051483154,
      "epoch": 0.5384245917387128,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2786,
      "mean_token_accuracy": 0.20097937434911728,
      "num_tokens": 12857713.0,
      "step": 5605
    },
    {
      "entropy": 5.379673480987549,
      "epoch": 0.5389048991354467,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.2624,
      "mean_token_accuracy": 0.2018478959798813,
      "num_tokens": 12869967.0,
      "step": 5610
    },
    {
      "entropy": 5.492646980285644,
      "epoch": 0.5393852065321806,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19776214063167571,
      "num_tokens": 12881230.0,
      "step": 5615
    },
    {
      "entropy": 5.448351192474365,
      "epoch": 0.5398655139289145,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497774652090923,
      "loss": 5.2605,
      "mean_token_accuracy": 0.2008052185177803,
      "num_tokens": 12892376.0,
      "step": 5620
    },
    {
      "entropy": 5.450916433334351,
      "epoch": 0.5403458213256485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.2683,
      "mean_token_accuracy": 0.2019094258546829,
      "num_tokens": 12903709.0,
      "step": 5625
    },
    {
      "entropy": 5.377445697784424,
      "epoch": 0.5408261287223823,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20498238056898116,
      "num_tokens": 12914958.0,
      "step": 5630
    },
    {
      "entropy": 5.372557878494263,
      "epoch": 0.5413064361191162,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20101455003023147,
      "num_tokens": 12927418.0,
      "step": 5635
    },
    {
      "entropy": 5.437803649902344,
      "epoch": 0.5417867435158501,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.2651,
      "mean_token_accuracy": 0.2012587755918503,
      "num_tokens": 12939874.0,
      "step": 5640
    },
    {
      "entropy": 5.488529539108276,
      "epoch": 0.5422670509125841,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19846536368131637,
      "num_tokens": 12951113.0,
      "step": 5645
    },
    {
      "entropy": 5.358415174484253,
      "epoch": 0.542747358309318,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.178,
      "mean_token_accuracy": 0.20335119515657424,
      "num_tokens": 12961940.0,
      "step": 5650
    },
    {
      "entropy": 5.318795442581177,
      "epoch": 0.5432276657060519,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19999839067459108,
      "num_tokens": 12973938.0,
      "step": 5655
    },
    {
      "entropy": 5.446343898773193,
      "epoch": 0.5437079731027857,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20178124606609343,
      "num_tokens": 12985400.0,
      "step": 5660
    },
    {
      "entropy": 5.417103862762451,
      "epoch": 0.5441882804995197,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.2407,
      "mean_token_accuracy": 0.20964264422655104,
      "num_tokens": 12996402.0,
      "step": 5665
    },
    {
      "entropy": 5.344844722747803,
      "epoch": 0.5446685878962536,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.1284,
      "mean_token_accuracy": 0.21215670555830002,
      "num_tokens": 13006991.0,
      "step": 5670
    },
    {
      "entropy": 5.425043201446533,
      "epoch": 0.5451488952929875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.2295,
      "mean_token_accuracy": 0.20265681445598602,
      "num_tokens": 13016969.0,
      "step": 5675
    },
    {
      "entropy": 5.384352016448974,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497716561458634,
      "loss": 5.1603,
      "mean_token_accuracy": 0.20864049792289735,
      "num_tokens": 13027759.0,
      "step": 5680
    },
    {
      "entropy": 5.512132787704468,
      "epoch": 0.5461095100864554,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1933292031288147,
      "num_tokens": 13039881.0,
      "step": 5685
    },
    {
      "entropy": 5.450646924972534,
      "epoch": 0.5465898174831892,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.2281,
      "mean_token_accuracy": 0.2028298258781433,
      "num_tokens": 13051443.0,
      "step": 5690
    },
    {
      "entropy": 5.4808813571929935,
      "epoch": 0.5470701248799231,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19113280028104782,
      "num_tokens": 13063474.0,
      "step": 5695
    },
    {
      "entropy": 5.38072190284729,
      "epoch": 0.547550432276657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.1591,
      "mean_token_accuracy": 0.21121865510940552,
      "num_tokens": 13074365.0,
      "step": 5700
    },
    {
      "entropy": 5.384839630126953,
      "epoch": 0.548030739673391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.2279,
      "mean_token_accuracy": 0.20203657448291779,
      "num_tokens": 13087354.0,
      "step": 5705
    },
    {
      "entropy": 5.395923471450805,
      "epoch": 0.5485110470701249,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.2118,
      "mean_token_accuracy": 0.2014988273382187,
      "num_tokens": 13099306.0,
      "step": 5710
    },
    {
      "entropy": 5.477517414093017,
      "epoch": 0.5489913544668588,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.3482,
      "mean_token_accuracy": 0.19400807470083237,
      "num_tokens": 13111259.0,
      "step": 5715
    },
    {
      "entropy": 5.509419107437134,
      "epoch": 0.5494716618635928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.261,
      "mean_token_accuracy": 0.19410279840230943,
      "num_tokens": 13123589.0,
      "step": 5720
    },
    {
      "entropy": 5.385268211364746,
      "epoch": 0.5499519692603266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19971895813941956,
      "num_tokens": 13136485.0,
      "step": 5725
    },
    {
      "entropy": 5.445385360717774,
      "epoch": 0.5504322766570605,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.2307,
      "mean_token_accuracy": 0.1987402245402336,
      "num_tokens": 13148067.0,
      "step": 5730
    },
    {
      "entropy": 5.429761600494385,
      "epoch": 0.5509125840537944,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19749991297721864,
      "num_tokens": 13159813.0,
      "step": 5735
    },
    {
      "entropy": 5.470585727691651,
      "epoch": 0.5513928914505284,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.2851,
      "mean_token_accuracy": 0.20123465061187745,
      "num_tokens": 13170828.0,
      "step": 5740
    },
    {
      "entropy": 5.350544357299805,
      "epoch": 0.5518731988472623,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.1144,
      "mean_token_accuracy": 0.21216228753328323,
      "num_tokens": 13181883.0,
      "step": 5745
    },
    {
      "entropy": 5.395107269287109,
      "epoch": 0.5523535062439962,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.1864,
      "mean_token_accuracy": 0.20682404637336732,
      "num_tokens": 13193530.0,
      "step": 5750
    },
    {
      "entropy": 5.518691444396973,
      "epoch": 0.55283381364073,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.4119,
      "mean_token_accuracy": 0.18553577214479447,
      "num_tokens": 13205822.0,
      "step": 5755
    },
    {
      "entropy": 5.390845012664795,
      "epoch": 0.553314121037464,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.1375,
      "mean_token_accuracy": 0.2063609778881073,
      "num_tokens": 13216698.0,
      "step": 5760
    },
    {
      "entropy": 5.372720384597779,
      "epoch": 0.5537944284341979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.2362,
      "mean_token_accuracy": 0.20510985404253007,
      "num_tokens": 13228389.0,
      "step": 5765
    },
    {
      "entropy": 5.459476041793823,
      "epoch": 0.5542747358309318,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.2863,
      "mean_token_accuracy": 0.19380225241184235,
      "num_tokens": 13239628.0,
      "step": 5770
    },
    {
      "entropy": 5.45081353187561,
      "epoch": 0.5547550432276657,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20101425051689148,
      "num_tokens": 13251149.0,
      "step": 5775
    },
    {
      "entropy": 5.557024431228638,
      "epoch": 0.5552353506243997,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497618089174132,
      "loss": 5.3309,
      "mean_token_accuracy": 0.19553734362125397,
      "num_tokens": 13264846.0,
      "step": 5780
    },
    {
      "entropy": 5.340891885757446,
      "epoch": 0.5557156580211335,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2120337665081024,
      "num_tokens": 13275409.0,
      "step": 5785
    },
    {
      "entropy": 5.256704425811767,
      "epoch": 0.5561959654178674,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21279428601264955,
      "num_tokens": 13287173.0,
      "step": 5790
    },
    {
      "entropy": 5.466112899780273,
      "epoch": 0.5566762728146013,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497603139994675,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20267048478126526,
      "num_tokens": 13298225.0,
      "step": 5795
    },
    {
      "entropy": 5.411417007446289,
      "epoch": 0.5571565802113353,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.1918,
      "mean_token_accuracy": 0.2040710151195526,
      "num_tokens": 13309685.0,
      "step": 5800
    },
    {
      "entropy": 5.349876117706299,
      "epoch": 0.5576368876080692,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.201,
      "mean_token_accuracy": 0.20868574678897858,
      "num_tokens": 13320837.0,
      "step": 5805
    },
    {
      "entropy": 5.454625225067138,
      "epoch": 0.5581171950048031,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1992785006761551,
      "num_tokens": 13333512.0,
      "step": 5810
    },
    {
      "entropy": 5.4970136165618895,
      "epoch": 0.5585975024015369,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19946846067905427,
      "num_tokens": 13345189.0,
      "step": 5815
    },
    {
      "entropy": 5.389864587783814,
      "epoch": 0.5590778097982709,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.1908,
      "mean_token_accuracy": 0.20144879668951035,
      "num_tokens": 13356123.0,
      "step": 5820
    },
    {
      "entropy": 5.391473960876465,
      "epoch": 0.5595581171950048,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.2333,
      "mean_token_accuracy": 0.2022359102964401,
      "num_tokens": 13366902.0,
      "step": 5825
    },
    {
      "entropy": 5.475438785552979,
      "epoch": 0.5600384245917387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.273,
      "mean_token_accuracy": 0.19751251935958863,
      "num_tokens": 13377567.0,
      "step": 5830
    },
    {
      "entropy": 5.451048517227173,
      "epoch": 0.5605187319884726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.2281,
      "mean_token_accuracy": 0.20986885875463485,
      "num_tokens": 13387297.0,
      "step": 5835
    },
    {
      "entropy": 5.444889259338379,
      "epoch": 0.5609990393852066,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.2529,
      "mean_token_accuracy": 0.20367213785648347,
      "num_tokens": 13397723.0,
      "step": 5840
    },
    {
      "entropy": 5.440711307525635,
      "epoch": 0.5614793467819404,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.3125,
      "mean_token_accuracy": 0.19717175215482713,
      "num_tokens": 13409875.0,
      "step": 5845
    },
    {
      "entropy": 5.372532749176026,
      "epoch": 0.5619596541786743,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.2079,
      "mean_token_accuracy": 0.20235518962144852,
      "num_tokens": 13421338.0,
      "step": 5850
    },
    {
      "entropy": 5.458660697937011,
      "epoch": 0.5624399615754082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.2462,
      "mean_token_accuracy": 0.19968458116054535,
      "num_tokens": 13431373.0,
      "step": 5855
    },
    {
      "entropy": 5.4430427074432375,
      "epoch": 0.5629202689721422,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.2574,
      "mean_token_accuracy": 0.2052182987332344,
      "num_tokens": 13443158.0,
      "step": 5860
    },
    {
      "entropy": 5.493396043777466,
      "epoch": 0.5634005763688761,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.3251,
      "mean_token_accuracy": 0.19516831785440444,
      "num_tokens": 13454559.0,
      "step": 5865
    },
    {
      "entropy": 5.5587399959564205,
      "epoch": 0.56388088376561,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.3882,
      "mean_token_accuracy": 0.19617153257131575,
      "num_tokens": 13466934.0,
      "step": 5870
    },
    {
      "entropy": 5.428758668899536,
      "epoch": 0.5643611911623438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.1403,
      "mean_token_accuracy": 0.21330775022506715,
      "num_tokens": 13477770.0,
      "step": 5875
    },
    {
      "entropy": 5.4610990524292,
      "epoch": 0.5648414985590778,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.3431,
      "mean_token_accuracy": 0.19222095012664794,
      "num_tokens": 13490093.0,
      "step": 5880
    },
    {
      "entropy": 5.52604112625122,
      "epoch": 0.5653218059558117,
      "grad_norm": 1.25,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20205526053905487,
      "num_tokens": 13500663.0,
      "step": 5885
    },
    {
      "entropy": 5.442627429962158,
      "epoch": 0.5658021133525456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.3041,
      "mean_token_accuracy": 0.1966984361410141,
      "num_tokens": 13512369.0,
      "step": 5890
    },
    {
      "entropy": 5.422051334381104,
      "epoch": 0.5662824207492796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.1741,
      "mean_token_accuracy": 0.1990967184305191,
      "num_tokens": 13523181.0,
      "step": 5895
    },
    {
      "entropy": 5.411604356765747,
      "epoch": 0.5667627281460135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.1446,
      "mean_token_accuracy": 0.19783616960048675,
      "num_tokens": 13534113.0,
      "step": 5900
    },
    {
      "entropy": 5.355008316040039,
      "epoch": 0.5672430355427474,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.188,
      "mean_token_accuracy": 0.20168602615594863,
      "num_tokens": 13545037.0,
      "step": 5905
    },
    {
      "entropy": 5.4220489978790285,
      "epoch": 0.5677233429394812,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19897332787513733,
      "num_tokens": 13555541.0,
      "step": 5910
    },
    {
      "entropy": 5.389899873733521,
      "epoch": 0.5682036503362152,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.2105,
      "mean_token_accuracy": 0.20374743044376373,
      "num_tokens": 13566951.0,
      "step": 5915
    },
    {
      "entropy": 5.459915590286255,
      "epoch": 0.5686839577329491,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.287,
      "mean_token_accuracy": 0.19948874711990355,
      "num_tokens": 13578492.0,
      "step": 5920
    },
    {
      "entropy": 5.4552897930145265,
      "epoch": 0.569164265129683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.183,
      "mean_token_accuracy": 0.2121192216873169,
      "num_tokens": 13589183.0,
      "step": 5925
    },
    {
      "entropy": 5.458286571502685,
      "epoch": 0.5696445725264169,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.3108,
      "mean_token_accuracy": 0.19659883230924607,
      "num_tokens": 13600588.0,
      "step": 5930
    },
    {
      "entropy": 5.3862463474273685,
      "epoch": 0.5701248799231509,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.1979,
      "mean_token_accuracy": 0.20755694359540938,
      "num_tokens": 13612107.0,
      "step": 5935
    },
    {
      "entropy": 5.540802001953125,
      "epoch": 0.5706051873198847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19629850685596467,
      "num_tokens": 13623973.0,
      "step": 5940
    },
    {
      "entropy": 5.400304937362671,
      "epoch": 0.5710854947166186,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.161,
      "mean_token_accuracy": 0.2056756779551506,
      "num_tokens": 13634994.0,
      "step": 5945
    },
    {
      "entropy": 5.467379760742188,
      "epoch": 0.5715658021133525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.3338,
      "mean_token_accuracy": 0.1981966733932495,
      "num_tokens": 13646361.0,
      "step": 5950
    },
    {
      "entropy": 5.423090648651123,
      "epoch": 0.5720461095100865,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497440788529968,
      "loss": 5.2022,
      "mean_token_accuracy": 0.196497642993927,
      "num_tokens": 13656975.0,
      "step": 5955
    },
    {
      "entropy": 5.415749454498291,
      "epoch": 0.5725264169068204,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1987801671028137,
      "num_tokens": 13668434.0,
      "step": 5960
    },
    {
      "entropy": 5.467151117324829,
      "epoch": 0.5730067243035543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.2347,
      "mean_token_accuracy": 0.20059672445058824,
      "num_tokens": 13679266.0,
      "step": 5965
    },
    {
      "entropy": 5.36699423789978,
      "epoch": 0.5734870317002881,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.246,
      "mean_token_accuracy": 0.19388700574636458,
      "num_tokens": 13690921.0,
      "step": 5970
    },
    {
      "entropy": 5.427312994003296,
      "epoch": 0.5739673390970221,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497420122679239,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20325016975402832,
      "num_tokens": 13702329.0,
      "step": 5975
    },
    {
      "entropy": 5.384410715103149,
      "epoch": 0.574447646493756,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.1722,
      "mean_token_accuracy": 0.21002744883298874,
      "num_tokens": 13713356.0,
      "step": 5980
    },
    {
      "entropy": 5.311866712570191,
      "epoch": 0.5749279538904899,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20121487677097322,
      "num_tokens": 13724718.0,
      "step": 5985
    },
    {
      "entropy": 5.310837936401367,
      "epoch": 0.5754082612872238,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.1827,
      "mean_token_accuracy": 0.20771329551935197,
      "num_tokens": 13736113.0,
      "step": 5990
    },
    {
      "entropy": 5.484830284118653,
      "epoch": 0.5758885686839578,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.2038,
      "mean_token_accuracy": 0.20306566208600998,
      "num_tokens": 13748322.0,
      "step": 5995
    },
    {
      "entropy": 5.364419174194336,
      "epoch": 0.5763688760806917,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.2097,
      "mean_token_accuracy": 0.19368440806865692,
      "num_tokens": 13759433.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.168700758740124,
      "eval_loss": 5.236293315887451,
      "eval_mean_token_accuracy": 0.20920352415176702,
      "eval_num_tokens": 13759433.0,
      "eval_runtime": 20.4353,
      "eval_samples_per_second": 1605.798,
      "eval_steps_per_second": 200.731,
      "step": 6000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3129573368954880.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}