hin-deva-10mb-after-ppt-Dp-100mb-ckpt500_seed3407/checkpoint-27000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.5936599423631126,
  "eval_steps": 3000,
  "global_step": 27000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.307085371017456,
      "epoch": 0.0004803073967339097,
      "grad_norm": 1.203125,
      "learning_rate": 2e-06,
      "loss": 6.6711,
      "mean_token_accuracy": 0.11963684856891632,
      "num_tokens": 10855.0,
      "step": 5
    },
    {
      "entropy": 7.261483955383301,
      "epoch": 0.0009606147934678194,
      "grad_norm": 1.09375,
      "learning_rate": 4.5e-06,
      "loss": 6.611,
      "mean_token_accuracy": 0.12699446231126785,
      "num_tokens": 24110.0,
      "step": 10
    },
    {
      "entropy": 7.322568655014038,
      "epoch": 0.001440922190201729,
      "grad_norm": 1.171875,
      "learning_rate": 7e-06,
      "loss": 6.712,
      "mean_token_accuracy": 0.11978077217936516,
      "num_tokens": 35984.0,
      "step": 15
    },
    {
      "entropy": 7.314567613601684,
      "epoch": 0.0019212295869356388,
      "grad_norm": 1.0859375,
      "learning_rate": 9.5e-06,
      "loss": 6.8292,
      "mean_token_accuracy": 0.11400551423430443,
      "num_tokens": 48152.0,
      "step": 20
    },
    {
      "entropy": 7.27342529296875,
      "epoch": 0.0024015369836695487,
      "grad_norm": 1.125,
      "learning_rate": 1.2e-05,
      "loss": 6.7211,
      "mean_token_accuracy": 0.12296620458364486,
      "num_tokens": 59810.0,
      "step": 25
    },
    {
      "entropy": 7.184682607650757,
      "epoch": 0.002881844380403458,
      "grad_norm": 1.0,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 6.5527,
      "mean_token_accuracy": 0.12848290130496026,
      "num_tokens": 70852.0,
      "step": 30
    },
    {
      "entropy": 7.165819931030273,
      "epoch": 0.0033621517771373678,
      "grad_norm": 1.1171875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 6.6676,
      "mean_token_accuracy": 0.12854331284761428,
      "num_tokens": 83378.0,
      "step": 35
    },
    {
      "entropy": 7.211078310012818,
      "epoch": 0.0038424591738712775,
      "grad_norm": 1.0859375,
      "learning_rate": 1.95e-05,
      "loss": 6.7318,
      "mean_token_accuracy": 0.12352305576205254,
      "num_tokens": 95505.0,
      "step": 40
    },
    {
      "entropy": 7.143542957305908,
      "epoch": 0.004322766570605188,
      "grad_norm": 0.94921875,
      "learning_rate": 2.2e-05,
      "loss": 6.6624,
      "mean_token_accuracy": 0.13083189800381662,
      "num_tokens": 106812.0,
      "step": 45
    },
    {
      "entropy": 7.1696985244750975,
      "epoch": 0.004803073967339097,
      "grad_norm": 0.94140625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 6.6905,
      "mean_token_accuracy": 0.12702840194106102,
      "num_tokens": 118572.0,
      "step": 50
    },
    {
      "entropy": 7.134132242202758,
      "epoch": 0.005283381364073006,
      "grad_norm": 1.0234375,
      "learning_rate": 2.7e-05,
      "loss": 6.6333,
      "mean_token_accuracy": 0.13059643805027008,
      "num_tokens": 130051.0,
      "step": 55
    },
    {
      "entropy": 7.118882560729981,
      "epoch": 0.005763688760806916,
      "grad_norm": 1.1015625,
      "learning_rate": 2.95e-05,
      "loss": 6.7479,
      "mean_token_accuracy": 0.12762305215001107,
      "num_tokens": 141920.0,
      "step": 60
    },
    {
      "entropy": 7.106608819961548,
      "epoch": 0.006243996157540826,
      "grad_norm": 0.91015625,
      "learning_rate": 3.2e-05,
      "loss": 6.6002,
      "mean_token_accuracy": 0.13259410262107849,
      "num_tokens": 152706.0,
      "step": 65
    },
    {
      "entropy": 7.095578145980835,
      "epoch": 0.0067243035542747355,
      "grad_norm": 0.9140625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 6.6396,
      "mean_token_accuracy": 0.1325068600475788,
      "num_tokens": 165253.0,
      "step": 70
    },
    {
      "entropy": 7.086773347854614,
      "epoch": 0.007204610951008645,
      "grad_norm": 0.88671875,
      "learning_rate": 3.7e-05,
      "loss": 6.5959,
      "mean_token_accuracy": 0.13050568848848343,
      "num_tokens": 176708.0,
      "step": 75
    },
    {
      "entropy": 7.126611709594727,
      "epoch": 0.007684918347742555,
      "grad_norm": 1.125,
      "learning_rate": 3.95e-05,
      "loss": 6.69,
      "mean_token_accuracy": 0.1262853056192398,
      "num_tokens": 188240.0,
      "step": 80
    },
    {
      "entropy": 7.042779111862183,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.078125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 6.496,
      "mean_token_accuracy": 0.14095853343605996,
      "num_tokens": 198355.0,
      "step": 85
    },
    {
      "entropy": 7.078471279144287,
      "epoch": 0.008645533141210375,
      "grad_norm": 1.125,
      "learning_rate": 4.45e-05,
      "loss": 6.586,
      "mean_token_accuracy": 0.12663363963365554,
      "num_tokens": 209497.0,
      "step": 90
    },
    {
      "entropy": 7.093791770935058,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.09375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 6.6097,
      "mean_token_accuracy": 0.1298912487924099,
      "num_tokens": 220859.0,
      "step": 95
    },
    {
      "entropy": 7.139126348495483,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.015625,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.7121,
      "mean_token_accuracy": 0.12592171207070352,
      "num_tokens": 231550.0,
      "step": 100
    },
    {
      "entropy": 7.0799497127532955,
      "epoch": 0.010086455331412104,
      "grad_norm": 0.96875,
      "learning_rate": 5.2e-05,
      "loss": 6.6082,
      "mean_token_accuracy": 0.12669628784060477,
      "num_tokens": 244210.0,
      "step": 105
    },
    {
      "entropy": 7.108440780639649,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.03125,
      "learning_rate": 5.45e-05,
      "loss": 6.5846,
      "mean_token_accuracy": 0.13026389181613923,
      "num_tokens": 255745.0,
      "step": 110
    },
    {
      "entropy": 7.094820356369018,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.1328125,
      "learning_rate": 5.7e-05,
      "loss": 6.581,
      "mean_token_accuracy": 0.1345028355717659,
      "num_tokens": 266180.0,
      "step": 115
    },
    {
      "entropy": 7.168249034881592,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.0,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 6.6584,
      "mean_token_accuracy": 0.13001366779208184,
      "num_tokens": 277736.0,
      "step": 120
    },
    {
      "entropy": 7.125609493255615,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.234375,
      "learning_rate": 6.2e-05,
      "loss": 6.6671,
      "mean_token_accuracy": 0.1278264619410038,
      "num_tokens": 289069.0,
      "step": 125
    },
    {
      "entropy": 7.0424315452575685,
      "epoch": 0.012487992315081652,
      "grad_norm": 0.95703125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 6.6483,
      "mean_token_accuracy": 0.1312246046960354,
      "num_tokens": 300240.0,
      "step": 130
    },
    {
      "entropy": 7.105067348480224,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.0625,
      "learning_rate": 6.7e-05,
      "loss": 6.6951,
      "mean_token_accuracy": 0.13201999366283418,
      "num_tokens": 311698.0,
      "step": 135
    },
    {
      "entropy": 7.135102462768555,
      "epoch": 0.013448607108549471,
      "grad_norm": 1.03125,
      "learning_rate": 6.950000000000001e-05,
      "loss": 6.652,
      "mean_token_accuracy": 0.129649805277586,
      "num_tokens": 322844.0,
      "step": 140
    },
    {
      "entropy": 7.104988622665405,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.03125,
      "learning_rate": 7.2e-05,
      "loss": 6.6544,
      "mean_token_accuracy": 0.13054293021559715,
      "num_tokens": 335092.0,
      "step": 145
    },
    {
      "entropy": 7.036244344711304,
      "epoch": 0.01440922190201729,
      "grad_norm": 0.97265625,
      "learning_rate": 7.45e-05,
      "loss": 6.6388,
      "mean_token_accuracy": 0.13229672610759735,
      "num_tokens": 347033.0,
      "step": 150
    },
    {
      "entropy": 7.0629229068756105,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.1953125,
      "learning_rate": 7.7e-05,
      "loss": 6.6368,
      "mean_token_accuracy": 0.12350815683603286,
      "num_tokens": 358696.0,
      "step": 155
    },
    {
      "entropy": 7.103863048553467,
      "epoch": 0.01536983669548511,
      "grad_norm": 1.1171875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 6.6527,
      "mean_token_accuracy": 0.12880626693367958,
      "num_tokens": 369390.0,
      "step": 160
    },
    {
      "entropy": 7.0600316524505615,
      "epoch": 0.01585014409221902,
      "grad_norm": 1.015625,
      "learning_rate": 8.2e-05,
      "loss": 6.5942,
      "mean_token_accuracy": 0.12720233500003814,
      "num_tokens": 380540.0,
      "step": 165
    },
    {
      "entropy": 7.0939164638519285,
      "epoch": 0.01633045148895293,
      "grad_norm": 1.15625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 6.5928,
      "mean_token_accuracy": 0.1307683579623699,
      "num_tokens": 391243.0,
      "step": 170
    },
    {
      "entropy": 7.08874363899231,
      "epoch": 0.01681075888568684,
      "grad_norm": 0.94921875,
      "learning_rate": 8.7e-05,
      "loss": 6.5721,
      "mean_token_accuracy": 0.1347414717078209,
      "num_tokens": 403336.0,
      "step": 175
    },
    {
      "entropy": 7.054573011398316,
      "epoch": 0.01729106628242075,
      "grad_norm": 1.21875,
      "learning_rate": 8.95e-05,
      "loss": 6.5372,
      "mean_token_accuracy": 0.1348336383700371,
      "num_tokens": 413886.0,
      "step": 180
    },
    {
      "entropy": 7.076615619659424,
      "epoch": 0.01777137367915466,
      "grad_norm": 1.2265625,
      "learning_rate": 9.2e-05,
      "loss": 6.6369,
      "mean_token_accuracy": 0.12773096263408662,
      "num_tokens": 425277.0,
      "step": 185
    },
    {
      "entropy": 7.169283103942871,
      "epoch": 0.01825168107588857,
      "grad_norm": 1.09375,
      "learning_rate": 9.45e-05,
      "loss": 6.6817,
      "mean_token_accuracy": 0.13388459086418153,
      "num_tokens": 436868.0,
      "step": 190
    },
    {
      "entropy": 7.070893096923828,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.1328125,
      "learning_rate": 9.7e-05,
      "loss": 6.5585,
      "mean_token_accuracy": 0.13712915256619454,
      "num_tokens": 448349.0,
      "step": 195
    },
    {
      "entropy": 7.00331335067749,
      "epoch": 0.01921229586935639,
      "grad_norm": 1.1015625,
      "learning_rate": 9.95e-05,
      "loss": 6.5787,
      "mean_token_accuracy": 0.13295828625559808,
      "num_tokens": 459447.0,
      "step": 200
    },
    {
      "entropy": 7.017018413543701,
      "epoch": 0.0196926032660903,
      "grad_norm": 1.09375,
      "learning_rate": 0.000102,
      "loss": 6.5664,
      "mean_token_accuracy": 0.13398413211107255,
      "num_tokens": 470734.0,
      "step": 205
    },
    {
      "entropy": 7.074815893173218,
      "epoch": 0.020172910662824207,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00010449999999999999,
      "loss": 6.6675,
      "mean_token_accuracy": 0.1243060939013958,
      "num_tokens": 482015.0,
      "step": 210
    },
    {
      "entropy": 7.051287984848022,
      "epoch": 0.020653218059558116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000107,
      "loss": 6.5853,
      "mean_token_accuracy": 0.12988396733999252,
      "num_tokens": 493339.0,
      "step": 215
    },
    {
      "entropy": 7.021688842773438,
      "epoch": 0.021133525456292025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001095,
      "loss": 6.5124,
      "mean_token_accuracy": 0.13405952081084252,
      "num_tokens": 504924.0,
      "step": 220
    },
    {
      "entropy": 7.087656021118164,
      "epoch": 0.021613832853025938,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000112,
      "loss": 6.5272,
      "mean_token_accuracy": 0.12774203419685365,
      "num_tokens": 516603.0,
      "step": 225
    },
    {
      "entropy": 6.912145376205444,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001145,
      "loss": 6.4572,
      "mean_token_accuracy": 0.13284713625907899,
      "num_tokens": 528347.0,
      "step": 230
    },
    {
      "entropy": 7.009493589401245,
      "epoch": 0.022574447646493755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00011700000000000001,
      "loss": 6.5941,
      "mean_token_accuracy": 0.13424275815486908,
      "num_tokens": 539328.0,
      "step": 235
    },
    {
      "entropy": 7.143858766555786,
      "epoch": 0.023054755043227664,
      "grad_norm": 1.171875,
      "learning_rate": 0.00011949999999999999,
      "loss": 6.5522,
      "mean_token_accuracy": 0.1398538187146187,
      "num_tokens": 549297.0,
      "step": 240
    },
    {
      "entropy": 6.905237674713135,
      "epoch": 0.023535062439961577,
      "grad_norm": 1.15625,
      "learning_rate": 0.000122,
      "loss": 6.545,
      "mean_token_accuracy": 0.13216322883963585,
      "num_tokens": 560306.0,
      "step": 245
    },
    {
      "entropy": 7.09604697227478,
      "epoch": 0.024015369836695485,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001245,
      "loss": 6.6323,
      "mean_token_accuracy": 0.1321306437253952,
      "num_tokens": 571972.0,
      "step": 250
    },
    {
      "entropy": 7.053035020828247,
      "epoch": 0.024495677233429394,
      "grad_norm": 1.21875,
      "learning_rate": 0.000127,
      "loss": 6.5172,
      "mean_token_accuracy": 0.13756523206830024,
      "num_tokens": 582962.0,
      "step": 255
    },
    {
      "entropy": 6.858367395401001,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.0,
      "learning_rate": 0.0001295,
      "loss": 6.5103,
      "mean_token_accuracy": 0.13402715027332307,
      "num_tokens": 597193.0,
      "step": 260
    },
    {
      "entropy": 7.156196880340576,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000132,
      "loss": 6.5892,
      "mean_token_accuracy": 0.12901883125305175,
      "num_tokens": 608982.0,
      "step": 265
    },
    {
      "entropy": 7.014424085617065,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.140625,
      "learning_rate": 0.00013450000000000002,
      "loss": 6.6212,
      "mean_token_accuracy": 0.12978230491280557,
      "num_tokens": 619953.0,
      "step": 270
    },
    {
      "entropy": 7.096269750595093,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.6321,
      "mean_token_accuracy": 0.12898936718702317,
      "num_tokens": 631039.0,
      "step": 275
    },
    {
      "entropy": 7.03494119644165,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001395,
      "loss": 6.6062,
      "mean_token_accuracy": 0.13531998470425605,
      "num_tokens": 642656.0,
      "step": 280
    },
    {
      "entropy": 6.969197034835815,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.15625,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.4944,
      "mean_token_accuracy": 0.1461842603981495,
      "num_tokens": 653748.0,
      "step": 285
    },
    {
      "entropy": 6.980546951293945,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001445,
      "loss": 6.4709,
      "mean_token_accuracy": 0.1350526325404644,
      "num_tokens": 665618.0,
      "step": 290
    },
    {
      "entropy": 6.976346349716186,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000147,
      "loss": 6.5103,
      "mean_token_accuracy": 0.13303453028202056,
      "num_tokens": 677329.0,
      "step": 295
    },
    {
      "entropy": 6.96361517906189,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001495,
      "loss": 6.5113,
      "mean_token_accuracy": 0.1333675727248192,
      "num_tokens": 688278.0,
      "step": 300
    },
    {
      "entropy": 6.986789083480835,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000152,
      "loss": 6.5677,
      "mean_token_accuracy": 0.13771163299679756,
      "num_tokens": 700739.0,
      "step": 305
    },
    {
      "entropy": 6.966262865066528,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.15625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.5046,
      "mean_token_accuracy": 0.13294478878378868,
      "num_tokens": 712527.0,
      "step": 310
    },
    {
      "entropy": 6.959698724746704,
      "epoch": 0.03025936599423631,
      "grad_norm": 1.0625,
      "learning_rate": 0.000157,
      "loss": 6.5456,
      "mean_token_accuracy": 0.13097512274980544,
      "num_tokens": 724514.0,
      "step": 315
    },
    {
      "entropy": 7.038305854797363,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001595,
      "loss": 6.5092,
      "mean_token_accuracy": 0.13729717060923577,
      "num_tokens": 735679.0,
      "step": 320
    },
    {
      "entropy": 6.877438640594482,
      "epoch": 0.03121998078770413,
      "grad_norm": 0.984375,
      "learning_rate": 0.000162,
      "loss": 6.4986,
      "mean_token_accuracy": 0.13833012878894807,
      "num_tokens": 747896.0,
      "step": 325
    },
    {
      "entropy": 7.030805873870849,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.125,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.5506,
      "mean_token_accuracy": 0.13383892253041269,
      "num_tokens": 759081.0,
      "step": 330
    },
    {
      "entropy": 6.936169385910034,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.234375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.502,
      "mean_token_accuracy": 0.1336835578083992,
      "num_tokens": 770459.0,
      "step": 335
    },
    {
      "entropy": 6.922156190872192,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.4557,
      "mean_token_accuracy": 0.13467910513281822,
      "num_tokens": 783960.0,
      "step": 340
    },
    {
      "entropy": 6.945554304122925,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.5724,
      "mean_token_accuracy": 0.13674034401774407,
      "num_tokens": 795425.0,
      "step": 345
    },
    {
      "entropy": 6.9888596534729,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.5241,
      "mean_token_accuracy": 0.13427431806921958,
      "num_tokens": 807536.0,
      "step": 350
    },
    {
      "entropy": 7.014977931976318,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000177,
      "loss": 6.5882,
      "mean_token_accuracy": 0.13020211309194565,
      "num_tokens": 818801.0,
      "step": 355
    },
    {
      "entropy": 6.9366690158844,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0001795,
      "loss": 6.4507,
      "mean_token_accuracy": 0.13554891422390938,
      "num_tokens": 831497.0,
      "step": 360
    },
    {
      "entropy": 6.8712386131286625,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000182,
      "loss": 6.4424,
      "mean_token_accuracy": 0.13640913739800453,
      "num_tokens": 842491.0,
      "step": 365
    },
    {
      "entropy": 7.009519624710083,
      "epoch": 0.03554274735830932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001845,
      "loss": 6.523,
      "mean_token_accuracy": 0.12829140126705169,
      "num_tokens": 854560.0,
      "step": 370
    },
    {
      "entropy": 6.867582225799561,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000187,
      "loss": 6.4306,
      "mean_token_accuracy": 0.14245563745498657,
      "num_tokens": 866688.0,
      "step": 375
    },
    {
      "entropy": 6.9899519920349125,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001895,
      "loss": 6.5572,
      "mean_token_accuracy": 0.12518782839179038,
      "num_tokens": 879484.0,
      "step": 380
    },
    {
      "entropy": 6.972244501113892,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.25,
      "learning_rate": 0.000192,
      "loss": 6.5485,
      "mean_token_accuracy": 0.12844221517443657,
      "num_tokens": 890807.0,
      "step": 385
    },
    {
      "entropy": 6.982624101638794,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.359375,
      "learning_rate": 0.0001945,
      "loss": 6.4655,
      "mean_token_accuracy": 0.1351318120956421,
      "num_tokens": 901759.0,
      "step": 390
    },
    {
      "entropy": 6.820593404769897,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.4239,
      "mean_token_accuracy": 0.13946242481470109,
      "num_tokens": 912212.0,
      "step": 395
    },
    {
      "entropy": 6.997915410995484,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.5082,
      "mean_token_accuracy": 0.1373220317065716,
      "num_tokens": 923947.0,
      "step": 400
    },
    {
      "entropy": 6.887757110595703,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000202,
      "loss": 6.4624,
      "mean_token_accuracy": 0.13495285138487817,
      "num_tokens": 935732.0,
      "step": 405
    },
    {
      "entropy": 6.860993576049805,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.4234,
      "mean_token_accuracy": 0.13301072269678116,
      "num_tokens": 948261.0,
      "step": 410
    },
    {
      "entropy": 6.998539113998413,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.359375,
      "learning_rate": 0.000207,
      "loss": 6.5408,
      "mean_token_accuracy": 0.13510932326316832,
      "num_tokens": 959574.0,
      "step": 415
    },
    {
      "entropy": 6.97111144065857,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002095,
      "loss": 6.6136,
      "mean_token_accuracy": 0.12621866688132286,
      "num_tokens": 970329.0,
      "step": 420
    },
    {
      "entropy": 6.985296583175659,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000212,
      "loss": 6.4701,
      "mean_token_accuracy": 0.14006434381008148,
      "num_tokens": 982037.0,
      "step": 425
    },
    {
      "entropy": 6.804930353164673,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002145,
      "loss": 6.4271,
      "mean_token_accuracy": 0.13567045107483863,
      "num_tokens": 994612.0,
      "step": 430
    },
    {
      "entropy": 6.955875444412231,
      "epoch": 0.04178674351585014,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.5617,
      "mean_token_accuracy": 0.13497636914253236,
      "num_tokens": 1005960.0,
      "step": 435
    },
    {
      "entropy": 6.869253540039063,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002195,
      "loss": 6.4174,
      "mean_token_accuracy": 0.13668350800871848,
      "num_tokens": 1017618.0,
      "step": 440
    },
    {
      "entropy": 6.8875102519989015,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000222,
      "loss": 6.48,
      "mean_token_accuracy": 0.13301057890057563,
      "num_tokens": 1029307.0,
      "step": 445
    },
    {
      "entropy": 6.912690162658691,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002245,
      "loss": 6.4972,
      "mean_token_accuracy": 0.13038710951805116,
      "num_tokens": 1042027.0,
      "step": 450
    },
    {
      "entropy": 6.889484977722168,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.4819,
      "mean_token_accuracy": 0.13361246138811111,
      "num_tokens": 1053125.0,
      "step": 455
    },
    {
      "entropy": 6.828107833862305,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.171875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.4905,
      "mean_token_accuracy": 0.13674674183130264,
      "num_tokens": 1064908.0,
      "step": 460
    },
    {
      "entropy": 6.9115313529968265,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.203125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.3581,
      "mean_token_accuracy": 0.14502883404493333,
      "num_tokens": 1076328.0,
      "step": 465
    },
    {
      "entropy": 6.896181726455689,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.328125,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.5895,
      "mean_token_accuracy": 0.1260202206671238,
      "num_tokens": 1088469.0,
      "step": 470
    },
    {
      "entropy": 6.823418760299683,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.1875,
      "learning_rate": 0.000237,
      "loss": 6.391,
      "mean_token_accuracy": 0.1406915917992592,
      "num_tokens": 1099408.0,
      "step": 475
    },
    {
      "entropy": 6.765524959564209,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002395,
      "loss": 6.3342,
      "mean_token_accuracy": 0.14521979689598083,
      "num_tokens": 1111101.0,
      "step": 480
    },
    {
      "entropy": 6.744671630859375,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.15625,
      "learning_rate": 0.000242,
      "loss": 6.3479,
      "mean_token_accuracy": 0.14791416078805925,
      "num_tokens": 1122877.0,
      "step": 485
    },
    {
      "entropy": 6.7991739273071286,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.25,
      "learning_rate": 0.0002445,
      "loss": 6.454,
      "mean_token_accuracy": 0.1279207557439804,
      "num_tokens": 1133956.0,
      "step": 490
    },
    {
      "entropy": 6.823109483718872,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000247,
      "loss": 6.3641,
      "mean_token_accuracy": 0.13983470499515532,
      "num_tokens": 1146101.0,
      "step": 495
    },
    {
      "entropy": 6.855337381362915,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.296875,
      "learning_rate": 0.0002495,
      "loss": 6.4502,
      "mean_token_accuracy": 0.14883086159825326,
      "num_tokens": 1157432.0,
      "step": 500
    },
    {
      "entropy": 6.822028875350952,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000252,
      "loss": 6.4439,
      "mean_token_accuracy": 0.14014339298009873,
      "num_tokens": 1167601.0,
      "step": 505
    },
    {
      "entropy": 6.793749713897705,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002545,
      "loss": 6.3433,
      "mean_token_accuracy": 0.14071932584047317,
      "num_tokens": 1178818.0,
      "step": 510
    },
    {
      "entropy": 6.961408185958862,
      "epoch": 0.0494716618635927,
      "grad_norm": 1.109375,
      "learning_rate": 0.000257,
      "loss": 6.5897,
      "mean_token_accuracy": 0.13712238147854805,
      "num_tokens": 1189977.0,
      "step": 515
    },
    {
      "entropy": 6.825463342666626,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002595,
      "loss": 6.4273,
      "mean_token_accuracy": 0.13879103511571883,
      "num_tokens": 1201039.0,
      "step": 520
    },
    {
      "entropy": 6.885947799682617,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000262,
      "loss": 6.4418,
      "mean_token_accuracy": 0.13869711384177208,
      "num_tokens": 1212573.0,
      "step": 525
    },
    {
      "entropy": 6.831294298171997,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.3877,
      "mean_token_accuracy": 0.14377668723464013,
      "num_tokens": 1223382.0,
      "step": 530
    },
    {
      "entropy": 6.8336584091186525,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.0,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.4756,
      "mean_token_accuracy": 0.1397824764251709,
      "num_tokens": 1236501.0,
      "step": 535
    },
    {
      "entropy": 6.854681539535522,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.4854,
      "mean_token_accuracy": 0.1359047219157219,
      "num_tokens": 1246798.0,
      "step": 540
    },
    {
      "entropy": 6.818022680282593,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.203125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.3993,
      "mean_token_accuracy": 0.1383517175912857,
      "num_tokens": 1258182.0,
      "step": 545
    },
    {
      "entropy": 6.725576877593994,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.25,
      "learning_rate": 0.0002745,
      "loss": 6.3529,
      "mean_token_accuracy": 0.13154633566737176,
      "num_tokens": 1270273.0,
      "step": 550
    },
    {
      "entropy": 6.824942541122437,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000277,
      "loss": 6.4187,
      "mean_token_accuracy": 0.13395696729421616,
      "num_tokens": 1281136.0,
      "step": 555
    },
    {
      "entropy": 6.798213529586792,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002795,
      "loss": 6.388,
      "mean_token_accuracy": 0.13719379529356956,
      "num_tokens": 1294488.0,
      "step": 560
    },
    {
      "entropy": 6.690643072128296,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.342,
      "mean_token_accuracy": 0.13738902136683465,
      "num_tokens": 1304113.0,
      "step": 565
    },
    {
      "entropy": 6.811877346038818,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002845,
      "loss": 6.3715,
      "mean_token_accuracy": 0.1435349702835083,
      "num_tokens": 1315417.0,
      "step": 570
    },
    {
      "entropy": 6.705857038497925,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000287,
      "loss": 6.3487,
      "mean_token_accuracy": 0.13941997289657593,
      "num_tokens": 1328084.0,
      "step": 575
    },
    {
      "entropy": 6.830380821228028,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002895,
      "loss": 6.4948,
      "mean_token_accuracy": 0.13673060759902,
      "num_tokens": 1338696.0,
      "step": 580
    },
    {
      "entropy": 6.816842555999756,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.046875,
      "learning_rate": 0.000292,
      "loss": 6.4362,
      "mean_token_accuracy": 0.13864873573184014,
      "num_tokens": 1350240.0,
      "step": 585
    },
    {
      "entropy": 6.8044596195220945,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002945,
      "loss": 6.3876,
      "mean_token_accuracy": 0.13224058821797371,
      "num_tokens": 1361720.0,
      "step": 590
    },
    {
      "entropy": 6.702146100997925,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.375,
      "learning_rate": 0.000297,
      "loss": 6.388,
      "mean_token_accuracy": 0.14646777287125587,
      "num_tokens": 1373286.0,
      "step": 595
    },
    {
      "entropy": 6.792327165603638,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002995,
      "loss": 6.3357,
      "mean_token_accuracy": 0.14013056233525276,
      "num_tokens": 1384274.0,
      "step": 600
    },
    {
      "entropy": 6.695337629318237,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.15625,
      "learning_rate": 0.000302,
      "loss": 6.3866,
      "mean_token_accuracy": 0.13806435391306876,
      "num_tokens": 1395355.0,
      "step": 605
    },
    {
      "entropy": 6.67833251953125,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003045,
      "loss": 6.3185,
      "mean_token_accuracy": 0.14838905334472657,
      "num_tokens": 1406664.0,
      "step": 610
    },
    {
      "entropy": 6.84452452659607,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.1875,
      "learning_rate": 0.000307,
      "loss": 6.4734,
      "mean_token_accuracy": 0.1365173079073429,
      "num_tokens": 1418450.0,
      "step": 615
    },
    {
      "entropy": 6.745831251144409,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003095,
      "loss": 6.365,
      "mean_token_accuracy": 0.1397569701075554,
      "num_tokens": 1430048.0,
      "step": 620
    },
    {
      "entropy": 6.811714601516724,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000312,
      "loss": 6.4099,
      "mean_token_accuracy": 0.13645492121577263,
      "num_tokens": 1441820.0,
      "step": 625
    },
    {
      "entropy": 6.656697416305542,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003145,
      "loss": 6.2619,
      "mean_token_accuracy": 0.14607831984758377,
      "num_tokens": 1453209.0,
      "step": 630
    },
    {
      "entropy": 6.677804946899414,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.171875,
      "learning_rate": 0.000317,
      "loss": 6.2859,
      "mean_token_accuracy": 0.139230377972126,
      "num_tokens": 1465423.0,
      "step": 635
    },
    {
      "entropy": 6.6918044090271,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003195,
      "loss": 6.2776,
      "mean_token_accuracy": 0.14577876925468444,
      "num_tokens": 1476575.0,
      "step": 640
    },
    {
      "entropy": 6.6395305633544925,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.296875,
      "learning_rate": 0.000322,
      "loss": 6.2269,
      "mean_token_accuracy": 0.14229501113295556,
      "num_tokens": 1486932.0,
      "step": 645
    },
    {
      "entropy": 6.6645488262176515,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.125,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.2634,
      "mean_token_accuracy": 0.1412537418305874,
      "num_tokens": 1498494.0,
      "step": 650
    },
    {
      "entropy": 6.728798723220825,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.3314,
      "mean_token_accuracy": 0.1367237947881222,
      "num_tokens": 1509937.0,
      "step": 655
    },
    {
      "entropy": 6.730763149261475,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.390625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.3493,
      "mean_token_accuracy": 0.14772669598460197,
      "num_tokens": 1519823.0,
      "step": 660
    },
    {
      "entropy": 6.728877162933349,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.377,
      "mean_token_accuracy": 0.1385086067020893,
      "num_tokens": 1529943.0,
      "step": 665
    },
    {
      "entropy": 6.646539354324341,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.218,
      "mean_token_accuracy": 0.1462228536605835,
      "num_tokens": 1540618.0,
      "step": 670
    },
    {
      "entropy": 6.621165466308594,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000337,
      "loss": 6.3349,
      "mean_token_accuracy": 0.13347838148474694,
      "num_tokens": 1553208.0,
      "step": 675
    },
    {
      "entropy": 6.75033278465271,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003395,
      "loss": 6.3079,
      "mean_token_accuracy": 0.14778160005807878,
      "num_tokens": 1563975.0,
      "step": 680
    },
    {
      "entropy": 6.730441284179688,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.25,
      "learning_rate": 0.000342,
      "loss": 6.3601,
      "mean_token_accuracy": 0.13537456542253495,
      "num_tokens": 1575998.0,
      "step": 685
    },
    {
      "entropy": 6.659586429595947,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.25,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.263,
      "mean_token_accuracy": 0.14849890321493148,
      "num_tokens": 1586041.0,
      "step": 690
    },
    {
      "entropy": 6.605801343917847,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.34375,
      "learning_rate": 0.000347,
      "loss": 6.324,
      "mean_token_accuracy": 0.14218652099370957,
      "num_tokens": 1597531.0,
      "step": 695
    },
    {
      "entropy": 6.669734811782837,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003495,
      "loss": 6.3116,
      "mean_token_accuracy": 0.13976099193096161,
      "num_tokens": 1609255.0,
      "step": 700
    },
    {
      "entropy": 6.609857988357544,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000352,
      "loss": 6.2983,
      "mean_token_accuracy": 0.14360181987285614,
      "num_tokens": 1621098.0,
      "step": 705
    },
    {
      "entropy": 6.6083897113800045,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003545,
      "loss": 6.189,
      "mean_token_accuracy": 0.15515244603157044,
      "num_tokens": 1631941.0,
      "step": 710
    },
    {
      "entropy": 6.57919979095459,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000357,
      "loss": 6.2074,
      "mean_token_accuracy": 0.14852219969034194,
      "num_tokens": 1643117.0,
      "step": 715
    },
    {
      "entropy": 6.576744127273559,
      "epoch": 0.069164265129683,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003595,
      "loss": 6.2372,
      "mean_token_accuracy": 0.1395713485777378,
      "num_tokens": 1653595.0,
      "step": 720
    },
    {
      "entropy": 6.663236331939697,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000362,
      "loss": 6.2792,
      "mean_token_accuracy": 0.14607299417257308,
      "num_tokens": 1664495.0,
      "step": 725
    },
    {
      "entropy": 6.556261014938355,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003645,
      "loss": 6.2639,
      "mean_token_accuracy": 0.1435217887163162,
      "num_tokens": 1674923.0,
      "step": 730
    },
    {
      "entropy": 6.618684720993042,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000367,
      "loss": 6.1667,
      "mean_token_accuracy": 0.1526644468307495,
      "num_tokens": 1685904.0,
      "step": 735
    },
    {
      "entropy": 6.627186965942383,
      "epoch": 0.07108549471661864,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003695,
      "loss": 6.3877,
      "mean_token_accuracy": 0.13355778381228448,
      "num_tokens": 1699133.0,
      "step": 740
    },
    {
      "entropy": 6.548422527313233,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.0625,
      "learning_rate": 0.000372,
      "loss": 6.2018,
      "mean_token_accuracy": 0.14151064530014992,
      "num_tokens": 1711559.0,
      "step": 745
    },
    {
      "entropy": 6.550561428070068,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003745,
      "loss": 6.1238,
      "mean_token_accuracy": 0.1509994685649872,
      "num_tokens": 1722526.0,
      "step": 750
    },
    {
      "entropy": 6.532330465316773,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.203125,
      "learning_rate": 0.000377,
      "loss": 6.1889,
      "mean_token_accuracy": 0.14655965566635132,
      "num_tokens": 1734261.0,
      "step": 755
    },
    {
      "entropy": 6.53015832901001,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003795,
      "loss": 6.1803,
      "mean_token_accuracy": 0.15340202897787095,
      "num_tokens": 1745151.0,
      "step": 760
    },
    {
      "entropy": 6.69319715499878,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000382,
      "loss": 6.3355,
      "mean_token_accuracy": 0.1429338976740837,
      "num_tokens": 1755463.0,
      "step": 765
    },
    {
      "entropy": 6.623394536972046,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003845,
      "loss": 6.2828,
      "mean_token_accuracy": 0.13968928754329682,
      "num_tokens": 1767717.0,
      "step": 770
    },
    {
      "entropy": 6.437202119827271,
      "epoch": 0.074447646493756,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.1355,
      "mean_token_accuracy": 0.15134122371673583,
      "num_tokens": 1779115.0,
      "step": 775
    },
    {
      "entropy": 6.419523668289185,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.087,
      "mean_token_accuracy": 0.15281283408403395,
      "num_tokens": 1789644.0,
      "step": 780
    },
    {
      "entropy": 6.6132158756256105,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.2997,
      "mean_token_accuracy": 0.14345607832074164,
      "num_tokens": 1800606.0,
      "step": 785
    },
    {
      "entropy": 6.524192380905151,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.301,
      "mean_token_accuracy": 0.13897362276911734,
      "num_tokens": 1812168.0,
      "step": 790
    },
    {
      "entropy": 6.73820424079895,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.2386,
      "mean_token_accuracy": 0.14347319155931473,
      "num_tokens": 1823830.0,
      "step": 795
    },
    {
      "entropy": 6.445359086990356,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003995,
      "loss": 6.2403,
      "mean_token_accuracy": 0.1467433527112007,
      "num_tokens": 1835611.0,
      "step": 800
    },
    {
      "entropy": 6.597848510742187,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.09375,
      "learning_rate": 0.000402,
      "loss": 6.3054,
      "mean_token_accuracy": 0.13762886747717856,
      "num_tokens": 1847036.0,
      "step": 805
    },
    {
      "entropy": 6.562679290771484,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004045,
      "loss": 6.2378,
      "mean_token_accuracy": 0.14037195593118668,
      "num_tokens": 1857476.0,
      "step": 810
    },
    {
      "entropy": 6.5669573783874515,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2621,
      "mean_token_accuracy": 0.14760917276144028,
      "num_tokens": 1869073.0,
      "step": 815
    },
    {
      "entropy": 6.387801313400269,
      "epoch": 0.0787704130643612,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004095,
      "loss": 6.1412,
      "mean_token_accuracy": 0.1509813189506531,
      "num_tokens": 1880439.0,
      "step": 820
    },
    {
      "entropy": 6.699515724182129,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000412,
      "loss": 6.2976,
      "mean_token_accuracy": 0.14118361324071885,
      "num_tokens": 1891600.0,
      "step": 825
    },
    {
      "entropy": 6.5128673076629635,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004145,
      "loss": 6.2569,
      "mean_token_accuracy": 0.14543531611561775,
      "num_tokens": 1903126.0,
      "step": 830
    },
    {
      "entropy": 6.672891998291016,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.109375,
      "learning_rate": 0.000417,
      "loss": 6.226,
      "mean_token_accuracy": 0.14782591313123702,
      "num_tokens": 1913913.0,
      "step": 835
    },
    {
      "entropy": 6.366758394241333,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004195,
      "loss": 6.1501,
      "mean_token_accuracy": 0.1501932203769684,
      "num_tokens": 1924961.0,
      "step": 840
    },
    {
      "entropy": 6.692445278167725,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.265625,
      "learning_rate": 0.000422,
      "loss": 6.2407,
      "mean_token_accuracy": 0.1381278119981289,
      "num_tokens": 1936773.0,
      "step": 845
    },
    {
      "entropy": 6.406199932098389,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004245,
      "loss": 6.1643,
      "mean_token_accuracy": 0.1459251821041107,
      "num_tokens": 1948190.0,
      "step": 850
    },
    {
      "entropy": 6.452398300170898,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.171875,
      "learning_rate": 0.000427,
      "loss": 6.1575,
      "mean_token_accuracy": 0.14858585894107817,
      "num_tokens": 1960038.0,
      "step": 855
    },
    {
      "entropy": 6.57236967086792,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004295,
      "loss": 6.2384,
      "mean_token_accuracy": 0.14626160115003586,
      "num_tokens": 1970535.0,
      "step": 860
    },
    {
      "entropy": 6.537123394012451,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.171875,
      "learning_rate": 0.000432,
      "loss": 6.2265,
      "mean_token_accuracy": 0.142296252399683,
      "num_tokens": 1981386.0,
      "step": 865
    },
    {
      "entropy": 6.4313184261322025,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004345,
      "loss": 6.1375,
      "mean_token_accuracy": 0.14900052845478057,
      "num_tokens": 1993196.0,
      "step": 870
    },
    {
      "entropy": 6.4812110424041744,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.046875,
      "learning_rate": 0.000437,
      "loss": 6.1562,
      "mean_token_accuracy": 0.15512351393699647,
      "num_tokens": 2004756.0,
      "step": 875
    },
    {
      "entropy": 6.437265253067016,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004395,
      "loss": 6.1392,
      "mean_token_accuracy": 0.14693072885274888,
      "num_tokens": 2016020.0,
      "step": 880
    },
    {
      "entropy": 6.458638095855713,
      "epoch": 0.08501440922190202,
      "grad_norm": 1.125,
      "learning_rate": 0.000442,
      "loss": 6.1858,
      "mean_token_accuracy": 0.14666839689016342,
      "num_tokens": 2027747.0,
      "step": 885
    },
    {
      "entropy": 6.551916837692261,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004445,
      "loss": 6.1604,
      "mean_token_accuracy": 0.14383722916245462,
      "num_tokens": 2038841.0,
      "step": 890
    },
    {
      "entropy": 6.44727840423584,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000447,
      "loss": 6.2109,
      "mean_token_accuracy": 0.14256930723786354,
      "num_tokens": 2049905.0,
      "step": 895
    },
    {
      "entropy": 6.512820863723755,
      "epoch": 0.08645533141210375,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2817,
      "mean_token_accuracy": 0.14085745438933372,
      "num_tokens": 2062492.0,
      "step": 900
    },
    {
      "entropy": 6.474342775344849,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.216,
      "mean_token_accuracy": 0.14198037534952163,
      "num_tokens": 2073840.0,
      "step": 905
    },
    {
      "entropy": 6.469565916061401,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.1233,
      "mean_token_accuracy": 0.14328503012657165,
      "num_tokens": 2085720.0,
      "step": 910
    },
    {
      "entropy": 6.508474731445313,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2079,
      "mean_token_accuracy": 0.14646697416901588,
      "num_tokens": 2096649.0,
      "step": 915
    },
    {
      "entropy": 6.526723051071167,
      "epoch": 0.08837656099903939,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2087,
      "mean_token_accuracy": 0.14670456051826478,
      "num_tokens": 2109030.0,
      "step": 920
    },
    {
      "entropy": 6.422917222976684,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000462,
      "loss": 6.1155,
      "mean_token_accuracy": 0.15043116360902786,
      "num_tokens": 2121384.0,
      "step": 925
    },
    {
      "entropy": 6.429614877700805,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004645,
      "loss": 6.0734,
      "mean_token_accuracy": 0.15336936861276626,
      "num_tokens": 2131377.0,
      "step": 930
    },
    {
      "entropy": 6.468717241287232,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.140625,
      "learning_rate": 0.000467,
      "loss": 6.2611,
      "mean_token_accuracy": 0.14346477389335632,
      "num_tokens": 2142364.0,
      "step": 935
    },
    {
      "entropy": 6.449568080902099,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004695,
      "loss": 6.1302,
      "mean_token_accuracy": 0.14927371144294738,
      "num_tokens": 2153040.0,
      "step": 940
    },
    {
      "entropy": 6.434846210479736,
      "epoch": 0.09077809798270893,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000472,
      "loss": 6.1895,
      "mean_token_accuracy": 0.15058233961462975,
      "num_tokens": 2165571.0,
      "step": 945
    },
    {
      "entropy": 6.456413507461548,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004745,
      "loss": 6.1612,
      "mean_token_accuracy": 0.14620548188686372,
      "num_tokens": 2177241.0,
      "step": 950
    },
    {
      "entropy": 6.455846166610717,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477,
      "loss": 6.1925,
      "mean_token_accuracy": 0.13922726288437842,
      "num_tokens": 2187475.0,
      "step": 955
    },
    {
      "entropy": 6.435314750671386,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004795,
      "loss": 6.2183,
      "mean_token_accuracy": 0.1472764365375042,
      "num_tokens": 2198185.0,
      "step": 960
    },
    {
      "entropy": 6.411659860610962,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.125,
      "learning_rate": 0.000482,
      "loss": 6.1554,
      "mean_token_accuracy": 0.14865822941064835,
      "num_tokens": 2210404.0,
      "step": 965
    },
    {
      "entropy": 6.521021556854248,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004845,
      "loss": 6.1382,
      "mean_token_accuracy": 0.14820332527160646,
      "num_tokens": 2222188.0,
      "step": 970
    },
    {
      "entropy": 6.401497077941895,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.09375,
      "learning_rate": 0.000487,
      "loss": 6.1825,
      "mean_token_accuracy": 0.1503289483487606,
      "num_tokens": 2233418.0,
      "step": 975
    },
    {
      "entropy": 6.386258792877197,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004895,
      "loss": 6.1549,
      "mean_token_accuracy": 0.15336488783359528,
      "num_tokens": 2245053.0,
      "step": 980
    },
    {
      "entropy": 6.497596883773804,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.140625,
      "learning_rate": 0.000492,
      "loss": 6.1799,
      "mean_token_accuracy": 0.1518462046980858,
      "num_tokens": 2256375.0,
      "step": 985
    },
    {
      "entropy": 6.297069692611695,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004945,
      "loss": 6.0222,
      "mean_token_accuracy": 0.1536286234855652,
      "num_tokens": 2267074.0,
      "step": 990
    },
    {
      "entropy": 6.3477600574493405,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000497,
      "loss": 6.0739,
      "mean_token_accuracy": 0.15113484859466553,
      "num_tokens": 2277168.0,
      "step": 995
    },
    {
      "entropy": 6.337840461730957,
      "epoch": 0.09606147934678194,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995,
      "loss": 6.0562,
      "mean_token_accuracy": 0.15290841460227966,
      "num_tokens": 2288178.0,
      "step": 1000
    },
    {
      "entropy": 6.316500282287597,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.0619,
      "mean_token_accuracy": 0.1515717551112175,
      "num_tokens": 2299765.0,
      "step": 1005
    },
    {
      "entropy": 6.519514703750611,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.1835,
      "mean_token_accuracy": 0.13907498270273208,
      "num_tokens": 2312047.0,
      "step": 1010
    },
    {
      "entropy": 6.364095211029053,
      "epoch": 0.09750240153698367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.1611,
      "mean_token_accuracy": 0.14061366766691208,
      "num_tokens": 2324118.0,
      "step": 1015
    },
    {
      "entropy": 6.4445971012115475,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.1611,
      "mean_token_accuracy": 0.14967067837715148,
      "num_tokens": 2335171.0,
      "step": 1020
    },
    {
      "entropy": 6.398542642593384,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.0743,
      "mean_token_accuracy": 0.1492784783244133,
      "num_tokens": 2346338.0,
      "step": 1025
    },
    {
      "entropy": 6.369733762741089,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.1011,
      "mean_token_accuracy": 0.15191911309957504,
      "num_tokens": 2357185.0,
      "step": 1030
    },
    {
      "entropy": 6.378099775314331,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.0369,
      "mean_token_accuracy": 0.15464141294360162,
      "num_tokens": 2368494.0,
      "step": 1035
    },
    {
      "entropy": 6.379990863800049,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.2349,
      "mean_token_accuracy": 0.14129094332456588,
      "num_tokens": 2380800.0,
      "step": 1040
    },
    {
      "entropy": 6.467003870010376,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.1819,
      "mean_token_accuracy": 0.1439610406756401,
      "num_tokens": 2391753.0,
      "step": 1045
    },
    {
      "entropy": 6.27177791595459,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.0646,
      "mean_token_accuracy": 0.14684219062328338,
      "num_tokens": 2403324.0,
      "step": 1050
    },
    {
      "entropy": 6.4170433521270756,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.1568,
      "mean_token_accuracy": 0.1471158519387245,
      "num_tokens": 2414873.0,
      "step": 1055
    },
    {
      "entropy": 6.226349878311157,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999996363169212,
      "loss": 5.9522,
      "mean_token_accuracy": 0.15969930738210678,
      "num_tokens": 2425308.0,
      "step": 1060
    },
    {
      "entropy": 6.280011129379273,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999995720638266,
      "loss": 5.9956,
      "mean_token_accuracy": 0.15745754390954972,
      "num_tokens": 2436835.0,
      "step": 1065
    },
    {
      "entropy": 6.318831586837769,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.21875,
      "learning_rate": 0.00049999950258691,
      "loss": 6.1016,
      "mean_token_accuracy": 0.15324254631996154,
      "num_tokens": 2446798.0,
      "step": 1070
    },
    {
      "entropy": 6.259381628036499,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999994278861731,
      "loss": 6.0097,
      "mean_token_accuracy": 0.15251461267471314,
      "num_tokens": 2457308.0,
      "step": 1075
    },
    {
      "entropy": 6.345122861862182,
      "epoch": 0.1037463976945245,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.0775,
      "mean_token_accuracy": 0.14187726974487305,
      "num_tokens": 2468917.0,
      "step": 1080
    },
    {
      "entropy": 6.341993236541748,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.0947,
      "mean_token_accuracy": 0.145965027064085,
      "num_tokens": 2481363.0,
      "step": 1085
    },
    {
      "entropy": 6.322693538665772,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.0875,
      "mean_token_accuracy": 0.15120186507701874,
      "num_tokens": 2493082.0,
      "step": 1090
    },
    {
      "entropy": 6.3057647228240965,
      "epoch": 0.10518731988472622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.0435,
      "mean_token_accuracy": 0.15195624977350236,
      "num_tokens": 2503849.0,
      "step": 1095
    },
    {
      "entropy": 6.302676057815551,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.0935,
      "mean_token_accuracy": 0.15190254151821136,
      "num_tokens": 2514528.0,
      "step": 1100
    },
    {
      "entropy": 6.300235176086426,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999988699816299,
      "loss": 6.0734,
      "mean_token_accuracy": 0.15657310485839843,
      "num_tokens": 2524971.0,
      "step": 1105
    },
    {
      "entropy": 6.231389999389648,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999987587142058,
      "loss": 5.9787,
      "mean_token_accuracy": 0.1494237408041954,
      "num_tokens": 2535674.0,
      "step": 1110
    },
    {
      "entropy": 6.4110383033752445,
      "epoch": 0.10710854947166186,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.2173,
      "mean_token_accuracy": 0.14188248962163924,
      "num_tokens": 2547108.0,
      "step": 1115
    },
    {
      "entropy": 6.270652484893799,
      "epoch": 0.10758885686839577,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.0312,
      "mean_token_accuracy": 0.14654332846403123,
      "num_tokens": 2559474.0,
      "step": 1120
    },
    {
      "entropy": 6.178351354598999,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999983935691265,
      "loss": 5.877,
      "mean_token_accuracy": 0.16631512492895126,
      "num_tokens": 2571264.0,
      "step": 1125
    },
    {
      "entropy": 6.240466594696045,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499998261406507,
      "loss": 6.0402,
      "mean_token_accuracy": 0.1550907276570797,
      "num_tokens": 2583731.0,
      "step": 1130
    },
    {
      "entropy": 6.455864334106446,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.0865,
      "mean_token_accuracy": 0.15334258675575257,
      "num_tokens": 2595497.0,
      "step": 1135
    },
    {
      "entropy": 6.156699085235596,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.0553,
      "mean_token_accuracy": 0.1544922299683094,
      "num_tokens": 2607358.0,
      "step": 1140
    },
    {
      "entropy": 6.313141107559204,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999978335759121,
      "loss": 5.9824,
      "mean_token_accuracy": 0.16003676876425743,
      "num_tokens": 2618936.0,
      "step": 1145
    },
    {
      "entropy": 6.253233003616333,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.1273,
      "mean_token_accuracy": 0.1448728360235691,
      "num_tokens": 2631840.0,
      "step": 1150
    },
    {
      "entropy": 6.283588409423828,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499997522236602,
      "loss": 6.0696,
      "mean_token_accuracy": 0.15917213410139083,
      "num_tokens": 2642412.0,
      "step": 1155
    },
    {
      "entropy": 6.2728049755096436,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.0413,
      "mean_token_accuracy": 0.15381678342819213,
      "num_tokens": 2653890.0,
      "step": 1160
    },
    {
      "entropy": 6.398951053619385,
      "epoch": 0.11191162343900096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.0976,
      "mean_token_accuracy": 0.1579543687403202,
      "num_tokens": 2664888.0,
      "step": 1165
    },
    {
      "entropy": 6.157506227493286,
      "epoch": 0.11239193083573487,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999970160493391,
      "loss": 6.0067,
      "mean_token_accuracy": 0.1552426412701607,
      "num_tokens": 2675550.0,
      "step": 1170
    },
    {
      "entropy": 6.209622144699097,
      "epoch": 0.11287223823246878,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999968368727209,
      "loss": 6.01,
      "mean_token_accuracy": 0.15791884660720826,
      "num_tokens": 2688022.0,
      "step": 1175
    },
    {
      "entropy": 6.314128446578979,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999966524723442,
      "loss": 6.0035,
      "mean_token_accuracy": 0.15661349296569824,
      "num_tokens": 2698737.0,
      "step": 1180
    },
    {
      "entropy": 6.17528281211853,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999964628482135,
      "loss": 5.9803,
      "mean_token_accuracy": 0.1625732421875,
      "num_tokens": 2709844.0,
      "step": 1185
    },
    {
      "entropy": 6.311813545227051,
      "epoch": 0.1143131604226705,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.0243,
      "mean_token_accuracy": 0.15992799401283264,
      "num_tokens": 2720273.0,
      "step": 1190
    },
    {
      "entropy": 6.231012916564941,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.125,
      "learning_rate": 0.000499996067928707,
      "loss": 6.0555,
      "mean_token_accuracy": 0.1537340298295021,
      "num_tokens": 2731354.0,
      "step": 1195
    },
    {
      "entropy": 6.250294923782349,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.0383,
      "mean_token_accuracy": 0.15496493577957154,
      "num_tokens": 2742966.0,
      "step": 1200
    },
    {
      "entropy": 6.222562694549561,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999956521142383,
      "loss": 5.9677,
      "mean_token_accuracy": 0.1614091530442238,
      "num_tokens": 2755010.0,
      "step": 1205
    },
    {
      "entropy": 6.238761901855469,
      "epoch": 0.11623439000960614,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999954363714051,
      "loss": 5.9963,
      "mean_token_accuracy": 0.1529179558157921,
      "num_tokens": 2766176.0,
      "step": 1210
    },
    {
      "entropy": 6.375168800354004,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.1416,
      "mean_token_accuracy": 0.1510852135717869,
      "num_tokens": 2777861.0,
      "step": 1215
    },
    {
      "entropy": 6.179434299468994,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499994989214566,
      "loss": 5.9563,
      "mean_token_accuracy": 0.15724676698446274,
      "num_tokens": 2788725.0,
      "step": 1220
    },
    {
      "entropy": 6.156388998031616,
      "epoch": 0.11767531219980788,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999947578005705,
      "loss": 5.9829,
      "mean_token_accuracy": 0.1558725953102112,
      "num_tokens": 2801613.0,
      "step": 1225
    },
    {
      "entropy": 6.306823587417602,
      "epoch": 0.11815561959654179,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.0591,
      "mean_token_accuracy": 0.15092146396636963,
      "num_tokens": 2812474.0,
      "step": 1230
    },
    {
      "entropy": 6.149747133255005,
      "epoch": 0.1186359269932757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999942793014544,
      "loss": 5.9613,
      "mean_token_accuracy": 0.15863570123910903,
      "num_tokens": 2823178.0,
      "step": 1235
    },
    {
      "entropy": 6.148617744445801,
      "epoch": 0.11911623439000961,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499994032216345,
      "loss": 5.9582,
      "mean_token_accuracy": 0.15689361244440078,
      "num_tokens": 2836486.0,
      "step": 1240
    },
    {
      "entropy": 6.2577982425689695,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999937799075422,
      "loss": 6.0325,
      "mean_token_accuracy": 0.1582626670598984,
      "num_tokens": 2847902.0,
      "step": 1245
    },
    {
      "entropy": 6.061917066574097,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499993522375052,
      "loss": 5.9155,
      "mean_token_accuracy": 0.15855048298835756,
      "num_tokens": 2859991.0,
      "step": 1250
    },
    {
      "entropy": 6.334481048583984,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999932596188802,
      "loss": 6.103,
      "mean_token_accuracy": 0.14992666617035866,
      "num_tokens": 2870269.0,
      "step": 1255
    },
    {
      "entropy": 6.16114387512207,
      "epoch": 0.12103746397694524,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999929916390331,
      "loss": 5.974,
      "mean_token_accuracy": 0.15265119820833206,
      "num_tokens": 2882191.0,
      "step": 1260
    },
    {
      "entropy": 6.185744333267212,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999927184355169,
      "loss": 5.9814,
      "mean_token_accuracy": 0.15251912623643876,
      "num_tokens": 2892775.0,
      "step": 1265
    },
    {
      "entropy": 6.141828775405884,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999924400083377,
      "loss": 5.9617,
      "mean_token_accuracy": 0.15787012949585916,
      "num_tokens": 2904750.0,
      "step": 1270
    },
    {
      "entropy": 6.286509704589844,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.0399,
      "mean_token_accuracy": 0.1519160158932209,
      "num_tokens": 2916150.0,
      "step": 1275
    },
    {
      "entropy": 6.238752746582032,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.0001,
      "mean_token_accuracy": 0.15371812880039215,
      "num_tokens": 2928452.0,
      "step": 1280
    },
    {
      "entropy": 6.158025789260864,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999915733848886,
      "loss": 5.999,
      "mean_token_accuracy": 0.15188550055027009,
      "num_tokens": 2940577.0,
      "step": 1285
    },
    {
      "entropy": 6.150114059448242,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499991274063124,
      "loss": 5.9678,
      "mean_token_accuracy": 0.1559542365372181,
      "num_tokens": 2952302.0,
      "step": 1290
    },
    {
      "entropy": 6.235889863967896,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999909695177301,
      "loss": 6.029,
      "mean_token_accuracy": 0.15738558918237686,
      "num_tokens": 2964611.0,
      "step": 1295
    },
    {
      "entropy": 6.222271394729614,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499990659748714,
      "loss": 6.0135,
      "mean_token_accuracy": 0.1533857174217701,
      "num_tokens": 2975668.0,
      "step": 1300
    },
    {
      "entropy": 6.2896746635437015,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999903447560828,
      "loss": 6.0706,
      "mean_token_accuracy": 0.15294826924800872,
      "num_tokens": 2987303.0,
      "step": 1305
    },
    {
      "entropy": 6.250040912628174,
      "epoch": 0.12584053794428435,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999900245398439,
      "loss": 5.9698,
      "mean_token_accuracy": 0.15828934013843537,
      "num_tokens": 3000400.0,
      "step": 1310
    },
    {
      "entropy": 6.187452220916748,
      "epoch": 0.12632084534101826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1555943712592125,
      "num_tokens": 3012336.0,
      "step": 1315
    },
    {
      "entropy": 6.189295625686645,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999893684365729,
      "loss": 5.9769,
      "mean_token_accuracy": 0.15539684668183326,
      "num_tokens": 3023004.0,
      "step": 1320
    },
    {
      "entropy": 6.154639005661011,
      "epoch": 0.12728146013448607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999890325495559,
      "loss": 6.0444,
      "mean_token_accuracy": 0.15357957929372787,
      "num_tokens": 3035147.0,
      "step": 1325
    },
    {
      "entropy": 6.1711499214172365,
      "epoch": 0.12776176753121998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.8703,
      "mean_token_accuracy": 0.15838308334350587,
      "num_tokens": 3045611.0,
      "step": 1330
    },
    {
      "entropy": 6.135772943496704,
      "epoch": 0.1282420749279539,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999883451047981,
      "loss": 5.8891,
      "mean_token_accuracy": 0.1591949164867401,
      "num_tokens": 3056420.0,
      "step": 1335
    },
    {
      "entropy": 6.1266110897064205,
      "epoch": 0.1287223823246878,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.8705,
      "mean_token_accuracy": 0.16175246238708496,
      "num_tokens": 3068770.0,
      "step": 1340
    },
    {
      "entropy": 6.193431091308594,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999876367657954,
      "loss": 6.0132,
      "mean_token_accuracy": 0.14821962267160416,
      "num_tokens": 3080806.0,
      "step": 1345
    },
    {
      "entropy": 6.241663312911987,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999872747609725,
      "loss": 6.0259,
      "mean_token_accuracy": 0.15041830465197564,
      "num_tokens": 3091769.0,
      "step": 1350
    },
    {
      "entropy": 6.161892700195312,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999869075326132,
      "loss": 5.9515,
      "mean_token_accuracy": 0.15427336543798448,
      "num_tokens": 3103121.0,
      "step": 1355
    },
    {
      "entropy": 6.167838478088379,
      "epoch": 0.13064361191162344,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499986535080726,
      "loss": 5.9257,
      "mean_token_accuracy": 0.1637582778930664,
      "num_tokens": 3115606.0,
      "step": 1360
    },
    {
      "entropy": 6.070174360275269,
      "epoch": 0.13112391930835735,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999861574053196,
      "loss": 5.8353,
      "mean_token_accuracy": 0.15959180593490602,
      "num_tokens": 3127961.0,
      "step": 1365
    },
    {
      "entropy": 6.0476092338562015,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999857745064027,
      "loss": 5.8378,
      "mean_token_accuracy": 0.16003572195768356,
      "num_tokens": 3138316.0,
      "step": 1370
    },
    {
      "entropy": 6.09722580909729,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499985386383984,
      "loss": 5.8327,
      "mean_token_accuracy": 0.16383897215127946,
      "num_tokens": 3150818.0,
      "step": 1375
    },
    {
      "entropy": 6.173034286499023,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999849930380729,
      "loss": 5.9896,
      "mean_token_accuracy": 0.15335828065872192,
      "num_tokens": 3162066.0,
      "step": 1380
    },
    {
      "entropy": 6.091350126266479,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999845944686781,
      "loss": 5.9441,
      "mean_token_accuracy": 0.15436633825302123,
      "num_tokens": 3172209.0,
      "step": 1385
    },
    {
      "entropy": 6.033679103851318,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.7728,
      "mean_token_accuracy": 0.17072101235389708,
      "num_tokens": 3183248.0,
      "step": 1390
    },
    {
      "entropy": 6.115644407272339,
      "epoch": 0.1340057636887608,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.8801,
      "mean_token_accuracy": 0.1629835158586502,
      "num_tokens": 3194748.0,
      "step": 1395
    },
    {
      "entropy": 6.074833250045776,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.8424,
      "mean_token_accuracy": 0.16948780417442322,
      "num_tokens": 3205669.0,
      "step": 1400
    },
    {
      "entropy": 6.105074739456176,
      "epoch": 0.13496637848222862,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999829479564518,
      "loss": 5.9361,
      "mean_token_accuracy": 0.1520385757088661,
      "num_tokens": 3216035.0,
      "step": 1405
    },
    {
      "entropy": 6.140946578979492,
      "epoch": 0.13544668587896252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499982523269781,
      "loss": 5.9206,
      "mean_token_accuracy": 0.15891541689634323,
      "num_tokens": 3227192.0,
      "step": 1410
    },
    {
      "entropy": 6.1240339279174805,
      "epoch": 0.13592699327569643,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999820933596842,
      "loss": 5.9534,
      "mean_token_accuracy": 0.1581968516111374,
      "num_tokens": 3240237.0,
      "step": 1415
    },
    {
      "entropy": 6.090185928344726,
      "epoch": 0.13640730067243034,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499981658226171,
      "loss": 5.8259,
      "mean_token_accuracy": 0.16865789741277695,
      "num_tokens": 3251963.0,
      "step": 1420
    },
    {
      "entropy": 6.063898086547852,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499981217869252,
      "loss": 5.9439,
      "mean_token_accuracy": 0.1649617299437523,
      "num_tokens": 3263101.0,
      "step": 1425
    },
    {
      "entropy": 6.1431561470031735,
      "epoch": 0.1373679154658982,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000499980772288937,
      "loss": 5.8366,
      "mean_token_accuracy": 0.16629794836044312,
      "num_tokens": 3275100.0,
      "step": 1430
    },
    {
      "entropy": 6.106657552719116,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999803214852367,
      "loss": 5.9272,
      "mean_token_accuracy": 0.15512363463640214,
      "num_tokens": 3287025.0,
      "step": 1435
    },
    {
      "entropy": 6.167088365554809,
      "epoch": 0.138328530259366,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999798654581613,
      "loss": 5.9322,
      "mean_token_accuracy": 0.15951334238052367,
      "num_tokens": 3299867.0,
      "step": 1440
    },
    {
      "entropy": 6.086508417129517,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.8733,
      "mean_token_accuracy": 0.15924615412950516,
      "num_tokens": 3311183.0,
      "step": 1445
    },
    {
      "entropy": 6.092142724990845,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999789377339279,
      "loss": 5.9375,
      "mean_token_accuracy": 0.1590658202767372,
      "num_tokens": 3322247.0,
      "step": 1450
    },
    {
      "entropy": 6.110898780822754,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.8514,
      "mean_token_accuracy": 0.16008831560611725,
      "num_tokens": 3333369.0,
      "step": 1455
    },
    {
      "entropy": 6.026701641082764,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999779891163231,
      "loss": 5.8907,
      "mean_token_accuracy": 0.1609531357884407,
      "num_tokens": 3345876.0,
      "step": 1460
    },
    {
      "entropy": 6.066957569122314,
      "epoch": 0.14073006724303555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999775069725339,
      "loss": 5.778,
      "mean_token_accuracy": 0.16528390794992448,
      "num_tokens": 3357323.0,
      "step": 1465
    },
    {
      "entropy": 6.024584341049194,
      "epoch": 0.14121037463976946,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499977019605435,
      "loss": 5.8666,
      "mean_token_accuracy": 0.1610391244292259,
      "num_tokens": 3367689.0,
      "step": 1470
    },
    {
      "entropy": 5.971542263031006,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.8225,
      "mean_token_accuracy": 0.16424132734537125,
      "num_tokens": 3379472.0,
      "step": 1475
    },
    {
      "entropy": 6.139827394485474,
      "epoch": 0.14217098943323728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.8726,
      "mean_token_accuracy": 0.15778491795063018,
      "num_tokens": 3390929.0,
      "step": 1480
    },
    {
      "entropy": 6.217739486694336,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.8688,
      "mean_token_accuracy": 0.16537885516881942,
      "num_tokens": 3401242.0,
      "step": 1485
    },
    {
      "entropy": 5.921372890472412,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.8606,
      "mean_token_accuracy": 0.16257801353931428,
      "num_tokens": 3411169.0,
      "step": 1490
    },
    {
      "entropy": 6.042217111587524,
      "epoch": 0.143611911623439,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.6965,
      "mean_token_accuracy": 0.16992068588733672,
      "num_tokens": 3423265.0,
      "step": 1495
    },
    {
      "entropy": 5.984377431869507,
      "epoch": 0.1440922190201729,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.8066,
      "mean_token_accuracy": 0.1671621523797512,
      "num_tokens": 3434793.0,
      "step": 1500
    },
    {
      "entropy": 5.961929225921631,
      "epoch": 0.14457252641690682,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499973461784038,
      "loss": 5.7414,
      "mean_token_accuracy": 0.17664009779691697,
      "num_tokens": 3445732.0,
      "step": 1505
    },
    {
      "entropy": 6.045278692245484,
      "epoch": 0.14505283381364073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9086,
      "mean_token_accuracy": 0.16037625521421434,
      "num_tokens": 3457090.0,
      "step": 1510
    },
    {
      "entropy": 6.174110746383667,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000499972398254517,
      "loss": 5.8945,
      "mean_token_accuracy": 0.15567544400691985,
      "num_tokens": 3468087.0,
      "step": 1515
    },
    {
      "entropy": 6.075960445404053,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499971858654964,
      "loss": 5.8393,
      "mean_token_accuracy": 0.16277710646390914,
      "num_tokens": 3478820.0,
      "step": 1520
    },
    {
      "entropy": 6.011015510559082,
      "epoch": 0.14649375600384246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.864,
      "mean_token_accuracy": 0.15902443826198578,
      "num_tokens": 3489878.0,
      "step": 1525
    },
    {
      "entropy": 6.096222686767578,
      "epoch": 0.14697406340057637,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.849,
      "mean_token_accuracy": 0.16232544481754302,
      "num_tokens": 3500944.0,
      "step": 1530
    },
    {
      "entropy": 5.966727018356323,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.8314,
      "mean_token_accuracy": 0.16640263050794601,
      "num_tokens": 3511383.0,
      "step": 1535
    },
    {
      "entropy": 6.095454359054566,
      "epoch": 0.14793467819404418,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999696480250929,
      "loss": 5.921,
      "mean_token_accuracy": 0.1610638305544853,
      "num_tokens": 3523300.0,
      "step": 1540
    },
    {
      "entropy": 6.080443000793457,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.8507,
      "mean_token_accuracy": 0.15561971068382263,
      "num_tokens": 3534371.0,
      "step": 1545
    },
    {
      "entropy": 6.039649724960327,
      "epoch": 0.148895292987512,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.8226,
      "mean_token_accuracy": 0.16583748608827592,
      "num_tokens": 3544847.0,
      "step": 1550
    },
    {
      "entropy": 6.0535636901855465,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.7929,
      "mean_token_accuracy": 0.16670662462711333,
      "num_tokens": 3555859.0,
      "step": 1555
    },
    {
      "entropy": 5.992221164703369,
      "epoch": 0.14985590778097982,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8236,
      "mean_token_accuracy": 0.166217839717865,
      "num_tokens": 3568283.0,
      "step": 1560
    },
    {
      "entropy": 5.987648916244507,
      "epoch": 0.15033621517771373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.8857,
      "mean_token_accuracy": 0.15877313315868377,
      "num_tokens": 3581442.0,
      "step": 1565
    },
    {
      "entropy": 6.158979511260986,
      "epoch": 0.15081652257444764,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049996617538674,
      "loss": 5.9279,
      "mean_token_accuracy": 0.15592343956232071,
      "num_tokens": 3594055.0,
      "step": 1570
    },
    {
      "entropy": 6.06167197227478,
      "epoch": 0.15129682997118155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.8897,
      "mean_token_accuracy": 0.15909540206193923,
      "num_tokens": 3605952.0,
      "step": 1575
    },
    {
      "entropy": 6.10714282989502,
      "epoch": 0.15177713736791545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.9099,
      "mean_token_accuracy": 0.1594424292445183,
      "num_tokens": 3618544.0,
      "step": 1580
    },
    {
      "entropy": 6.059627628326416,
      "epoch": 0.15225744476464936,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.8529,
      "mean_token_accuracy": 0.164482718706131,
      "num_tokens": 3630445.0,
      "step": 1585
    },
    {
      "entropy": 6.061960506439209,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.8683,
      "mean_token_accuracy": 0.16716394126415252,
      "num_tokens": 3642516.0,
      "step": 1590
    },
    {
      "entropy": 5.999404573440552,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.7843,
      "mean_token_accuracy": 0.1711465001106262,
      "num_tokens": 3654425.0,
      "step": 1595
    },
    {
      "entropy": 5.989068651199341,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.8055,
      "mean_token_accuracy": 0.1635123535990715,
      "num_tokens": 3666521.0,
      "step": 1600
    },
    {
      "entropy": 6.065861177444458,
      "epoch": 0.15417867435158503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.8508,
      "mean_token_accuracy": 0.15598922967910767,
      "num_tokens": 3679121.0,
      "step": 1605
    },
    {
      "entropy": 6.048035669326782,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.8645,
      "mean_token_accuracy": 0.15967089980840682,
      "num_tokens": 3691095.0,
      "step": 1610
    },
    {
      "entropy": 5.985746097564697,
      "epoch": 0.15513928914505284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8272,
      "mean_token_accuracy": 0.16446878761053085,
      "num_tokens": 3703426.0,
      "step": 1615
    },
    {
      "entropy": 6.084816551208496,
      "epoch": 0.15561959654178675,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.8925,
      "mean_token_accuracy": 0.16607261896133424,
      "num_tokens": 3715429.0,
      "step": 1620
    },
    {
      "entropy": 6.091441202163696,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.8779,
      "mean_token_accuracy": 0.16095825433731079,
      "num_tokens": 3726327.0,
      "step": 1625
    },
    {
      "entropy": 6.048316621780396,
      "epoch": 0.15658021133525457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.8316,
      "mean_token_accuracy": 0.15777417570352553,
      "num_tokens": 3736818.0,
      "step": 1630
    },
    {
      "entropy": 6.080630207061768,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.8335,
      "mean_token_accuracy": 0.16248745918273927,
      "num_tokens": 3747776.0,
      "step": 1635
    },
    {
      "entropy": 5.891999816894531,
      "epoch": 0.1575408261287224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.6917,
      "mean_token_accuracy": 0.1702808439731598,
      "num_tokens": 3758990.0,
      "step": 1640
    },
    {
      "entropy": 5.972564077377319,
      "epoch": 0.1580211335254563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.7612,
      "mean_token_accuracy": 0.165286822617054,
      "num_tokens": 3769686.0,
      "step": 1645
    },
    {
      "entropy": 6.024297380447388,
      "epoch": 0.1585014409221902,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8409,
      "mean_token_accuracy": 0.16243061274290085,
      "num_tokens": 3781815.0,
      "step": 1650
    },
    {
      "entropy": 6.015322732925415,
      "epoch": 0.15898174831892412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.8773,
      "mean_token_accuracy": 0.1623807817697525,
      "num_tokens": 3793392.0,
      "step": 1655
    },
    {
      "entropy": 6.032832622528076,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.826,
      "mean_token_accuracy": 0.16092282980680467,
      "num_tokens": 3804974.0,
      "step": 1660
    },
    {
      "entropy": 5.871552133560181,
      "epoch": 0.15994236311239193,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049995393823994,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1742493450641632,
      "num_tokens": 3817166.0,
      "step": 1665
    },
    {
      "entropy": 6.0467840194702145,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.798,
      "mean_token_accuracy": 0.16097687929868698,
      "num_tokens": 3828151.0,
      "step": 1670
    },
    {
      "entropy": 5.955699825286866,
      "epoch": 0.16090297790585975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499952540439365,
      "loss": 5.81,
      "mean_token_accuracy": 0.164180888235569,
      "num_tokens": 3839439.0,
      "step": 1675
    },
    {
      "entropy": 6.0351283073425295,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.8468,
      "mean_token_accuracy": 0.16053693741559982,
      "num_tokens": 3851694.0,
      "step": 1680
    },
    {
      "entropy": 5.977187013626098,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049995112174787,
      "loss": 5.8331,
      "mean_token_accuracy": 0.1628563791513443,
      "num_tokens": 3863593.0,
      "step": 1685
    },
    {
      "entropy": 5.929252243041992,
      "epoch": 0.16234390009606148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.7497,
      "mean_token_accuracy": 0.17348818629980087,
      "num_tokens": 3874588.0,
      "step": 1690
    },
    {
      "entropy": 6.014736175537109,
      "epoch": 0.1628242074927954,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.8407,
      "mean_token_accuracy": 0.163246688246727,
      "num_tokens": 3884662.0,
      "step": 1695
    },
    {
      "entropy": 6.109212112426758,
      "epoch": 0.1633045148895293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.9364,
      "mean_token_accuracy": 0.1559630572795868,
      "num_tokens": 3896569.0,
      "step": 1700
    },
    {
      "entropy": 6.078909492492675,
      "epoch": 0.1637848222862632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.7962,
      "mean_token_accuracy": 0.16575282365083693,
      "num_tokens": 3907691.0,
      "step": 1705
    },
    {
      "entropy": 6.001922178268432,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.8099,
      "mean_token_accuracy": 0.16902267709374427,
      "num_tokens": 3918794.0,
      "step": 1710
    },
    {
      "entropy": 5.976260614395142,
      "epoch": 0.16474543707973102,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.7941,
      "mean_token_accuracy": 0.17384963780641555,
      "num_tokens": 3929773.0,
      "step": 1715
    },
    {
      "entropy": 5.973721170425415,
      "epoch": 0.16522574447646493,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.8159,
      "mean_token_accuracy": 0.16380798667669297,
      "num_tokens": 3940264.0,
      "step": 1720
    },
    {
      "entropy": 5.907494354248047,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7114,
      "mean_token_accuracy": 0.1722415491938591,
      "num_tokens": 3951500.0,
      "step": 1725
    },
    {
      "entropy": 6.023552942276001,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.7824,
      "mean_token_accuracy": 0.1692255288362503,
      "num_tokens": 3963580.0,
      "step": 1730
    },
    {
      "entropy": 5.8660492420196535,
      "epoch": 0.16666666666666666,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.7285,
      "mean_token_accuracy": 0.17743493169546126,
      "num_tokens": 3975994.0,
      "step": 1735
    },
    {
      "entropy": 6.00585618019104,
      "epoch": 0.16714697406340057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499942945525439,
      "loss": 5.745,
      "mean_token_accuracy": 0.16235984563827516,
      "num_tokens": 3987897.0,
      "step": 1740
    },
    {
      "entropy": 5.995152854919434,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.8941,
      "mean_token_accuracy": 0.15724924057722092,
      "num_tokens": 3999829.0,
      "step": 1745
    },
    {
      "entropy": 5.933386898040771,
      "epoch": 0.16810758885686838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.7299,
      "mean_token_accuracy": 0.17098538130521773,
      "num_tokens": 4010882.0,
      "step": 1750
    },
    {
      "entropy": 5.917968273162842,
      "epoch": 0.1685878962536023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.7206,
      "mean_token_accuracy": 0.16878145337104797,
      "num_tokens": 4021423.0,
      "step": 1755
    },
    {
      "entropy": 6.074040222167969,
      "epoch": 0.1690682036503362,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.8527,
      "mean_token_accuracy": 0.1553002655506134,
      "num_tokens": 4033590.0,
      "step": 1760
    },
    {
      "entropy": 5.8794393062591555,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.125,
      "learning_rate": 0.00049993902015901,
      "loss": 5.7225,
      "mean_token_accuracy": 0.17117420434951783,
      "num_tokens": 4043978.0,
      "step": 1765
    },
    {
      "entropy": 6.0071815013885494,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.8558,
      "mean_token_accuracy": 0.16220006048679353,
      "num_tokens": 4054513.0,
      "step": 1770
    },
    {
      "entropy": 6.068150615692138,
      "epoch": 0.17050912584053796,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.7949,
      "mean_token_accuracy": 0.16809019297361374,
      "num_tokens": 4066019.0,
      "step": 1775
    },
    {
      "entropy": 5.9036009311676025,
      "epoch": 0.17098943323727187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.7542,
      "mean_token_accuracy": 0.17212767601013185,
      "num_tokens": 4077688.0,
      "step": 1780
    },
    {
      "entropy": 6.022014522552491,
      "epoch": 0.17146974063400577,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.8467,
      "mean_token_accuracy": 0.16922961324453353,
      "num_tokens": 4089803.0,
      "step": 1785
    },
    {
      "entropy": 5.994244956970215,
      "epoch": 0.17195004803073968,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.7916,
      "mean_token_accuracy": 0.1677403450012207,
      "num_tokens": 4101969.0,
      "step": 1790
    },
    {
      "entropy": 5.974350070953369,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.7811,
      "mean_token_accuracy": 0.16280766427516938,
      "num_tokens": 4113567.0,
      "step": 1795
    },
    {
      "entropy": 5.9534135341644285,
      "epoch": 0.1729106628242075,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7357,
      "mean_token_accuracy": 0.1677995041012764,
      "num_tokens": 4125191.0,
      "step": 1800
    },
    {
      "entropy": 5.986348867416382,
      "epoch": 0.1733909702209414,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.7731,
      "mean_token_accuracy": 0.16389087438583375,
      "num_tokens": 4135050.0,
      "step": 1805
    },
    {
      "entropy": 5.944538688659668,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.8327,
      "mean_token_accuracy": 0.1667800709605217,
      "num_tokens": 4146874.0,
      "step": 1810
    },
    {
      "entropy": 5.925647783279419,
      "epoch": 0.17435158501440923,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.7567,
      "mean_token_accuracy": 0.16681902706623078,
      "num_tokens": 4158118.0,
      "step": 1815
    },
    {
      "entropy": 5.988029193878174,
      "epoch": 0.17483189241114314,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.7486,
      "mean_token_accuracy": 0.17824842482805253,
      "num_tokens": 4169035.0,
      "step": 1820
    },
    {
      "entropy": 5.898885297775268,
      "epoch": 0.17531219980787704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499929066597658,
      "loss": 5.7262,
      "mean_token_accuracy": 0.17178142070770264,
      "num_tokens": 4180314.0,
      "step": 1825
    },
    {
      "entropy": 6.002191352844238,
      "epoch": 0.17579250720461095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.8002,
      "mean_token_accuracy": 0.1674297496676445,
      "num_tokens": 4192238.0,
      "step": 1830
    },
    {
      "entropy": 5.8907599449157715,
      "epoch": 0.17627281460134486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.7252,
      "mean_token_accuracy": 0.16727315485477448,
      "num_tokens": 4203346.0,
      "step": 1835
    },
    {
      "entropy": 5.895097637176514,
      "epoch": 0.17675312199807877,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.7977,
      "mean_token_accuracy": 0.1713564246892929,
      "num_tokens": 4213763.0,
      "step": 1840
    },
    {
      "entropy": 5.920028400421143,
      "epoch": 0.17723342939481268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.754,
      "mean_token_accuracy": 0.16816631257534026,
      "num_tokens": 4225553.0,
      "step": 1845
    },
    {
      "entropy": 6.013068199157715,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499924697335629,
      "loss": 5.6628,
      "mean_token_accuracy": 0.17668255865573884,
      "num_tokens": 4236058.0,
      "step": 1850
    },
    {
      "entropy": 5.8163008213043215,
      "epoch": 0.1781940441882805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.7584,
      "mean_token_accuracy": 0.17156918495893478,
      "num_tokens": 4248299.0,
      "step": 1855
    },
    {
      "entropy": 5.956279039382935,
      "epoch": 0.1786743515850144,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.6868,
      "mean_token_accuracy": 0.17354945242404937,
      "num_tokens": 4259704.0,
      "step": 1860
    },
    {
      "entropy": 5.824860429763794,
      "epoch": 0.17915465898174832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499922013113246,
      "loss": 5.5708,
      "mean_token_accuracy": 0.1824245736002922,
      "num_tokens": 4270480.0,
      "step": 1865
    },
    {
      "entropy": 5.922730875015259,
      "epoch": 0.17963496637848222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.8211,
      "mean_token_accuracy": 0.16856807470321655,
      "num_tokens": 4282104.0,
      "step": 1870
    },
    {
      "entropy": 6.019510555267334,
      "epoch": 0.18011527377521613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.7974,
      "mean_token_accuracy": 0.16135748475790024,
      "num_tokens": 4294251.0,
      "step": 1875
    },
    {
      "entropy": 5.900371026992798,
      "epoch": 0.18059558117195004,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.6804,
      "mean_token_accuracy": 0.1701715335249901,
      "num_tokens": 4305569.0,
      "step": 1880
    },
    {
      "entropy": 5.941131639480591,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.8057,
      "mean_token_accuracy": 0.1641265444457531,
      "num_tokens": 4317845.0,
      "step": 1885
    },
    {
      "entropy": 5.888107252120972,
      "epoch": 0.18155619596541786,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.6809,
      "mean_token_accuracy": 0.17280717492103576,
      "num_tokens": 4329987.0,
      "step": 1890
    },
    {
      "entropy": 5.836309289932251,
      "epoch": 0.18203650336215177,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.6926,
      "mean_token_accuracy": 0.17172161340713502,
      "num_tokens": 4341628.0,
      "step": 1895
    },
    {
      "entropy": 5.965258646011352,
      "epoch": 0.18251681075888568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.7545,
      "mean_token_accuracy": 0.16519593745470046,
      "num_tokens": 4352379.0,
      "step": 1900
    },
    {
      "entropy": 5.800844955444336,
      "epoch": 0.1829971181556196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7045,
      "mean_token_accuracy": 0.17391971200704576,
      "num_tokens": 4364800.0,
      "step": 1905
    },
    {
      "entropy": 5.960105085372925,
      "epoch": 0.1834774255523535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.7164,
      "mean_token_accuracy": 0.1672212541103363,
      "num_tokens": 4376048.0,
      "step": 1910
    },
    {
      "entropy": 5.888760232925415,
      "epoch": 0.1839577329490874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.7058,
      "mean_token_accuracy": 0.17140705287456512,
      "num_tokens": 4388072.0,
      "step": 1915
    },
    {
      "entropy": 5.9937965869903564,
      "epoch": 0.1844380403458213,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.8004,
      "mean_token_accuracy": 0.1623227506875992,
      "num_tokens": 4399843.0,
      "step": 1920
    },
    {
      "entropy": 5.803120040893555,
      "epoch": 0.18491834774255522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.6574,
      "mean_token_accuracy": 0.17487162053585054,
      "num_tokens": 4411373.0,
      "step": 1925
    },
    {
      "entropy": 5.846405506134033,
      "epoch": 0.18539865513928913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499909838400118,
      "loss": 5.6564,
      "mean_token_accuracy": 0.17340814918279648,
      "num_tokens": 4421857.0,
      "step": 1930
    },
    {
      "entropy": 5.95341854095459,
      "epoch": 0.18587896253602307,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.7408,
      "mean_token_accuracy": 0.16888684779405594,
      "num_tokens": 4432728.0,
      "step": 1935
    },
    {
      "entropy": 5.844529104232788,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.6742,
      "mean_token_accuracy": 0.1769302397966385,
      "num_tokens": 4444683.0,
      "step": 1940
    },
    {
      "entropy": 5.96117753982544,
      "epoch": 0.18683957732949089,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.8016,
      "mean_token_accuracy": 0.16333086788654327,
      "num_tokens": 4456961.0,
      "step": 1945
    },
    {
      "entropy": 5.875500202178955,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.6392,
      "mean_token_accuracy": 0.17937855124473573,
      "num_tokens": 4468424.0,
      "step": 1950
    },
    {
      "entropy": 5.871504831314087,
      "epoch": 0.1878001921229587,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.8396,
      "mean_token_accuracy": 0.15889065861701965,
      "num_tokens": 4479813.0,
      "step": 1955
    },
    {
      "entropy": 6.061702728271484,
      "epoch": 0.1882804995196926,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.8364,
      "mean_token_accuracy": 0.16994454562664033,
      "num_tokens": 4491172.0,
      "step": 1960
    },
    {
      "entropy": 5.844229984283447,
      "epoch": 0.18876080691642652,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.7157,
      "mean_token_accuracy": 0.17361883521080018,
      "num_tokens": 4503063.0,
      "step": 1965
    },
    {
      "entropy": 5.998745012283325,
      "epoch": 0.18924111431316043,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.8005,
      "mean_token_accuracy": 0.1544879749417305,
      "num_tokens": 4514188.0,
      "step": 1970
    },
    {
      "entropy": 5.911010074615478,
      "epoch": 0.18972142170989434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7089,
      "mean_token_accuracy": 0.17505204379558564,
      "num_tokens": 4525293.0,
      "step": 1975
    },
    {
      "entropy": 5.906450510025024,
      "epoch": 0.19020172910662825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.785,
      "mean_token_accuracy": 0.16963134258985518,
      "num_tokens": 4536589.0,
      "step": 1980
    },
    {
      "entropy": 5.855934715270996,
      "epoch": 0.19068203650336216,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.5373,
      "mean_token_accuracy": 0.1919272780418396,
      "num_tokens": 4547594.0,
      "step": 1985
    },
    {
      "entropy": 5.856652116775512,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.7494,
      "mean_token_accuracy": 0.17164870202541352,
      "num_tokens": 4559850.0,
      "step": 1990
    },
    {
      "entropy": 5.886022758483887,
      "epoch": 0.19164265129682997,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1702727809548378,
      "num_tokens": 4570727.0,
      "step": 1995
    },
    {
      "entropy": 5.75973162651062,
      "epoch": 0.19212295869356388,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.6355,
      "mean_token_accuracy": 0.174292254447937,
      "num_tokens": 4582248.0,
      "step": 2000
    },
    {
      "entropy": 5.904182481765747,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.6548,
      "mean_token_accuracy": 0.17286083698272706,
      "num_tokens": 4592604.0,
      "step": 2005
    },
    {
      "entropy": 5.794434881210327,
      "epoch": 0.1930835734870317,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499893642700184,
      "loss": 5.6982,
      "mean_token_accuracy": 0.17326663583517074,
      "num_tokens": 4604398.0,
      "step": 2010
    },
    {
      "entropy": 5.889737224578857,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.7798,
      "mean_token_accuracy": 0.16947566717863083,
      "num_tokens": 4616434.0,
      "step": 2015
    },
    {
      "entropy": 5.953923463821411,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.7198,
      "mean_token_accuracy": 0.17970746457576753,
      "num_tokens": 4627577.0,
      "step": 2020
    },
    {
      "entropy": 5.960914373397827,
      "epoch": 0.19452449567723343,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.7831,
      "mean_token_accuracy": 0.1678330808877945,
      "num_tokens": 4639698.0,
      "step": 2025
    },
    {
      "entropy": 5.958297538757324,
      "epoch": 0.19500480307396734,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.7914,
      "mean_token_accuracy": 0.16515226066112518,
      "num_tokens": 4650643.0,
      "step": 2030
    },
    {
      "entropy": 5.87400336265564,
      "epoch": 0.19548511047070125,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7203,
      "mean_token_accuracy": 0.1693568304181099,
      "num_tokens": 4662897.0,
      "step": 2035
    },
    {
      "entropy": 5.9375998973846436,
      "epoch": 0.19596541786743515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.7942,
      "mean_token_accuracy": 0.17414003014564514,
      "num_tokens": 4673701.0,
      "step": 2040
    },
    {
      "entropy": 5.967000341415405,
      "epoch": 0.19644572526416906,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.7591,
      "mean_token_accuracy": 0.16811660975217818,
      "num_tokens": 4685873.0,
      "step": 2045
    },
    {
      "entropy": 5.925240468978882,
      "epoch": 0.19692603266090297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.6411,
      "mean_token_accuracy": 0.17622891664505005,
      "num_tokens": 4697179.0,
      "step": 2050
    },
    {
      "entropy": 5.831101989746093,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.6862,
      "mean_token_accuracy": 0.1738536462187767,
      "num_tokens": 4707752.0,
      "step": 2055
    },
    {
      "entropy": 5.746025085449219,
      "epoch": 0.1978866474543708,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.5372,
      "mean_token_accuracy": 0.18245317041873932,
      "num_tokens": 4718413.0,
      "step": 2060
    },
    {
      "entropy": 5.855678606033325,
      "epoch": 0.1983669548511047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499881732778379,
      "loss": 5.7395,
      "mean_token_accuracy": 0.16523946076631546,
      "num_tokens": 4730033.0,
      "step": 2065
    },
    {
      "entropy": 5.843948650360107,
      "epoch": 0.1988472622478386,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499880618731189,
      "loss": 5.586,
      "mean_token_accuracy": 0.17743310034275056,
      "num_tokens": 4742084.0,
      "step": 2070
    },
    {
      "entropy": 5.899618673324585,
      "epoch": 0.19932756964457252,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.8189,
      "mean_token_accuracy": 0.16559504121541976,
      "num_tokens": 4753885.0,
      "step": 2075
    },
    {
      "entropy": 5.831307220458984,
      "epoch": 0.19980787704130643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.6661,
      "mean_token_accuracy": 0.17495889812707902,
      "num_tokens": 4765686.0,
      "step": 2080
    },
    {
      "entropy": 5.899104452133178,
      "epoch": 0.20028818443804033,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7021,
      "mean_token_accuracy": 0.16899442821741104,
      "num_tokens": 4777157.0,
      "step": 2085
    },
    {
      "entropy": 5.828914642333984,
      "epoch": 0.20076849183477424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6397,
      "mean_token_accuracy": 0.17771946042776107,
      "num_tokens": 4788583.0,
      "step": 2090
    },
    {
      "entropy": 5.88235239982605,
      "epoch": 0.20124879923150815,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.755,
      "mean_token_accuracy": 0.16652097702026367,
      "num_tokens": 4800749.0,
      "step": 2095
    },
    {
      "entropy": 5.819902420043945,
      "epoch": 0.2017291066282421,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.6686,
      "mean_token_accuracy": 0.17305531948804856,
      "num_tokens": 4812488.0,
      "step": 2100
    },
    {
      "entropy": 5.922385215759277,
      "epoch": 0.202209414024976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6469,
      "mean_token_accuracy": 0.17342206686735154,
      "num_tokens": 4823495.0,
      "step": 2105
    },
    {
      "entropy": 5.770005416870117,
      "epoch": 0.2026897214217099,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.7124,
      "mean_token_accuracy": 0.16901031583547593,
      "num_tokens": 4834450.0,
      "step": 2110
    },
    {
      "entropy": 5.839266920089722,
      "epoch": 0.20317002881844382,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.6129,
      "mean_token_accuracy": 0.18550968468189238,
      "num_tokens": 4846826.0,
      "step": 2115
    },
    {
      "entropy": 5.970982599258423,
      "epoch": 0.20365033621517772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.7977,
      "mean_token_accuracy": 0.16186073422431946,
      "num_tokens": 4859668.0,
      "step": 2120
    },
    {
      "entropy": 5.834139776229859,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.6906,
      "mean_token_accuracy": 0.17062558084726334,
      "num_tokens": 4871727.0,
      "step": 2125
    },
    {
      "entropy": 5.955994033813477,
      "epoch": 0.20461095100864554,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.7366,
      "mean_token_accuracy": 0.1794666439294815,
      "num_tokens": 4882191.0,
      "step": 2130
    },
    {
      "entropy": 5.847929811477661,
      "epoch": 0.20509125840537945,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.6706,
      "mean_token_accuracy": 0.17324328422546387,
      "num_tokens": 4892416.0,
      "step": 2135
    },
    {
      "entropy": 5.728474712371826,
      "epoch": 0.20557156580211336,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.6471,
      "mean_token_accuracy": 0.16920143961906434,
      "num_tokens": 4903572.0,
      "step": 2140
    },
    {
      "entropy": 5.977149581909179,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.7446,
      "mean_token_accuracy": 0.17007820159196854,
      "num_tokens": 4915417.0,
      "step": 2145
    },
    {
      "entropy": 5.810878467559815,
      "epoch": 0.20653218059558118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.6138,
      "mean_token_accuracy": 0.17479929476976394,
      "num_tokens": 4926943.0,
      "step": 2150
    },
    {
      "entropy": 5.80943193435669,
      "epoch": 0.2070124879923151,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.5847,
      "mean_token_accuracy": 0.17787374258041383,
      "num_tokens": 4937724.0,
      "step": 2155
    },
    {
      "entropy": 5.957506895065308,
      "epoch": 0.207492795389049,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.7368,
      "mean_token_accuracy": 0.1671774461865425,
      "num_tokens": 4949970.0,
      "step": 2160
    },
    {
      "entropy": 5.856918811798096,
      "epoch": 0.2079731027857829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.7383,
      "mean_token_accuracy": 0.16863747537136078,
      "num_tokens": 4961389.0,
      "step": 2165
    },
    {
      "entropy": 5.930571174621582,
      "epoch": 0.2084534101825168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.787,
      "mean_token_accuracy": 0.16769310384988784,
      "num_tokens": 4973888.0,
      "step": 2170
    },
    {
      "entropy": 5.846202564239502,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.7358,
      "mean_token_accuracy": 0.16144881695508956,
      "num_tokens": 4985423.0,
      "step": 2175
    },
    {
      "entropy": 5.814085817337036,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.7026,
      "mean_token_accuracy": 0.16614219099283217,
      "num_tokens": 4997711.0,
      "step": 2180
    },
    {
      "entropy": 5.891442108154297,
      "epoch": 0.20989433237271854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17229223549365996,
      "num_tokens": 5009633.0,
      "step": 2185
    },
    {
      "entropy": 5.876784420013427,
      "epoch": 0.21037463976945245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.6628,
      "mean_token_accuracy": 0.17913604974746705,
      "num_tokens": 5021523.0,
      "step": 2190
    },
    {
      "entropy": 5.9075572967529295,
      "epoch": 0.21085494716618636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7561,
      "mean_token_accuracy": 0.1698111966252327,
      "num_tokens": 5033513.0,
      "step": 2195
    },
    {
      "entropy": 5.961580562591553,
      "epoch": 0.21133525456292027,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7608,
      "mean_token_accuracy": 0.16819328367710112,
      "num_tokens": 5047055.0,
      "step": 2200
    },
    {
      "entropy": 5.845474815368652,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.6379,
      "mean_token_accuracy": 0.17564828246831893,
      "num_tokens": 5057613.0,
      "step": 2205
    },
    {
      "entropy": 5.840123701095581,
      "epoch": 0.21229586935638808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.7558,
      "mean_token_accuracy": 0.16806377619504928,
      "num_tokens": 5068886.0,
      "step": 2210
    },
    {
      "entropy": 5.85970664024353,
      "epoch": 0.212776176753122,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.7283,
      "mean_token_accuracy": 0.167040803283453,
      "num_tokens": 5079978.0,
      "step": 2215
    },
    {
      "entropy": 5.902343082427978,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.6819,
      "mean_token_accuracy": 0.1737098440527916,
      "num_tokens": 5091021.0,
      "step": 2220
    },
    {
      "entropy": 5.76426329612732,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.5926,
      "mean_token_accuracy": 0.1794767439365387,
      "num_tokens": 5101483.0,
      "step": 2225
    },
    {
      "entropy": 5.874709224700927,
      "epoch": 0.21421709894332372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.712,
      "mean_token_accuracy": 0.16954506933689117,
      "num_tokens": 5113116.0,
      "step": 2230
    },
    {
      "entropy": 5.8735052108764645,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.6648,
      "mean_token_accuracy": 0.16956303715705873,
      "num_tokens": 5124824.0,
      "step": 2235
    },
    {
      "entropy": 5.7920280456542965,
      "epoch": 0.21517771373679154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.5907,
      "mean_token_accuracy": 0.1711305171251297,
      "num_tokens": 5137567.0,
      "step": 2240
    },
    {
      "entropy": 5.7892828464508055,
      "epoch": 0.21565802113352545,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6577,
      "mean_token_accuracy": 0.17161570936441423,
      "num_tokens": 5149016.0,
      "step": 2245
    },
    {
      "entropy": 5.763986873626709,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.5473,
      "mean_token_accuracy": 0.17783051282167434,
      "num_tokens": 5160356.0,
      "step": 2250
    },
    {
      "entropy": 5.848583841323853,
      "epoch": 0.21661863592699326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7387,
      "mean_token_accuracy": 0.16622594892978668,
      "num_tokens": 5172100.0,
      "step": 2255
    },
    {
      "entropy": 5.9171970844268795,
      "epoch": 0.21709894332372717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.7146,
      "mean_token_accuracy": 0.16663871705532074,
      "num_tokens": 5183984.0,
      "step": 2260
    },
    {
      "entropy": 5.842860412597656,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.6805,
      "mean_token_accuracy": 0.16949303299188614,
      "num_tokens": 5195853.0,
      "step": 2265
    },
    {
      "entropy": 5.7196276664733885,
      "epoch": 0.21805955811719502,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.5312,
      "mean_token_accuracy": 0.17934199124574662,
      "num_tokens": 5206995.0,
      "step": 2270
    },
    {
      "entropy": 5.686021423339843,
      "epoch": 0.21853986551392893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17936804592609407,
      "num_tokens": 5219291.0,
      "step": 2275
    },
    {
      "entropy": 5.80483112335205,
      "epoch": 0.21902017291066284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.6432,
      "mean_token_accuracy": 0.17933146953582763,
      "num_tokens": 5230856.0,
      "step": 2280
    },
    {
      "entropy": 5.694868040084839,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.5055,
      "mean_token_accuracy": 0.1853230893611908,
      "num_tokens": 5242871.0,
      "step": 2285
    },
    {
      "entropy": 5.757525062561035,
      "epoch": 0.21998078770413065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.6145,
      "mean_token_accuracy": 0.17693334370851516,
      "num_tokens": 5253835.0,
      "step": 2290
    },
    {
      "entropy": 5.757721900939941,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.5914,
      "mean_token_accuracy": 0.1778374120593071,
      "num_tokens": 5266195.0,
      "step": 2295
    },
    {
      "entropy": 5.770969820022583,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.5701,
      "mean_token_accuracy": 0.17990957647562028,
      "num_tokens": 5277499.0,
      "step": 2300
    },
    {
      "entropy": 5.856108522415161,
      "epoch": 0.22142170989433238,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499822369630386,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1632763683795929,
      "num_tokens": 5288622.0,
      "step": 2305
    },
    {
      "entropy": 5.837642621994019,
      "epoch": 0.2219020172910663,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.7363,
      "mean_token_accuracy": 0.1674153745174408,
      "num_tokens": 5299664.0,
      "step": 2310
    },
    {
      "entropy": 5.8238342761993405,
      "epoch": 0.2223823246878002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6326,
      "mean_token_accuracy": 0.1764681726694107,
      "num_tokens": 5311627.0,
      "step": 2315
    },
    {
      "entropy": 5.69834566116333,
      "epoch": 0.2228626320845341,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1816352590918541,
      "num_tokens": 5323000.0,
      "step": 2320
    },
    {
      "entropy": 5.7694638729095455,
      "epoch": 0.22334293948126802,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.5689,
      "mean_token_accuracy": 0.18260359317064284,
      "num_tokens": 5333811.0,
      "step": 2325
    },
    {
      "entropy": 5.779385423660278,
      "epoch": 0.22382324687800192,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499815494300587,
      "loss": 5.5901,
      "mean_token_accuracy": 0.17485559582710267,
      "num_tokens": 5344762.0,
      "step": 2330
    },
    {
      "entropy": 5.806505537033081,
      "epoch": 0.22430355427473583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.5754,
      "mean_token_accuracy": 0.18179302662611008,
      "num_tokens": 5356112.0,
      "step": 2335
    },
    {
      "entropy": 5.832508087158203,
      "epoch": 0.22478386167146974,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.6984,
      "mean_token_accuracy": 0.1719257727265358,
      "num_tokens": 5367339.0,
      "step": 2340
    },
    {
      "entropy": 5.7249836921691895,
      "epoch": 0.22526416906820365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.6361,
      "mean_token_accuracy": 0.18138026893138887,
      "num_tokens": 5378627.0,
      "step": 2345
    },
    {
      "entropy": 5.868769311904908,
      "epoch": 0.22574447646493756,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.6632,
      "mean_token_accuracy": 0.17547774761915208,
      "num_tokens": 5390209.0,
      "step": 2350
    },
    {
      "entropy": 5.81194429397583,
      "epoch": 0.22622478386167147,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.6538,
      "mean_token_accuracy": 0.1727759525179863,
      "num_tokens": 5401578.0,
      "step": 2355
    },
    {
      "entropy": 5.7616051197052,
      "epoch": 0.22670509125840538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.575,
      "mean_token_accuracy": 0.17865398079156875,
      "num_tokens": 5413831.0,
      "step": 2360
    },
    {
      "entropy": 5.82645959854126,
      "epoch": 0.2271853986551393,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6293,
      "mean_token_accuracy": 0.17196177393198014,
      "num_tokens": 5425430.0,
      "step": 2365
    },
    {
      "entropy": 5.588806056976319,
      "epoch": 0.2276657060518732,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.4715,
      "mean_token_accuracy": 0.18928515762090684,
      "num_tokens": 5435353.0,
      "step": 2370
    },
    {
      "entropy": 5.8202369689941404,
      "epoch": 0.2281460134486071,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.6808,
      "mean_token_accuracy": 0.1762748748064041,
      "num_tokens": 5446925.0,
      "step": 2375
    },
    {
      "entropy": 5.709089374542236,
      "epoch": 0.228626320845341,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.5578,
      "mean_token_accuracy": 0.18197699040174484,
      "num_tokens": 5456613.0,
      "step": 2380
    },
    {
      "entropy": 5.799993944168091,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17672255784273147,
      "num_tokens": 5467790.0,
      "step": 2385
    },
    {
      "entropy": 5.814759111404419,
      "epoch": 0.22958693563880883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499798461122933,
      "loss": 5.5703,
      "mean_token_accuracy": 0.1812552884221077,
      "num_tokens": 5479166.0,
      "step": 2390
    },
    {
      "entropy": 5.6896861553192135,
      "epoch": 0.23006724303554274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.5596,
      "mean_token_accuracy": 0.18635332137346267,
      "num_tokens": 5490186.0,
      "step": 2395
    },
    {
      "entropy": 5.738391304016114,
      "epoch": 0.23054755043227665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.5424,
      "mean_token_accuracy": 0.17801296710968018,
      "num_tokens": 5500416.0,
      "step": 2400
    },
    {
      "entropy": 5.766972875595092,
      "epoch": 0.23102785782901056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.6202,
      "mean_token_accuracy": 0.1773225635290146,
      "num_tokens": 5512189.0,
      "step": 2405
    },
    {
      "entropy": 5.711191701889038,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499792616381322,
      "loss": 5.4809,
      "mean_token_accuracy": 0.18552721440792083,
      "num_tokens": 5523631.0,
      "step": 2410
    },
    {
      "entropy": 5.715830135345459,
      "epoch": 0.23198847262247838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5429,
      "mean_token_accuracy": 0.1831541910767555,
      "num_tokens": 5535637.0,
      "step": 2415
    },
    {
      "entropy": 5.736978912353516,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499789662696278,
      "loss": 5.5562,
      "mean_token_accuracy": 0.1847958281636238,
      "num_tokens": 5546470.0,
      "step": 2420
    },
    {
      "entropy": 5.8512872695922855,
      "epoch": 0.2329490874159462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.7116,
      "mean_token_accuracy": 0.1715322643518448,
      "num_tokens": 5558633.0,
      "step": 2425
    },
    {
      "entropy": 5.771648311614991,
      "epoch": 0.2334293948126801,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.6272,
      "mean_token_accuracy": 0.17206788063049316,
      "num_tokens": 5570427.0,
      "step": 2430
    },
    {
      "entropy": 5.735009336471558,
      "epoch": 0.23390970220941404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499785193026685,
      "loss": 5.6008,
      "mean_token_accuracy": 0.17571118026971816,
      "num_tokens": 5580991.0,
      "step": 2435
    },
    {
      "entropy": 5.840723323822021,
      "epoch": 0.23439000960614795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.6681,
      "mean_token_accuracy": 0.17219943851232528,
      "num_tokens": 5592777.0,
      "step": 2440
    },
    {
      "entropy": 5.714860486984253,
      "epoch": 0.23487031700288186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.5575,
      "mean_token_accuracy": 0.1838330253958702,
      "num_tokens": 5603326.0,
      "step": 2445
    },
    {
      "entropy": 5.682803630828857,
      "epoch": 0.23535062439961577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.542,
      "mean_token_accuracy": 0.18075062334537506,
      "num_tokens": 5614504.0,
      "step": 2450
    },
    {
      "entropy": 5.7689934253692625,
      "epoch": 0.23583093179634967,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.6115,
      "mean_token_accuracy": 0.18290183395147325,
      "num_tokens": 5625150.0,
      "step": 2455
    },
    {
      "entropy": 5.732327699661255,
      "epoch": 0.23631123919308358,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17373958230018616,
      "num_tokens": 5636815.0,
      "step": 2460
    },
    {
      "entropy": 5.774140119552612,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.6077,
      "mean_token_accuracy": 0.17858083546161652,
      "num_tokens": 5648272.0,
      "step": 2465
    },
    {
      "entropy": 5.742091226577759,
      "epoch": 0.2372718539865514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18127477020025254,
      "num_tokens": 5659227.0,
      "step": 2470
    },
    {
      "entropy": 5.748807668685913,
      "epoch": 0.2377521613832853,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.6483,
      "mean_token_accuracy": 0.17652334049344062,
      "num_tokens": 5670411.0,
      "step": 2475
    },
    {
      "entropy": 5.703849220275879,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.4874,
      "mean_token_accuracy": 0.18095139116048814,
      "num_tokens": 5681570.0,
      "step": 2480
    },
    {
      "entropy": 5.688530683517456,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499769954908461,
      "loss": 5.4769,
      "mean_token_accuracy": 0.18896123319864272,
      "num_tokens": 5693021.0,
      "step": 2485
    },
    {
      "entropy": 5.6701000213623045,
      "epoch": 0.23919308357348704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.565,
      "mean_token_accuracy": 0.1783182293176651,
      "num_tokens": 5704043.0,
      "step": 2490
    },
    {
      "entropy": 5.731194639205933,
      "epoch": 0.23967339097022095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.6502,
      "mean_token_accuracy": 0.17267276644706725,
      "num_tokens": 5715735.0,
      "step": 2495
    },
    {
      "entropy": 5.862462711334229,
      "epoch": 0.24015369836695485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.6205,
      "mean_token_accuracy": 0.1759555384516716,
      "num_tokens": 5725778.0,
      "step": 2500
    },
    {
      "entropy": 5.732063245773316,
      "epoch": 0.24063400576368876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6296,
      "mean_token_accuracy": 0.17513870894908906,
      "num_tokens": 5737224.0,
      "step": 2505
    },
    {
      "entropy": 5.8137345790863035,
      "epoch": 0.24111431316042267,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.6583,
      "mean_token_accuracy": 0.17411693036556244,
      "num_tokens": 5749226.0,
      "step": 2510
    },
    {
      "entropy": 5.8183518409729,
      "epoch": 0.24159462055715658,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.6312,
      "mean_token_accuracy": 0.17504109144210817,
      "num_tokens": 5760282.0,
      "step": 2515
    },
    {
      "entropy": 5.663213491439819,
      "epoch": 0.2420749279538905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997589777234,
      "loss": 5.5309,
      "mean_token_accuracy": 0.1830235943198204,
      "num_tokens": 5771756.0,
      "step": 2520
    },
    {
      "entropy": 5.753743457794189,
      "epoch": 0.2425552353506244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.5671,
      "mean_token_accuracy": 0.1693316325545311,
      "num_tokens": 5783237.0,
      "step": 2525
    },
    {
      "entropy": 5.788538599014283,
      "epoch": 0.2430355427473583,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.6374,
      "mean_token_accuracy": 0.17653653472661973,
      "num_tokens": 5794825.0,
      "step": 2530
    },
    {
      "entropy": 5.7611926078796385,
      "epoch": 0.24351585014409222,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.5517,
      "mean_token_accuracy": 0.1838692843914032,
      "num_tokens": 5805578.0,
      "step": 2535
    },
    {
      "entropy": 5.798102807998657,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6398,
      "mean_token_accuracy": 0.17110283374786378,
      "num_tokens": 5818201.0,
      "step": 2540
    },
    {
      "entropy": 5.721447992324829,
      "epoch": 0.24447646493756003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.5639,
      "mean_token_accuracy": 0.1799464926123619,
      "num_tokens": 5830319.0,
      "step": 2545
    },
    {
      "entropy": 5.658524894714356,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5081,
      "mean_token_accuracy": 0.18168970346450805,
      "num_tokens": 5840638.0,
      "step": 2550
    },
    {
      "entropy": 5.725489377975464,
      "epoch": 0.24543707973102785,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.5401,
      "mean_token_accuracy": 0.17641098499298097,
      "num_tokens": 5852472.0,
      "step": 2555
    },
    {
      "entropy": 5.8019109725952145,
      "epoch": 0.24591738712776176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.5784,
      "mean_token_accuracy": 0.17450501769781113,
      "num_tokens": 5863455.0,
      "step": 2560
    },
    {
      "entropy": 5.637508058547974,
      "epoch": 0.24639769452449567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.4857,
      "mean_token_accuracy": 0.18074321746826172,
      "num_tokens": 5873141.0,
      "step": 2565
    },
    {
      "entropy": 5.688677740097046,
      "epoch": 0.24687800192122958,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.5911,
      "mean_token_accuracy": 0.1751626744866371,
      "num_tokens": 5884363.0,
      "step": 2570
    },
    {
      "entropy": 5.763658571243286,
      "epoch": 0.2473583093179635,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.6443,
      "mean_token_accuracy": 0.1757771074771881,
      "num_tokens": 5895856.0,
      "step": 2575
    },
    {
      "entropy": 5.676737451553345,
      "epoch": 0.2478386167146974,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.4743,
      "mean_token_accuracy": 0.1850918784737587,
      "num_tokens": 5906657.0,
      "step": 2580
    },
    {
      "entropy": 5.710775470733642,
      "epoch": 0.2483189241114313,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17889538109302522,
      "num_tokens": 5919496.0,
      "step": 2585
    },
    {
      "entropy": 5.750392150878906,
      "epoch": 0.24879923150816521,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.6576,
      "mean_token_accuracy": 0.16503879725933074,
      "num_tokens": 5932593.0,
      "step": 2590
    },
    {
      "entropy": 5.737170696258545,
      "epoch": 0.24927953890489912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.5772,
      "mean_token_accuracy": 0.17579198330640794,
      "num_tokens": 5944080.0,
      "step": 2595
    },
    {
      "entropy": 5.748187398910522,
      "epoch": 0.24975984630163303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.5536,
      "mean_token_accuracy": 0.18597260415554046,
      "num_tokens": 5955703.0,
      "step": 2600
    },
    {
      "entropy": 5.728941822052002,
      "epoch": 0.25024015369836694,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6317,
      "mean_token_accuracy": 0.17275256365537645,
      "num_tokens": 5966979.0,
      "step": 2605
    },
    {
      "entropy": 5.767943668365478,
      "epoch": 0.2507204610951009,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.5183,
      "mean_token_accuracy": 0.1809815689921379,
      "num_tokens": 5978808.0,
      "step": 2610
    },
    {
      "entropy": 5.720742416381836,
      "epoch": 0.25120076849183476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.6934,
      "mean_token_accuracy": 0.1722423776984215,
      "num_tokens": 5992145.0,
      "step": 2615
    },
    {
      "entropy": 5.732787466049194,
      "epoch": 0.2516810758885687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.5557,
      "mean_token_accuracy": 0.17480167895555496,
      "num_tokens": 6003723.0,
      "step": 2620
    },
    {
      "entropy": 5.763635778427124,
      "epoch": 0.2521613832853026,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.5786,
      "mean_token_accuracy": 0.18482747673988342,
      "num_tokens": 6014064.0,
      "step": 2625
    },
    {
      "entropy": 5.620661401748658,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.4966,
      "mean_token_accuracy": 0.18726035058498383,
      "num_tokens": 6025455.0,
      "step": 2630
    },
    {
      "entropy": 5.755416059494019,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.6073,
      "mean_token_accuracy": 0.16819422841072082,
      "num_tokens": 6037347.0,
      "step": 2635
    },
    {
      "entropy": 5.697003602981567,
      "epoch": 0.25360230547550433,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18332696408033372,
      "num_tokens": 6049236.0,
      "step": 2640
    },
    {
      "entropy": 5.591462659835815,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5117,
      "mean_token_accuracy": 0.1839895710349083,
      "num_tokens": 6060982.0,
      "step": 2645
    },
    {
      "entropy": 5.785092067718506,
      "epoch": 0.25456292026897215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.5093,
      "mean_token_accuracy": 0.183223856985569,
      "num_tokens": 6070925.0,
      "step": 2650
    },
    {
      "entropy": 5.637601613998413,
      "epoch": 0.25504322766570603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18813761025667192,
      "num_tokens": 6081279.0,
      "step": 2655
    },
    {
      "entropy": 5.6759350299835205,
      "epoch": 0.25552353506243997,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.5709,
      "mean_token_accuracy": 0.18180250078439714,
      "num_tokens": 6091797.0,
      "step": 2660
    },
    {
      "entropy": 5.803980493545533,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.5878,
      "mean_token_accuracy": 0.18288871198892592,
      "num_tokens": 6103435.0,
      "step": 2665
    },
    {
      "entropy": 5.7139026641845705,
      "epoch": 0.2564841498559078,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499709036684407,
      "loss": 5.6485,
      "mean_token_accuracy": 0.1780319407582283,
      "num_tokens": 6114531.0,
      "step": 2670
    },
    {
      "entropy": 5.635207653045654,
      "epoch": 0.25696445725264166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.4849,
      "mean_token_accuracy": 0.1845824345946312,
      "num_tokens": 6126110.0,
      "step": 2675
    },
    {
      "entropy": 5.746867656707764,
      "epoch": 0.2574447646493756,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.5801,
      "mean_token_accuracy": 0.17680004984140396,
      "num_tokens": 6137114.0,
      "step": 2680
    },
    {
      "entropy": 5.682720375061035,
      "epoch": 0.2579250720461095,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.5435,
      "mean_token_accuracy": 0.1826479434967041,
      "num_tokens": 6148696.0,
      "step": 2685
    },
    {
      "entropy": 5.668850469589233,
      "epoch": 0.2584053794428434,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.5377,
      "mean_token_accuracy": 0.18099813014268876,
      "num_tokens": 6160235.0,
      "step": 2690
    },
    {
      "entropy": 5.727720451354981,
      "epoch": 0.25888568683957736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.5204,
      "mean_token_accuracy": 0.18033260554075242,
      "num_tokens": 6172031.0,
      "step": 2695
    },
    {
      "entropy": 5.558167171478272,
      "epoch": 0.25936599423631124,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.3993,
      "mean_token_accuracy": 0.19067923128604888,
      "num_tokens": 6183547.0,
      "step": 2700
    },
    {
      "entropy": 5.730989408493042,
      "epoch": 0.2598463016330452,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.5889,
      "mean_token_accuracy": 0.17200068533420562,
      "num_tokens": 6194768.0,
      "step": 2705
    },
    {
      "entropy": 5.720772361755371,
      "epoch": 0.26032660902977905,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17187345325946807,
      "num_tokens": 6206099.0,
      "step": 2710
    },
    {
      "entropy": 5.782607173919677,
      "epoch": 0.260806916426513,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.585,
      "mean_token_accuracy": 0.18310263007879257,
      "num_tokens": 6217530.0,
      "step": 2715
    },
    {
      "entropy": 5.787333059310913,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1762217938899994,
      "num_tokens": 6228564.0,
      "step": 2720
    },
    {
      "entropy": 5.637451457977295,
      "epoch": 0.2617675312199808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499689548463118,
      "loss": 5.4915,
      "mean_token_accuracy": 0.17907891869544984,
      "num_tokens": 6239945.0,
      "step": 2725
    },
    {
      "entropy": 5.713913059234619,
      "epoch": 0.2622478386167147,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6031,
      "mean_token_accuracy": 0.17165847718715668,
      "num_tokens": 6251829.0,
      "step": 2730
    },
    {
      "entropy": 5.83690390586853,
      "epoch": 0.2627281460134486,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.6997,
      "mean_token_accuracy": 0.16278707683086396,
      "num_tokens": 6264823.0,
      "step": 2735
    },
    {
      "entropy": 5.652474355697632,
      "epoch": 0.2632084534101825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.3794,
      "mean_token_accuracy": 0.18814935386180878,
      "num_tokens": 6276247.0,
      "step": 2740
    },
    {
      "entropy": 5.562798309326172,
      "epoch": 0.26368876080691644,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.4945,
      "mean_token_accuracy": 0.18557627648115158,
      "num_tokens": 6287593.0,
      "step": 2745
    },
    {
      "entropy": 5.779765367507935,
      "epoch": 0.2641690682036503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.6677,
      "mean_token_accuracy": 0.17469124346971512,
      "num_tokens": 6299918.0,
      "step": 2750
    },
    {
      "entropy": 5.652080011367798,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5093,
      "mean_token_accuracy": 0.18570335656404496,
      "num_tokens": 6311147.0,
      "step": 2755
    },
    {
      "entropy": 5.57460036277771,
      "epoch": 0.26512968299711814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1906318724155426,
      "num_tokens": 6323239.0,
      "step": 2760
    },
    {
      "entropy": 5.755684471130371,
      "epoch": 0.2656099903938521,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5058,
      "mean_token_accuracy": 0.1824209213256836,
      "num_tokens": 6334625.0,
      "step": 2765
    },
    {
      "entropy": 5.676227712631226,
      "epoch": 0.26609029779058596,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499673134056204,
      "loss": 5.5537,
      "mean_token_accuracy": 0.17851219475269317,
      "num_tokens": 6346068.0,
      "step": 2770
    },
    {
      "entropy": 5.677485752105713,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.5272,
      "mean_token_accuracy": 0.17988724559545516,
      "num_tokens": 6357097.0,
      "step": 2775
    },
    {
      "entropy": 5.752181959152222,
      "epoch": 0.2670509125840538,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.5887,
      "mean_token_accuracy": 0.16941747516393663,
      "num_tokens": 6367481.0,
      "step": 2780
    },
    {
      "entropy": 5.618309497833252,
      "epoch": 0.2675312199807877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.4942,
      "mean_token_accuracy": 0.18622181862592696,
      "num_tokens": 6378873.0,
      "step": 2785
    },
    {
      "entropy": 5.736754608154297,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.5408,
      "mean_token_accuracy": 0.17920784205198287,
      "num_tokens": 6390651.0,
      "step": 2790
    },
    {
      "entropy": 5.613878059387207,
      "epoch": 0.26849183477425553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.4195,
      "mean_token_accuracy": 0.19438366293907167,
      "num_tokens": 6401631.0,
      "step": 2795
    },
    {
      "entropy": 5.575555801391602,
      "epoch": 0.2689721421709894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.505,
      "mean_token_accuracy": 0.18674360662698747,
      "num_tokens": 6413875.0,
      "step": 2800
    },
    {
      "entropy": 5.771028232574463,
      "epoch": 0.26945244956772335,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.541,
      "mean_token_accuracy": 0.17710736989974976,
      "num_tokens": 6425764.0,
      "step": 2805
    },
    {
      "entropy": 5.621025276184082,
      "epoch": 0.26993275696445723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.4427,
      "mean_token_accuracy": 0.18712297081947327,
      "num_tokens": 6437911.0,
      "step": 2810
    },
    {
      "entropy": 5.597506284713745,
      "epoch": 0.27041306436119117,
      "grad_norm": 1.125,
      "learning_rate": 0.00049965629713367,
      "loss": 5.4682,
      "mean_token_accuracy": 0.18105038553476333,
      "num_tokens": 6449942.0,
      "step": 2815
    },
    {
      "entropy": 5.707987642288208,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.5698,
      "mean_token_accuracy": 0.18570097982883454,
      "num_tokens": 6461729.0,
      "step": 2820
    },
    {
      "entropy": 5.729322052001953,
      "epoch": 0.271373679154659,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.553,
      "mean_token_accuracy": 0.18336791172623634,
      "num_tokens": 6472046.0,
      "step": 2825
    },
    {
      "entropy": 5.643385314941407,
      "epoch": 0.27185398655139287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.5346,
      "mean_token_accuracy": 0.17891884595155716,
      "num_tokens": 6483308.0,
      "step": 2830
    },
    {
      "entropy": 5.559644651412964,
      "epoch": 0.2723342939481268,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.4033,
      "mean_token_accuracy": 0.18978026360273362,
      "num_tokens": 6495093.0,
      "step": 2835
    },
    {
      "entropy": 5.695954370498657,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.4873,
      "mean_token_accuracy": 0.18494741916656493,
      "num_tokens": 6505933.0,
      "step": 2840
    },
    {
      "entropy": 5.6591239929199215,
      "epoch": 0.2732949087415946,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.5485,
      "mean_token_accuracy": 0.17702232003211976,
      "num_tokens": 6517280.0,
      "step": 2845
    },
    {
      "entropy": 5.6757430076599125,
      "epoch": 0.2737752161383285,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.5964,
      "mean_token_accuracy": 0.17370087206363677,
      "num_tokens": 6528980.0,
      "step": 2850
    },
    {
      "entropy": 5.688251590728759,
      "epoch": 0.27425552353506244,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.5803,
      "mean_token_accuracy": 0.18222061842679976,
      "num_tokens": 6540670.0,
      "step": 2855
    },
    {
      "entropy": 5.640368938446045,
      "epoch": 0.2747358309317964,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.545,
      "mean_token_accuracy": 0.1811893939971924,
      "num_tokens": 6551302.0,
      "step": 2860
    },
    {
      "entropy": 5.676407194137573,
      "epoch": 0.27521613832853026,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.4962,
      "mean_token_accuracy": 0.17508159577846527,
      "num_tokens": 6563177.0,
      "step": 2865
    },
    {
      "entropy": 5.826377010345459,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.6048,
      "mean_token_accuracy": 0.17976640462875365,
      "num_tokens": 6573323.0,
      "step": 2870
    },
    {
      "entropy": 5.664675855636597,
      "epoch": 0.2761767531219981,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.5386,
      "mean_token_accuracy": 0.17416994124650956,
      "num_tokens": 6585382.0,
      "step": 2875
    },
    {
      "entropy": 5.553443956375122,
      "epoch": 0.276657060518732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.4332,
      "mean_token_accuracy": 0.1913624182343483,
      "num_tokens": 6596629.0,
      "step": 2880
    },
    {
      "entropy": 5.703445148468018,
      "epoch": 0.2771373679154659,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499629266625136,
      "loss": 5.5392,
      "mean_token_accuracy": 0.1806390777230263,
      "num_tokens": 6608408.0,
      "step": 2885
    },
    {
      "entropy": 5.732683229446411,
      "epoch": 0.27761767531219983,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.549,
      "mean_token_accuracy": 0.1764448568224907,
      "num_tokens": 6619375.0,
      "step": 2890
    },
    {
      "entropy": 5.728315591812134,
      "epoch": 0.2780979827089337,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.6042,
      "mean_token_accuracy": 0.17850979417562485,
      "num_tokens": 6631317.0,
      "step": 2895
    },
    {
      "entropy": 5.662656164169311,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4637,
      "mean_token_accuracy": 0.18195123523473739,
      "num_tokens": 6642009.0,
      "step": 2900
    },
    {
      "entropy": 5.683246278762818,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.6133,
      "mean_token_accuracy": 0.172495736181736,
      "num_tokens": 6654713.0,
      "step": 2905
    },
    {
      "entropy": 5.6839197158813475,
      "epoch": 0.27953890489913547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.5781,
      "mean_token_accuracy": 0.1797851651906967,
      "num_tokens": 6667157.0,
      "step": 2910
    },
    {
      "entropy": 5.665137672424317,
      "epoch": 0.28001921229586935,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5158,
      "mean_token_accuracy": 0.1891197681427002,
      "num_tokens": 6679573.0,
      "step": 2915
    },
    {
      "entropy": 5.741504096984864,
      "epoch": 0.2804995196926033,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.5958,
      "mean_token_accuracy": 0.17464567720890045,
      "num_tokens": 6691768.0,
      "step": 2920
    },
    {
      "entropy": 5.721974849700928,
      "epoch": 0.28097982708933716,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049961336167259,
      "loss": 5.556,
      "mean_token_accuracy": 0.17392836958169938,
      "num_tokens": 6701964.0,
      "step": 2925
    },
    {
      "entropy": 5.678666019439698,
      "epoch": 0.2814601344860711,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5394,
      "mean_token_accuracy": 0.1783669501543045,
      "num_tokens": 6713506.0,
      "step": 2930
    },
    {
      "entropy": 5.680933046340942,
      "epoch": 0.281940441882805,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5647,
      "mean_token_accuracy": 0.1761838361620903,
      "num_tokens": 6724616.0,
      "step": 2935
    },
    {
      "entropy": 5.70354528427124,
      "epoch": 0.2824207492795389,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5319,
      "mean_token_accuracy": 0.18018208295106888,
      "num_tokens": 6735054.0,
      "step": 2940
    },
    {
      "entropy": 5.704903030395508,
      "epoch": 0.2829010566762728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.589,
      "mean_token_accuracy": 0.18162762224674225,
      "num_tokens": 6746756.0,
      "step": 2945
    },
    {
      "entropy": 5.67103385925293,
      "epoch": 0.28338136407300674,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.454,
      "mean_token_accuracy": 0.18400953710079193,
      "num_tokens": 6759566.0,
      "step": 2950
    },
    {
      "entropy": 5.649921560287476,
      "epoch": 0.2838616714697406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.4738,
      "mean_token_accuracy": 0.18114203959703445,
      "num_tokens": 6771375.0,
      "step": 2955
    },
    {
      "entropy": 5.739535665512085,
      "epoch": 0.28434197886647455,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.5935,
      "mean_token_accuracy": 0.1726598024368286,
      "num_tokens": 6783252.0,
      "step": 2960
    },
    {
      "entropy": 5.614906644821167,
      "epoch": 0.28482228626320844,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.5751,
      "mean_token_accuracy": 0.17417452335357667,
      "num_tokens": 6795525.0,
      "step": 2965
    },
    {
      "entropy": 5.7367006778717045,
      "epoch": 0.28530259365994237,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.4922,
      "mean_token_accuracy": 0.18741291463375093,
      "num_tokens": 6807212.0,
      "step": 2970
    },
    {
      "entropy": 5.595437860488891,
      "epoch": 0.28578290105667625,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1817941963672638,
      "num_tokens": 6819367.0,
      "step": 2975
    },
    {
      "entropy": 5.637096786499024,
      "epoch": 0.2862632084534102,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.5174,
      "mean_token_accuracy": 0.18265894204378127,
      "num_tokens": 6830405.0,
      "step": 2980
    },
    {
      "entropy": 5.62329068183899,
      "epoch": 0.28674351585014407,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.4952,
      "mean_token_accuracy": 0.1821579173207283,
      "num_tokens": 6841479.0,
      "step": 2985
    },
    {
      "entropy": 5.622311162948608,
      "epoch": 0.287223823246878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5304,
      "mean_token_accuracy": 0.1803388327360153,
      "num_tokens": 6851585.0,
      "step": 2990
    },
    {
      "entropy": 5.658223676681518,
      "epoch": 0.2877041306436119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.5059,
      "mean_token_accuracy": 0.1853910952806473,
      "num_tokens": 6863176.0,
      "step": 2995
    },
    {
      "entropy": 5.698240089416504,
      "epoch": 0.2881844380403458,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.5613,
      "mean_token_accuracy": 0.18125778138637544,
      "num_tokens": 6874021.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.4392175834856635,
      "eval_loss": 5.524839878082275,
      "eval_mean_token_accuracy": 0.1890813923326899,
      "eval_num_tokens": 6874021.0,
      "eval_runtime": 20.7338,
      "eval_samples_per_second": 1582.678,
      "eval_steps_per_second": 197.841,
      "step": 3000
    },
    {
      "entropy": 5.700993347167969,
      "epoch": 0.2886647454370797,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.5182,
      "mean_token_accuracy": 0.17984089851379395,
      "num_tokens": 6884999.0,
      "step": 3005
    },
    {
      "entropy": 5.678546047210693,
      "epoch": 0.28914505283381364,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.51,
      "mean_token_accuracy": 0.18251369148492813,
      "num_tokens": 6897021.0,
      "step": 3010
    },
    {
      "entropy": 5.577228927612305,
      "epoch": 0.2896253602305475,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.5079,
      "mean_token_accuracy": 0.1816476598381996,
      "num_tokens": 6908320.0,
      "step": 3015
    },
    {
      "entropy": 5.693873882293701,
      "epoch": 0.29010566762728146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18809738755226135,
      "num_tokens": 6919957.0,
      "step": 3020
    },
    {
      "entropy": 5.694812393188476,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.4474,
      "mean_token_accuracy": 0.1813945174217224,
      "num_tokens": 6930531.0,
      "step": 3025
    },
    {
      "entropy": 5.677816247940063,
      "epoch": 0.2910662824207493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.5802,
      "mean_token_accuracy": 0.17819592505693435,
      "num_tokens": 6942556.0,
      "step": 3030
    },
    {
      "entropy": 5.662592124938965,
      "epoch": 0.2915465898174832,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5308,
      "mean_token_accuracy": 0.18774571269750595,
      "num_tokens": 6954410.0,
      "step": 3035
    },
    {
      "entropy": 5.652567958831787,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.5406,
      "mean_token_accuracy": 0.17681581825017928,
      "num_tokens": 6964970.0,
      "step": 3040
    },
    {
      "entropy": 5.637769889831543,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.4309,
      "mean_token_accuracy": 0.1918710544705391,
      "num_tokens": 6976016.0,
      "step": 3045
    },
    {
      "entropy": 5.602601861953735,
      "epoch": 0.2929875120076849,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.4599,
      "mean_token_accuracy": 0.1865808606147766,
      "num_tokens": 6987199.0,
      "step": 3050
    },
    {
      "entropy": 5.6089927673339846,
      "epoch": 0.29346781940441885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.4809,
      "mean_token_accuracy": 0.19195563346147537,
      "num_tokens": 6999174.0,
      "step": 3055
    },
    {
      "entropy": 5.653221988677979,
      "epoch": 0.29394812680115273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499557218580945,
      "loss": 5.5526,
      "mean_token_accuracy": 0.1775237575173378,
      "num_tokens": 7012148.0,
      "step": 3060
    },
    {
      "entropy": 5.551672124862671,
      "epoch": 0.29442843419788667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1860169753432274,
      "num_tokens": 7023238.0,
      "step": 3065
    },
    {
      "entropy": 5.6040300846099855,
      "epoch": 0.29490874159462055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.5052,
      "mean_token_accuracy": 0.18213031440973282,
      "num_tokens": 7034944.0,
      "step": 3070
    },
    {
      "entropy": 5.660197782516479,
      "epoch": 0.2953890489913545,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499550745832149,
      "loss": 5.3774,
      "mean_token_accuracy": 0.1897702306509018,
      "num_tokens": 7046880.0,
      "step": 3075
    },
    {
      "entropy": 5.606032466888427,
      "epoch": 0.29586935638808837,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.4575,
      "mean_token_accuracy": 0.18246425986289977,
      "num_tokens": 7057678.0,
      "step": 3080
    },
    {
      "entropy": 5.530951976776123,
      "epoch": 0.2963496637848223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.4289,
      "mean_token_accuracy": 0.18984924107789994,
      "num_tokens": 7068336.0,
      "step": 3085
    },
    {
      "entropy": 5.612528896331787,
      "epoch": 0.2968299711815562,
      "grad_norm": 1.0,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5012,
      "mean_token_accuracy": 0.1788366675376892,
      "num_tokens": 7080341.0,
      "step": 3090
    },
    {
      "entropy": 5.7698705196380615,
      "epoch": 0.2973102785782901,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.6488,
      "mean_token_accuracy": 0.17575520128011704,
      "num_tokens": 7090618.0,
      "step": 3095
    },
    {
      "entropy": 5.592949104309082,
      "epoch": 0.297790585975024,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.3945,
      "mean_token_accuracy": 0.19016729593276976,
      "num_tokens": 7101843.0,
      "step": 3100
    },
    {
      "entropy": 5.547535753250122,
      "epoch": 0.29827089337175794,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.467,
      "mean_token_accuracy": 0.18589547425508499,
      "num_tokens": 7112014.0,
      "step": 3105
    },
    {
      "entropy": 5.645080947875977,
      "epoch": 0.2987512007684918,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.5408,
      "mean_token_accuracy": 0.17661455273628235,
      "num_tokens": 7123860.0,
      "step": 3110
    },
    {
      "entropy": 5.68672022819519,
      "epoch": 0.29923150816522576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.517,
      "mean_token_accuracy": 0.1843189463019371,
      "num_tokens": 7135901.0,
      "step": 3115
    },
    {
      "entropy": 5.60950345993042,
      "epoch": 0.29971181556195964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.4764,
      "mean_token_accuracy": 0.19353332966566086,
      "num_tokens": 7148743.0,
      "step": 3120
    },
    {
      "entropy": 5.700359869003296,
      "epoch": 0.3001921229586936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.522,
      "mean_token_accuracy": 0.1854911118745804,
      "num_tokens": 7160731.0,
      "step": 3125
    },
    {
      "entropy": 5.662644529342652,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5154,
      "mean_token_accuracy": 0.18299192190170288,
      "num_tokens": 7172861.0,
      "step": 3130
    },
    {
      "entropy": 5.613388013839722,
      "epoch": 0.3011527377521614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.4631,
      "mean_token_accuracy": 0.180388443171978,
      "num_tokens": 7183954.0,
      "step": 3135
    },
    {
      "entropy": 5.5486509799957275,
      "epoch": 0.3016330451488953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18174690902233123,
      "num_tokens": 7195466.0,
      "step": 3140
    },
    {
      "entropy": 5.612729692459107,
      "epoch": 0.3021133525456292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.422,
      "mean_token_accuracy": 0.19040560573339463,
      "num_tokens": 7207173.0,
      "step": 3145
    },
    {
      "entropy": 5.657872819900513,
      "epoch": 0.3025936599423631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.5693,
      "mean_token_accuracy": 0.18644312173128127,
      "num_tokens": 7220095.0,
      "step": 3150
    },
    {
      "entropy": 5.569757556915283,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.4857,
      "mean_token_accuracy": 0.1882634222507477,
      "num_tokens": 7230664.0,
      "step": 3155
    },
    {
      "entropy": 5.642200469970703,
      "epoch": 0.3035542747358309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.4315,
      "mean_token_accuracy": 0.18630266934633255,
      "num_tokens": 7241278.0,
      "step": 3160
    },
    {
      "entropy": 5.545270013809204,
      "epoch": 0.30403458213256485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.5928,
      "mean_token_accuracy": 0.1777158111333847,
      "num_tokens": 7252999.0,
      "step": 3165
    },
    {
      "entropy": 5.653474807739258,
      "epoch": 0.3045148895292987,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.4169,
      "mean_token_accuracy": 0.1864437311887741,
      "num_tokens": 7263949.0,
      "step": 3170
    },
    {
      "entropy": 5.639758348464966,
      "epoch": 0.30499519692603266,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.4714,
      "mean_token_accuracy": 0.18542171269655228,
      "num_tokens": 7275467.0,
      "step": 3175
    },
    {
      "entropy": 5.585669755935669,
      "epoch": 0.30547550432276654,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.4421,
      "mean_token_accuracy": 0.19193329364061357,
      "num_tokens": 7285534.0,
      "step": 3180
    },
    {
      "entropy": 5.6947675228118895,
      "epoch": 0.3059558117195005,
      "grad_norm": 1.0,
      "learning_rate": 0.000499501845410965,
      "loss": 5.5569,
      "mean_token_accuracy": 0.1815490484237671,
      "num_tokens": 7297252.0,
      "step": 3185
    },
    {
      "entropy": 5.607858324050904,
      "epoch": 0.30643611911623436,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.4299,
      "mean_token_accuracy": 0.19153623431921005,
      "num_tokens": 7308492.0,
      "step": 3190
    },
    {
      "entropy": 5.569702529907227,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.4581,
      "mean_token_accuracy": 0.18491947501897812,
      "num_tokens": 7319703.0,
      "step": 3195
    },
    {
      "entropy": 5.674600601196289,
      "epoch": 0.30739673390970224,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499494981699318,
      "loss": 5.4274,
      "mean_token_accuracy": 0.1937693625688553,
      "num_tokens": 7331022.0,
      "step": 3200
    },
    {
      "entropy": 5.56959900856018,
      "epoch": 0.3078770413064361,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499492683371151,
      "loss": 5.4772,
      "mean_token_accuracy": 0.18553764522075653,
      "num_tokens": 7342977.0,
      "step": 3205
    },
    {
      "entropy": 5.66892671585083,
      "epoch": 0.30835734870317005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.4562,
      "mean_token_accuracy": 0.18036793321371078,
      "num_tokens": 7353227.0,
      "step": 3210
    },
    {
      "entropy": 5.670237398147583,
      "epoch": 0.30883765609990393,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5121,
      "mean_token_accuracy": 0.1748569056391716,
      "num_tokens": 7364165.0,
      "step": 3215
    },
    {
      "entropy": 5.555734395980835,
      "epoch": 0.30931796349663787,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.4433,
      "mean_token_accuracy": 0.18816640377044677,
      "num_tokens": 7374800.0,
      "step": 3220
    },
    {
      "entropy": 5.496872520446777,
      "epoch": 0.30979827089337175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.3932,
      "mean_token_accuracy": 0.1932916134595871,
      "num_tokens": 7386360.0,
      "step": 3225
    },
    {
      "entropy": 5.606559085845947,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4317,
      "mean_token_accuracy": 0.18842349499464034,
      "num_tokens": 7397066.0,
      "step": 3230
    },
    {
      "entropy": 5.683773517608643,
      "epoch": 0.31075888568683957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.5188,
      "mean_token_accuracy": 0.1811457246541977,
      "num_tokens": 7408349.0,
      "step": 3235
    },
    {
      "entropy": 5.629186201095581,
      "epoch": 0.3112391930835735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.5533,
      "mean_token_accuracy": 0.17683551460504532,
      "num_tokens": 7420343.0,
      "step": 3240
    },
    {
      "entropy": 5.65942530632019,
      "epoch": 0.3117195004803074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.4622,
      "mean_token_accuracy": 0.18691066205501555,
      "num_tokens": 7431683.0,
      "step": 3245
    },
    {
      "entropy": 5.466778755187988,
      "epoch": 0.3121998078770413,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499471763881684,
      "loss": 5.3551,
      "mean_token_accuracy": 0.1865973576903343,
      "num_tokens": 7443327.0,
      "step": 3250
    },
    {
      "entropy": 5.634066247940064,
      "epoch": 0.3126801152737752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.4648,
      "mean_token_accuracy": 0.18632208704948425,
      "num_tokens": 7454002.0,
      "step": 3255
    },
    {
      "entropy": 5.543521547317505,
      "epoch": 0.31316042267050914,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499467057777609,
      "loss": 5.4556,
      "mean_token_accuracy": 0.18688408136367798,
      "num_tokens": 7464074.0,
      "step": 3260
    },
    {
      "entropy": 5.55138783454895,
      "epoch": 0.313640730067243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.3983,
      "mean_token_accuracy": 0.19326114803552627,
      "num_tokens": 7475543.0,
      "step": 3265
    },
    {
      "entropy": 5.635407209396362,
      "epoch": 0.31412103746397696,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.4507,
      "mean_token_accuracy": 0.18384750932455063,
      "num_tokens": 7486818.0,
      "step": 3270
    },
    {
      "entropy": 5.62159948348999,
      "epoch": 0.31460134486071084,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499459959535695,
      "loss": 5.4938,
      "mean_token_accuracy": 0.18008041083812715,
      "num_tokens": 7499046.0,
      "step": 3275
    },
    {
      "entropy": 5.600229930877686,
      "epoch": 0.3150816522574448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.4287,
      "mean_token_accuracy": 0.18267867863178253,
      "num_tokens": 7509853.0,
      "step": 3280
    },
    {
      "entropy": 5.477385759353638,
      "epoch": 0.31556195965417866,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.2819,
      "mean_token_accuracy": 0.1905076265335083,
      "num_tokens": 7521091.0,
      "step": 3285
    },
    {
      "entropy": 5.560299873352051,
      "epoch": 0.3160422670509126,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1956005111336708,
      "num_tokens": 7531317.0,
      "step": 3290
    },
    {
      "entropy": 5.59887285232544,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.5365,
      "mean_token_accuracy": 0.18225086703896523,
      "num_tokens": 7542822.0,
      "step": 3295
    },
    {
      "entropy": 5.675664710998535,
      "epoch": 0.3170028818443804,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.4889,
      "mean_token_accuracy": 0.17975055873394014,
      "num_tokens": 7552434.0,
      "step": 3300
    },
    {
      "entropy": 5.594845724105835,
      "epoch": 0.3174831892411143,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.3642,
      "mean_token_accuracy": 0.19091499149799346,
      "num_tokens": 7563895.0,
      "step": 3305
    },
    {
      "entropy": 5.409893989562988,
      "epoch": 0.31796349663784823,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.3848,
      "mean_token_accuracy": 0.19512586593627929,
      "num_tokens": 7575391.0,
      "step": 3310
    },
    {
      "entropy": 5.639042043685913,
      "epoch": 0.3184438040345821,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.5321,
      "mean_token_accuracy": 0.19067961126565933,
      "num_tokens": 7587119.0,
      "step": 3315
    },
    {
      "entropy": 5.650514125823975,
      "epoch": 0.31892411143131605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.5053,
      "mean_token_accuracy": 0.18139485269784927,
      "num_tokens": 7598615.0,
      "step": 3320
    },
    {
      "entropy": 5.662230205535889,
      "epoch": 0.31940441882804993,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.4847,
      "mean_token_accuracy": 0.19163189232349395,
      "num_tokens": 7610159.0,
      "step": 3325
    },
    {
      "entropy": 5.627119827270508,
      "epoch": 0.31988472622478387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.451,
      "mean_token_accuracy": 0.1873135432600975,
      "num_tokens": 7620922.0,
      "step": 3330
    },
    {
      "entropy": 5.546042776107788,
      "epoch": 0.32036503362151775,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.3591,
      "mean_token_accuracy": 0.19363973736763002,
      "num_tokens": 7632343.0,
      "step": 3335
    },
    {
      "entropy": 5.685207319259644,
      "epoch": 0.3208453410182517,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.568,
      "mean_token_accuracy": 0.17316349148750304,
      "num_tokens": 7644748.0,
      "step": 3340
    },
    {
      "entropy": 5.689049005508423,
      "epoch": 0.32132564841498557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499426214287049,
      "loss": 5.5029,
      "mean_token_accuracy": 0.19029539227485656,
      "num_tokens": 7655449.0,
      "step": 3345
    },
    {
      "entropy": 5.639818286895752,
      "epoch": 0.3218059558117195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.4696,
      "mean_token_accuracy": 0.18956416994333267,
      "num_tokens": 7665623.0,
      "step": 3350
    },
    {
      "entropy": 5.548650217056275,
      "epoch": 0.3222862632084534,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19147212505340577,
      "num_tokens": 7676860.0,
      "step": 3355
    },
    {
      "entropy": 5.647919225692749,
      "epoch": 0.3227665706051873,
      "grad_norm": 1.0,
      "learning_rate": 0.000499418850289057,
      "loss": 5.5559,
      "mean_token_accuracy": 0.18325860053300858,
      "num_tokens": 7687778.0,
      "step": 3360
    },
    {
      "entropy": 5.633452653884888,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.4644,
      "mean_token_accuracy": 0.18514435291290282,
      "num_tokens": 7699154.0,
      "step": 3365
    },
    {
      "entropy": 5.59450101852417,
      "epoch": 0.32372718539865514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18767904341220856,
      "num_tokens": 7711010.0,
      "step": 3370
    },
    {
      "entropy": 5.627596235275268,
      "epoch": 0.3242074927953891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18083676546812058,
      "num_tokens": 7721969.0,
      "step": 3375
    },
    {
      "entropy": 5.615005588531494,
      "epoch": 0.32468780019212296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.4211,
      "mean_token_accuracy": 0.18771904110908508,
      "num_tokens": 7733762.0,
      "step": 3380
    },
    {
      "entropy": 5.603797101974488,
      "epoch": 0.3251681075888569,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.4356,
      "mean_token_accuracy": 0.1839058965444565,
      "num_tokens": 7744614.0,
      "step": 3385
    },
    {
      "entropy": 5.560338926315308,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.3845,
      "mean_token_accuracy": 0.1955147460103035,
      "num_tokens": 7755799.0,
      "step": 3390
    },
    {
      "entropy": 5.513249158859253,
      "epoch": 0.3261287223823247,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049940148526788,
      "loss": 5.4396,
      "mean_token_accuracy": 0.1901847615838051,
      "num_tokens": 7768140.0,
      "step": 3395
    },
    {
      "entropy": 5.557335901260376,
      "epoch": 0.3266090297790586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4366,
      "mean_token_accuracy": 0.19033351689577102,
      "num_tokens": 7778494.0,
      "step": 3400
    },
    {
      "entropy": 5.683393812179565,
      "epoch": 0.3270893371757925,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.5064,
      "mean_token_accuracy": 0.1925898015499115,
      "num_tokens": 7789234.0,
      "step": 3405
    },
    {
      "entropy": 5.557097578048706,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.4937,
      "mean_token_accuracy": 0.18996243476867675,
      "num_tokens": 7800602.0,
      "step": 3410
    },
    {
      "entropy": 5.510222053527832,
      "epoch": 0.32804995196926034,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.3159,
      "mean_token_accuracy": 0.19364566057920457,
      "num_tokens": 7812803.0,
      "step": 3415
    },
    {
      "entropy": 5.749662780761719,
      "epoch": 0.3285302593659942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.614,
      "mean_token_accuracy": 0.17690031826496125,
      "num_tokens": 7825432.0,
      "step": 3420
    },
    {
      "entropy": 5.600550270080566,
      "epoch": 0.32901056676272816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18812805712223052,
      "num_tokens": 7837258.0,
      "step": 3425
    },
    {
      "entropy": 5.522640132904053,
      "epoch": 0.32949087415946204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.3049,
      "mean_token_accuracy": 0.1966621220111847,
      "num_tokens": 7847573.0,
      "step": 3430
    },
    {
      "entropy": 5.63411316871643,
      "epoch": 0.329971181556196,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.4645,
      "mean_token_accuracy": 0.18168777078390122,
      "num_tokens": 7857957.0,
      "step": 3435
    },
    {
      "entropy": 5.580896186828613,
      "epoch": 0.33045148895292986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.3706,
      "mean_token_accuracy": 0.1964847505092621,
      "num_tokens": 7867996.0,
      "step": 3440
    },
    {
      "entropy": 5.501073503494263,
      "epoch": 0.3309317963496638,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.4281,
      "mean_token_accuracy": 0.1835190922021866,
      "num_tokens": 7879245.0,
      "step": 3445
    },
    {
      "entropy": 5.646635055541992,
      "epoch": 0.3314121037463977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.5,
      "mean_token_accuracy": 0.17995858192443848,
      "num_tokens": 7891186.0,
      "step": 3450
    },
    {
      "entropy": 5.6572953224182125,
      "epoch": 0.3318924111431316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.5248,
      "mean_token_accuracy": 0.19139863401651383,
      "num_tokens": 7902231.0,
      "step": 3455
    },
    {
      "entropy": 5.600442361831665,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.488,
      "mean_token_accuracy": 0.18677273094654084,
      "num_tokens": 7913364.0,
      "step": 3460
    },
    {
      "entropy": 5.820091962814331,
      "epoch": 0.33285302593659943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.6781,
      "mean_token_accuracy": 0.1703316792845726,
      "num_tokens": 7925217.0,
      "step": 3465
    },
    {
      "entropy": 5.581819820404053,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.4198,
      "mean_token_accuracy": 0.1796149030327797,
      "num_tokens": 7937636.0,
      "step": 3470
    },
    {
      "entropy": 5.565121173858643,
      "epoch": 0.33381364073006725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4536,
      "mean_token_accuracy": 0.18050845563411713,
      "num_tokens": 7948958.0,
      "step": 3475
    },
    {
      "entropy": 5.6448219299316404,
      "epoch": 0.33429394812680113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.4741,
      "mean_token_accuracy": 0.1872809574007988,
      "num_tokens": 7960013.0,
      "step": 3480
    },
    {
      "entropy": 5.595401859283447,
      "epoch": 0.33477425552353507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.5031,
      "mean_token_accuracy": 0.18324995785951614,
      "num_tokens": 7971041.0,
      "step": 3485
    },
    {
      "entropy": 5.632128286361694,
      "epoch": 0.33525456292026895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.5098,
      "mean_token_accuracy": 0.17839510440826417,
      "num_tokens": 7982752.0,
      "step": 3490
    },
    {
      "entropy": 5.637563943862915,
      "epoch": 0.3357348703170029,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.3971,
      "mean_token_accuracy": 0.18504412919282914,
      "num_tokens": 7993681.0,
      "step": 3495
    },
    {
      "entropy": 5.538132953643799,
      "epoch": 0.33621517771373677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.4184,
      "mean_token_accuracy": 0.18824413269758225,
      "num_tokens": 8004727.0,
      "step": 3500
    },
    {
      "entropy": 5.585167694091797,
      "epoch": 0.3366954851104707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.4432,
      "mean_token_accuracy": 0.1928482621908188,
      "num_tokens": 8015423.0,
      "step": 3505
    },
    {
      "entropy": 5.588391065597534,
      "epoch": 0.3371757925072046,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.4039,
      "mean_token_accuracy": 0.19309359788894653,
      "num_tokens": 8027911.0,
      "step": 3510
    },
    {
      "entropy": 5.594974565505981,
      "epoch": 0.3376560999039385,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4243,
      "mean_token_accuracy": 0.18888370990753173,
      "num_tokens": 8038831.0,
      "step": 3515
    },
    {
      "entropy": 5.477750301361084,
      "epoch": 0.3381364073006724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.3115,
      "mean_token_accuracy": 0.19790522754192352,
      "num_tokens": 8049906.0,
      "step": 3520
    },
    {
      "entropy": 5.484724855422973,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.4212,
      "mean_token_accuracy": 0.18728912770748138,
      "num_tokens": 8061158.0,
      "step": 3525
    },
    {
      "entropy": 5.6147716522216795,
      "epoch": 0.3390970220941403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.3792,
      "mean_token_accuracy": 0.18561409711837767,
      "num_tokens": 8071958.0,
      "step": 3530
    },
    {
      "entropy": 5.567511987686157,
      "epoch": 0.33957732949087416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499329472516601,
      "loss": 5.38,
      "mean_token_accuracy": 0.18467946201562882,
      "num_tokens": 8084068.0,
      "step": 3535
    },
    {
      "entropy": 5.463061094284058,
      "epoch": 0.3400576368876081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.2918,
      "mean_token_accuracy": 0.20272473245859146,
      "num_tokens": 8096132.0,
      "step": 3540
    },
    {
      "entropy": 5.551108026504517,
      "epoch": 0.340537944284342,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.4552,
      "mean_token_accuracy": 0.1927696093916893,
      "num_tokens": 8107327.0,
      "step": 3545
    },
    {
      "entropy": 5.619629001617431,
      "epoch": 0.3410182516810759,
      "grad_norm": 1.0,
      "learning_rate": 0.000499321514656929,
      "loss": 5.4397,
      "mean_token_accuracy": 0.18920674920082092,
      "num_tokens": 8118584.0,
      "step": 3550
    },
    {
      "entropy": 5.625332069396973,
      "epoch": 0.3414985590778098,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.478,
      "mean_token_accuracy": 0.1815608412027359,
      "num_tokens": 8130081.0,
      "step": 3555
    },
    {
      "entropy": 5.557459163665771,
      "epoch": 0.34197886647454373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.441,
      "mean_token_accuracy": 0.18856000006198884,
      "num_tokens": 8141463.0,
      "step": 3560
    },
    {
      "entropy": 5.623035001754761,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.479,
      "mean_token_accuracy": 0.18447892665863036,
      "num_tokens": 8153863.0,
      "step": 3565
    },
    {
      "entropy": 5.488258695602417,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.3974,
      "mean_token_accuracy": 0.18443325012922288,
      "num_tokens": 8165695.0,
      "step": 3570
    },
    {
      "entropy": 5.602726173400879,
      "epoch": 0.34341978866474543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.457,
      "mean_token_accuracy": 0.1891469821333885,
      "num_tokens": 8176456.0,
      "step": 3575
    },
    {
      "entropy": 5.6182811737060545,
      "epoch": 0.34390009606147937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.4567,
      "mean_token_accuracy": 0.1888190761208534,
      "num_tokens": 8189050.0,
      "step": 3580
    },
    {
      "entropy": 5.482992887496948,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.3534,
      "mean_token_accuracy": 0.19064465761184693,
      "num_tokens": 8200011.0,
      "step": 3585
    },
    {
      "entropy": 5.55635633468628,
      "epoch": 0.3448607108549472,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499300064546858,
      "loss": 5.3727,
      "mean_token_accuracy": 0.19231539517641066,
      "num_tokens": 8211770.0,
      "step": 3590
    },
    {
      "entropy": 5.568660497665405,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.408,
      "mean_token_accuracy": 0.19162147790193557,
      "num_tokens": 8223582.0,
      "step": 3595
    },
    {
      "entropy": 5.642741394042969,
      "epoch": 0.345821325648415,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499294649942276,
      "loss": 5.5309,
      "mean_token_accuracy": 0.18496304154396057,
      "num_tokens": 8234336.0,
      "step": 3600
    },
    {
      "entropy": 5.624276733398437,
      "epoch": 0.3463016330451489,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4281,
      "mean_token_accuracy": 0.18882226347923278,
      "num_tokens": 8244605.0,
      "step": 3605
    },
    {
      "entropy": 5.608572101593017,
      "epoch": 0.3467819404418828,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.4808,
      "mean_token_accuracy": 0.18347105085849763,
      "num_tokens": 8255876.0,
      "step": 3610
    },
    {
      "entropy": 5.549061632156372,
      "epoch": 0.3472622478386167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.3814,
      "mean_token_accuracy": 0.19023108333349228,
      "num_tokens": 8266602.0,
      "step": 3615
    },
    {
      "entropy": 5.546348810195923,
      "epoch": 0.34774255523535064,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19489242136478424,
      "num_tokens": 8279804.0,
      "step": 3620
    },
    {
      "entropy": 5.629941320419311,
      "epoch": 0.3482228626320845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.4606,
      "mean_token_accuracy": 0.1908916175365448,
      "num_tokens": 8291020.0,
      "step": 3625
    },
    {
      "entropy": 5.618665075302124,
      "epoch": 0.34870317002881845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4171,
      "mean_token_accuracy": 0.19629807025194168,
      "num_tokens": 8302192.0,
      "step": 3630
    },
    {
      "entropy": 5.548134994506836,
      "epoch": 0.34918347742555234,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1896643429994583,
      "num_tokens": 8313649.0,
      "step": 3635
    },
    {
      "entropy": 5.500442028045654,
      "epoch": 0.34966378482228627,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.385,
      "mean_token_accuracy": 0.1900986060500145,
      "num_tokens": 8324694.0,
      "step": 3640
    },
    {
      "entropy": 5.660983419418335,
      "epoch": 0.35014409221902015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.5324,
      "mean_token_accuracy": 0.18272839337587357,
      "num_tokens": 8336517.0,
      "step": 3645
    },
    {
      "entropy": 5.593430089950561,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.4967,
      "mean_token_accuracy": 0.1796460419893265,
      "num_tokens": 8349001.0,
      "step": 3650
    },
    {
      "entropy": 5.585159015655518,
      "epoch": 0.35110470701248797,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.4108,
      "mean_token_accuracy": 0.1927414983510971,
      "num_tokens": 8361230.0,
      "step": 3655
    },
    {
      "entropy": 5.5338012218475345,
      "epoch": 0.3515850144092219,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.3264,
      "mean_token_accuracy": 0.19598029255867006,
      "num_tokens": 8372159.0,
      "step": 3660
    },
    {
      "entropy": 5.557580852508545,
      "epoch": 0.3520653218059558,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.4485,
      "mean_token_accuracy": 0.19024865478277206,
      "num_tokens": 8383228.0,
      "step": 3665
    },
    {
      "entropy": 5.60696702003479,
      "epoch": 0.3525456292026897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.3997,
      "mean_token_accuracy": 0.18926765471696855,
      "num_tokens": 8394582.0,
      "step": 3670
    },
    {
      "entropy": 5.584086465835571,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18775029927492143,
      "num_tokens": 8406784.0,
      "step": 3675
    },
    {
      "entropy": 5.534383869171142,
      "epoch": 0.35350624399615754,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.3932,
      "mean_token_accuracy": 0.194275338947773,
      "num_tokens": 8418405.0,
      "step": 3680
    },
    {
      "entropy": 5.582950401306152,
      "epoch": 0.3539865513928914,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3445,
      "mean_token_accuracy": 0.19232997596263884,
      "num_tokens": 8430432.0,
      "step": 3685
    },
    {
      "entropy": 5.566782712936401,
      "epoch": 0.35446685878962536,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.4912,
      "mean_token_accuracy": 0.18145429342985153,
      "num_tokens": 8442423.0,
      "step": 3690
    },
    {
      "entropy": 5.585980749130249,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18717589229345322,
      "num_tokens": 8454951.0,
      "step": 3695
    },
    {
      "entropy": 5.527453804016114,
      "epoch": 0.3554274735830932,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.3484,
      "mean_token_accuracy": 0.1946762263774872,
      "num_tokens": 8467758.0,
      "step": 3700
    },
    {
      "entropy": 5.585886812210083,
      "epoch": 0.3559077809798271,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4409,
      "mean_token_accuracy": 0.18967984169721602,
      "num_tokens": 8479728.0,
      "step": 3705
    },
    {
      "entropy": 5.6186473846435545,
      "epoch": 0.356388088376561,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.4939,
      "mean_token_accuracy": 0.18735775500535964,
      "num_tokens": 8492099.0,
      "step": 3710
    },
    {
      "entropy": 5.608113765716553,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.4456,
      "mean_token_accuracy": 0.1802011474967003,
      "num_tokens": 8504657.0,
      "step": 3715
    },
    {
      "entropy": 5.57549901008606,
      "epoch": 0.3573487031700288,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4066,
      "mean_token_accuracy": 0.19084483236074448,
      "num_tokens": 8514728.0,
      "step": 3720
    },
    {
      "entropy": 5.635108757019043,
      "epoch": 0.35782901056676275,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.5219,
      "mean_token_accuracy": 0.18584865778684617,
      "num_tokens": 8525588.0,
      "step": 3725
    },
    {
      "entropy": 5.581790542602539,
      "epoch": 0.35830931796349663,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4877,
      "mean_token_accuracy": 0.18325926661491393,
      "num_tokens": 8537468.0,
      "step": 3730
    },
    {
      "entropy": 5.597702932357788,
      "epoch": 0.35878962536023057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.3977,
      "mean_token_accuracy": 0.19316060990095138,
      "num_tokens": 8548598.0,
      "step": 3735
    },
    {
      "entropy": 5.409280014038086,
      "epoch": 0.35926993275696445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.2728,
      "mean_token_accuracy": 0.20413144379854203,
      "num_tokens": 8560301.0,
      "step": 3740
    },
    {
      "entropy": 5.53164644241333,
      "epoch": 0.3597502401536984,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.368,
      "mean_token_accuracy": 0.1967419356107712,
      "num_tokens": 8571186.0,
      "step": 3745
    },
    {
      "entropy": 5.545560312271118,
      "epoch": 0.36023054755043227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.3598,
      "mean_token_accuracy": 0.19761675149202346,
      "num_tokens": 8580725.0,
      "step": 3750
    },
    {
      "entropy": 5.526527690887451,
      "epoch": 0.3607108549471662,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.5078,
      "mean_token_accuracy": 0.1807180315256119,
      "num_tokens": 8594598.0,
      "step": 3755
    },
    {
      "entropy": 5.6951195240020756,
      "epoch": 0.3611911623439001,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.4955,
      "mean_token_accuracy": 0.18285927772521973,
      "num_tokens": 8606733.0,
      "step": 3760
    },
    {
      "entropy": 5.609960460662842,
      "epoch": 0.361671469740634,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.4313,
      "mean_token_accuracy": 0.1881916657090187,
      "num_tokens": 8618863.0,
      "step": 3765
    },
    {
      "entropy": 5.424804210662842,
      "epoch": 0.3621517771373679,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1962995320558548,
      "num_tokens": 8629270.0,
      "step": 3770
    },
    {
      "entropy": 5.53543701171875,
      "epoch": 0.36263208453410184,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.4789,
      "mean_token_accuracy": 0.19050425440073013,
      "num_tokens": 8641363.0,
      "step": 3775
    },
    {
      "entropy": 5.642627811431884,
      "epoch": 0.3631123919308357,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1923316612839699,
      "num_tokens": 8653233.0,
      "step": 3780
    },
    {
      "entropy": 5.533376264572143,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499190723117722,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18614102751016617,
      "num_tokens": 8665192.0,
      "step": 3785
    },
    {
      "entropy": 5.642783832550049,
      "epoch": 0.36407300672430354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.5196,
      "mean_token_accuracy": 0.18100242018699647,
      "num_tokens": 8677669.0,
      "step": 3790
    },
    {
      "entropy": 5.474308347702026,
      "epoch": 0.3645533141210375,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.314,
      "mean_token_accuracy": 0.19570468962192536,
      "num_tokens": 8688002.0,
      "step": 3795
    },
    {
      "entropy": 5.516477632522583,
      "epoch": 0.36503362151777136,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.367,
      "mean_token_accuracy": 0.18995878249406814,
      "num_tokens": 8698396.0,
      "step": 3800
    },
    {
      "entropy": 5.623310947418213,
      "epoch": 0.3655139289145053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.3885,
      "mean_token_accuracy": 0.19130971878767014,
      "num_tokens": 8711135.0,
      "step": 3805
    },
    {
      "entropy": 5.50729808807373,
      "epoch": 0.3659942363112392,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499176132348474,
      "loss": 5.4236,
      "mean_token_accuracy": 0.19352682530879975,
      "num_tokens": 8723707.0,
      "step": 3810
    },
    {
      "entropy": 5.572604703903198,
      "epoch": 0.3664745437079731,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.3627,
      "mean_token_accuracy": 0.1964678570628166,
      "num_tokens": 8734193.0,
      "step": 3815
    },
    {
      "entropy": 5.517657279968262,
      "epoch": 0.366954851104707,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.4572,
      "mean_token_accuracy": 0.18392777889966966,
      "num_tokens": 8745619.0,
      "step": 3820
    },
    {
      "entropy": 5.462949419021607,
      "epoch": 0.36743515850144093,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.3582,
      "mean_token_accuracy": 0.19223012924194335,
      "num_tokens": 8757331.0,
      "step": 3825
    },
    {
      "entropy": 5.527809762954712,
      "epoch": 0.3679154658981748,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.3655,
      "mean_token_accuracy": 0.19106081277132034,
      "num_tokens": 8768840.0,
      "step": 3830
    },
    {
      "entropy": 5.562812089920044,
      "epoch": 0.36839577329490875,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.3195,
      "mean_token_accuracy": 0.19561284482479097,
      "num_tokens": 8781214.0,
      "step": 3835
    },
    {
      "entropy": 5.584686708450318,
      "epoch": 0.3688760806916426,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.4033,
      "mean_token_accuracy": 0.1916833609342575,
      "num_tokens": 8791645.0,
      "step": 3840
    },
    {
      "entropy": 5.599692010879517,
      "epoch": 0.36935638808837656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17887818589806556,
      "num_tokens": 8803286.0,
      "step": 3845
    },
    {
      "entropy": 5.589219093322754,
      "epoch": 0.36983669548511044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3662,
      "mean_token_accuracy": 0.18918564915657043,
      "num_tokens": 8814207.0,
      "step": 3850
    },
    {
      "entropy": 5.53467755317688,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.417,
      "mean_token_accuracy": 0.18368727713823318,
      "num_tokens": 8825540.0,
      "step": 3855
    },
    {
      "entropy": 5.591689157485962,
      "epoch": 0.37079731027857826,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.547,
      "mean_token_accuracy": 0.17688291668891906,
      "num_tokens": 8838182.0,
      "step": 3860
    },
    {
      "entropy": 5.570683383941651,
      "epoch": 0.3712776176753122,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.4198,
      "mean_token_accuracy": 0.19610581248998643,
      "num_tokens": 8850211.0,
      "step": 3865
    },
    {
      "entropy": 5.517876291275025,
      "epoch": 0.37175792507204614,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.457,
      "mean_token_accuracy": 0.182083423435688,
      "num_tokens": 8862804.0,
      "step": 3870
    },
    {
      "entropy": 5.559181594848633,
      "epoch": 0.37223823246878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.405,
      "mean_token_accuracy": 0.18535222709178925,
      "num_tokens": 8874112.0,
      "step": 3875
    },
    {
      "entropy": 5.540090560913086,
      "epoch": 0.37271853986551395,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4274,
      "mean_token_accuracy": 0.18555488735437392,
      "num_tokens": 8885114.0,
      "step": 3880
    },
    {
      "entropy": 5.529178762435913,
      "epoch": 0.37319884726224783,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19386369735002518,
      "num_tokens": 8895555.0,
      "step": 3885
    },
    {
      "entropy": 5.464783573150635,
      "epoch": 0.37367915465898177,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.3138,
      "mean_token_accuracy": 0.20528001785278321,
      "num_tokens": 8906073.0,
      "step": 3890
    },
    {
      "entropy": 5.548325109481811,
      "epoch": 0.37415946205571565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.4661,
      "mean_token_accuracy": 0.18886378407478333,
      "num_tokens": 8916587.0,
      "step": 3895
    },
    {
      "entropy": 5.586668920516968,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3657,
      "mean_token_accuracy": 0.1910933658480644,
      "num_tokens": 8927923.0,
      "step": 3900
    },
    {
      "entropy": 5.475469779968262,
      "epoch": 0.37512007684918347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499119500825865,
      "loss": 5.405,
      "mean_token_accuracy": 0.1913732722401619,
      "num_tokens": 8939939.0,
      "step": 3905
    },
    {
      "entropy": 5.515975379943848,
      "epoch": 0.3756003842459174,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.3909,
      "mean_token_accuracy": 0.19644259959459304,
      "num_tokens": 8951748.0,
      "step": 3910
    },
    {
      "entropy": 5.5175879955291744,
      "epoch": 0.3760806916426513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.2979,
      "mean_token_accuracy": 0.201249860227108,
      "num_tokens": 8962922.0,
      "step": 3915
    },
    {
      "entropy": 5.4236226081848145,
      "epoch": 0.3765609990393852,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1982575923204422,
      "num_tokens": 8974013.0,
      "step": 3920
    },
    {
      "entropy": 5.478136730194092,
      "epoch": 0.3770413064361191,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499107339028103,
      "loss": 5.3664,
      "mean_token_accuracy": 0.19188770800828933,
      "num_tokens": 8986032.0,
      "step": 3925
    },
    {
      "entropy": 5.5743492603302,
      "epoch": 0.37752161383285304,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.3564,
      "mean_token_accuracy": 0.19044692516326905,
      "num_tokens": 8996889.0,
      "step": 3930
    },
    {
      "entropy": 5.498291397094727,
      "epoch": 0.3780019212295869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3435,
      "mean_token_accuracy": 0.19032660573720933,
      "num_tokens": 9007594.0,
      "step": 3935
    },
    {
      "entropy": 5.414751243591309,
      "epoch": 0.37848222862632086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.33,
      "mean_token_accuracy": 0.19158338755369186,
      "num_tokens": 9018097.0,
      "step": 3940
    },
    {
      "entropy": 5.559832191467285,
      "epoch": 0.37896253602305474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.4603,
      "mean_token_accuracy": 0.1812780097126961,
      "num_tokens": 9029554.0,
      "step": 3945
    },
    {
      "entropy": 5.55887622833252,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3256,
      "mean_token_accuracy": 0.19056572914123535,
      "num_tokens": 9042518.0,
      "step": 3950
    },
    {
      "entropy": 5.506370973587036,
      "epoch": 0.37992315081652256,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3644,
      "mean_token_accuracy": 0.19444247484207153,
      "num_tokens": 9054524.0,
      "step": 3955
    },
    {
      "entropy": 5.506084156036377,
      "epoch": 0.3804034582132565,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18973129987716675,
      "num_tokens": 9065375.0,
      "step": 3960
    },
    {
      "entropy": 5.506417083740234,
      "epoch": 0.3808837656099904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.2982,
      "mean_token_accuracy": 0.19565317034721375,
      "num_tokens": 9076338.0,
      "step": 3965
    },
    {
      "entropy": 5.502223873138428,
      "epoch": 0.3813640730067243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.3421,
      "mean_token_accuracy": 0.19122555553913118,
      "num_tokens": 9088407.0,
      "step": 3970
    },
    {
      "entropy": 5.494334125518799,
      "epoch": 0.3818443804034582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19051105678081512,
      "num_tokens": 9099833.0,
      "step": 3975
    },
    {
      "entropy": 5.3872777938842775,
      "epoch": 0.38232468780019213,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.2685,
      "mean_token_accuracy": 0.19974401593208313,
      "num_tokens": 9111126.0,
      "step": 3980
    },
    {
      "entropy": 5.513306760787964,
      "epoch": 0.382804995196926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.3689,
      "mean_token_accuracy": 0.1958555817604065,
      "num_tokens": 9121979.0,
      "step": 3985
    },
    {
      "entropy": 5.5826314926147464,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.3873,
      "mean_token_accuracy": 0.1911478102207184,
      "num_tokens": 9132929.0,
      "step": 3990
    },
    {
      "entropy": 5.551815366744995,
      "epoch": 0.38376560999039383,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3495,
      "mean_token_accuracy": 0.19590957462787628,
      "num_tokens": 9143903.0,
      "step": 3995
    },
    {
      "entropy": 5.567360591888428,
      "epoch": 0.38424591738712777,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.4273,
      "mean_token_accuracy": 0.19291835129261017,
      "num_tokens": 9154416.0,
      "step": 4000
    },
    {
      "entropy": 5.54535870552063,
      "epoch": 0.38472622478386165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.409,
      "mean_token_accuracy": 0.19129804521799088,
      "num_tokens": 9165613.0,
      "step": 4005
    },
    {
      "entropy": 5.470605134963989,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.3711,
      "mean_token_accuracy": 0.1947087824344635,
      "num_tokens": 9177884.0,
      "step": 4010
    },
    {
      "entropy": 5.600809383392334,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.4902,
      "mean_token_accuracy": 0.1800956666469574,
      "num_tokens": 9190041.0,
      "step": 4015
    },
    {
      "entropy": 5.623603248596192,
      "epoch": 0.3861671469740634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.3883,
      "mean_token_accuracy": 0.19113666117191314,
      "num_tokens": 9202390.0,
      "step": 4020
    },
    {
      "entropy": 5.474275207519531,
      "epoch": 0.3866474543707973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.3836,
      "mean_token_accuracy": 0.19006328135728837,
      "num_tokens": 9213437.0,
      "step": 4025
    },
    {
      "entropy": 5.555831718444824,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1827810600399971,
      "num_tokens": 9226310.0,
      "step": 4030
    },
    {
      "entropy": 5.534276008605957,
      "epoch": 0.38760806916426516,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.3874,
      "mean_token_accuracy": 0.1896564930677414,
      "num_tokens": 9236271.0,
      "step": 4035
    },
    {
      "entropy": 5.5598194122314455,
      "epoch": 0.38808837656099904,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.4553,
      "mean_token_accuracy": 0.1913477137684822,
      "num_tokens": 9247755.0,
      "step": 4040
    },
    {
      "entropy": 5.499641227722168,
      "epoch": 0.388568683957733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.3006,
      "mean_token_accuracy": 0.20061542093753815,
      "num_tokens": 9259821.0,
      "step": 4045
    },
    {
      "entropy": 5.5496776580810545,
      "epoch": 0.38904899135446686,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.4043,
      "mean_token_accuracy": 0.18962285965681075,
      "num_tokens": 9270663.0,
      "step": 4050
    },
    {
      "entropy": 5.5318115234375,
      "epoch": 0.3895292987512008,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.4028,
      "mean_token_accuracy": 0.1873584508895874,
      "num_tokens": 9282588.0,
      "step": 4055
    },
    {
      "entropy": 5.505263233184815,
      "epoch": 0.3900096061479347,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.3763,
      "mean_token_accuracy": 0.18254790604114532,
      "num_tokens": 9293368.0,
      "step": 4060
    },
    {
      "entropy": 5.559754848480225,
      "epoch": 0.3904899135446686,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18748684972524643,
      "num_tokens": 9305962.0,
      "step": 4065
    },
    {
      "entropy": 5.516952753067017,
      "epoch": 0.3909702209414025,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.3301,
      "mean_token_accuracy": 0.19609273076057435,
      "num_tokens": 9317955.0,
      "step": 4070
    },
    {
      "entropy": 5.497630071640015,
      "epoch": 0.3914505283381364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19048384428024293,
      "num_tokens": 9329491.0,
      "step": 4075
    },
    {
      "entropy": 5.501287317276001,
      "epoch": 0.3919308357348703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.4717,
      "mean_token_accuracy": 0.19690733551979064,
      "num_tokens": 9341612.0,
      "step": 4080
    },
    {
      "entropy": 5.572618722915649,
      "epoch": 0.39241114313160425,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.411,
      "mean_token_accuracy": 0.18688351958990096,
      "num_tokens": 9352302.0,
      "step": 4085
    },
    {
      "entropy": 5.529020738601685,
      "epoch": 0.3928914505283381,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.3502,
      "mean_token_accuracy": 0.19336572736501695,
      "num_tokens": 9363303.0,
      "step": 4090
    },
    {
      "entropy": 5.446750688552856,
      "epoch": 0.39337175792507206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1869595780968666,
      "num_tokens": 9375878.0,
      "step": 4095
    },
    {
      "entropy": 5.538749361038208,
      "epoch": 0.39385206532180594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1941888928413391,
      "num_tokens": 9388342.0,
      "step": 4100
    },
    {
      "entropy": 5.456574630737305,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.2692,
      "mean_token_accuracy": 0.19488134682178498,
      "num_tokens": 9399047.0,
      "step": 4105
    },
    {
      "entropy": 5.524734115600586,
      "epoch": 0.39481268011527376,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.3773,
      "mean_token_accuracy": 0.1977031409740448,
      "num_tokens": 9410172.0,
      "step": 4110
    },
    {
      "entropy": 5.5667112350463865,
      "epoch": 0.3952929875120077,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.5074,
      "mean_token_accuracy": 0.18664929568767546,
      "num_tokens": 9422287.0,
      "step": 4115
    },
    {
      "entropy": 5.591302490234375,
      "epoch": 0.3957732949087416,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.3713,
      "mean_token_accuracy": 0.19570473879575728,
      "num_tokens": 9433709.0,
      "step": 4120
    },
    {
      "entropy": 5.40889949798584,
      "epoch": 0.3962536023054755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.2625,
      "mean_token_accuracy": 0.19686949849128724,
      "num_tokens": 9445138.0,
      "step": 4125
    },
    {
      "entropy": 5.490507650375366,
      "epoch": 0.3967339097022094,
      "grad_norm": 1.125,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.4006,
      "mean_token_accuracy": 0.18909133076667786,
      "num_tokens": 9455263.0,
      "step": 4130
    },
    {
      "entropy": 5.5528974533081055,
      "epoch": 0.39721421709894333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19381564110517502,
      "num_tokens": 9465491.0,
      "step": 4135
    },
    {
      "entropy": 5.424878263473511,
      "epoch": 0.3976945244956772,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.324,
      "mean_token_accuracy": 0.1861468955874443,
      "num_tokens": 9477601.0,
      "step": 4140
    },
    {
      "entropy": 5.505374145507813,
      "epoch": 0.39817483189241115,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.3778,
      "mean_token_accuracy": 0.1953922465443611,
      "num_tokens": 9489385.0,
      "step": 4145
    },
    {
      "entropy": 5.459268808364868,
      "epoch": 0.39865513928914503,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.248,
      "mean_token_accuracy": 0.2002091407775879,
      "num_tokens": 9501131.0,
      "step": 4150
    },
    {
      "entropy": 5.464940404891967,
      "epoch": 0.39913544668587897,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498961495943777,
      "loss": 5.4137,
      "mean_token_accuracy": 0.19082237184047698,
      "num_tokens": 9513094.0,
      "step": 4155
    },
    {
      "entropy": 5.63197512626648,
      "epoch": 0.39961575408261285,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.522,
      "mean_token_accuracy": 0.18408708721399308,
      "num_tokens": 9524538.0,
      "step": 4160
    },
    {
      "entropy": 5.56934609413147,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.3291,
      "mean_token_accuracy": 0.20425097048282623,
      "num_tokens": 9535284.0,
      "step": 4165
    },
    {
      "entropy": 5.497316694259643,
      "epoch": 0.40057636887608067,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.4125,
      "mean_token_accuracy": 0.1898416116833687,
      "num_tokens": 9546472.0,
      "step": 4170
    },
    {
      "entropy": 5.518751907348633,
      "epoch": 0.4010566762728146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498948293859371,
      "loss": 5.407,
      "mean_token_accuracy": 0.18592777997255325,
      "num_tokens": 9558358.0,
      "step": 4175
    },
    {
      "entropy": 5.598344326019287,
      "epoch": 0.4015369836695485,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.4303,
      "mean_token_accuracy": 0.19122905135154725,
      "num_tokens": 9570653.0,
      "step": 4180
    },
    {
      "entropy": 5.527167081832886,
      "epoch": 0.4020172910662824,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.3722,
      "mean_token_accuracy": 0.19405449777841569,
      "num_tokens": 9582150.0,
      "step": 4185
    },
    {
      "entropy": 5.54949803352356,
      "epoch": 0.4024975984630163,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.3174,
      "mean_token_accuracy": 0.1968682125210762,
      "num_tokens": 9592462.0,
      "step": 4190
    },
    {
      "entropy": 5.459809732437134,
      "epoch": 0.40297790585975024,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19509333968162537,
      "num_tokens": 9604027.0,
      "step": 4195
    },
    {
      "entropy": 5.4236900329589846,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.3037,
      "mean_token_accuracy": 0.19438586235046387,
      "num_tokens": 9616325.0,
      "step": 4200
    },
    {
      "entropy": 5.422830820083618,
      "epoch": 0.40393852065321806,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.2652,
      "mean_token_accuracy": 0.1953098714351654,
      "num_tokens": 9628133.0,
      "step": 4205
    },
    {
      "entropy": 5.626201915740967,
      "epoch": 0.404418828049952,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.4747,
      "mean_token_accuracy": 0.18496284484863282,
      "num_tokens": 9639686.0,
      "step": 4210
    },
    {
      "entropy": 5.500379610061645,
      "epoch": 0.4048991354466859,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.3695,
      "mean_token_accuracy": 0.1925930693745613,
      "num_tokens": 9650222.0,
      "step": 4215
    },
    {
      "entropy": 5.430653524398804,
      "epoch": 0.4053794428434198,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.2787,
      "mean_token_accuracy": 0.20574321448802949,
      "num_tokens": 9661792.0,
      "step": 4220
    },
    {
      "entropy": 5.5395571231842045,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.3946,
      "mean_token_accuracy": 0.18767669349908828,
      "num_tokens": 9673000.0,
      "step": 4225
    },
    {
      "entropy": 5.5615856647491455,
      "epoch": 0.40634005763688763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19707095324993135,
      "num_tokens": 9685253.0,
      "step": 4230
    },
    {
      "entropy": 5.451553249359131,
      "epoch": 0.4068203650336215,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3231,
      "mean_token_accuracy": 0.19093127846717833,
      "num_tokens": 9697245.0,
      "step": 4235
    },
    {
      "entropy": 5.4586076736450195,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.3483,
      "mean_token_accuracy": 0.19377574026584626,
      "num_tokens": 9710095.0,
      "step": 4240
    },
    {
      "entropy": 5.475905895233154,
      "epoch": 0.40778097982708933,
      "grad_norm": 1.0,
      "learning_rate": 0.000498901431470218,
      "loss": 5.2338,
      "mean_token_accuracy": 0.20166248977184295,
      "num_tokens": 9721488.0,
      "step": 4245
    },
    {
      "entropy": 5.341732978820801,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20070856511592866,
      "num_tokens": 9733280.0,
      "step": 4250
    },
    {
      "entropy": 5.480173921585083,
      "epoch": 0.40874159462055715,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.3212,
      "mean_token_accuracy": 0.18952195197343827,
      "num_tokens": 9744514.0,
      "step": 4255
    },
    {
      "entropy": 5.479455947875977,
      "epoch": 0.4092219020172911,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.3254,
      "mean_token_accuracy": 0.19506131410598754,
      "num_tokens": 9754931.0,
      "step": 4260
    },
    {
      "entropy": 5.408810377120972,
      "epoch": 0.40970220941402496,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.2945,
      "mean_token_accuracy": 0.19845788776874543,
      "num_tokens": 9765635.0,
      "step": 4265
    },
    {
      "entropy": 5.524256181716919,
      "epoch": 0.4101825168107589,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1866440549492836,
      "num_tokens": 9777512.0,
      "step": 4270
    },
    {
      "entropy": 5.5750714302062985,
      "epoch": 0.4106628242074928,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.4248,
      "mean_token_accuracy": 0.1778908535838127,
      "num_tokens": 9790128.0,
      "step": 4275
    },
    {
      "entropy": 5.4882549285888675,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498877618185287,
      "loss": 5.3427,
      "mean_token_accuracy": 0.19739877730607985,
      "num_tokens": 9802549.0,
      "step": 4280
    },
    {
      "entropy": 5.437657594680786,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.3003,
      "mean_token_accuracy": 0.19505833089351654,
      "num_tokens": 9813736.0,
      "step": 4285
    },
    {
      "entropy": 5.389750909805298,
      "epoch": 0.41210374639769454,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.286,
      "mean_token_accuracy": 0.20560728162527084,
      "num_tokens": 9823785.0,
      "step": 4290
    },
    {
      "entropy": 5.458524465560913,
      "epoch": 0.4125840537944284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.2571,
      "mean_token_accuracy": 0.1967891976237297,
      "num_tokens": 9834910.0,
      "step": 4295
    },
    {
      "entropy": 5.488246631622315,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.2894,
      "mean_token_accuracy": 0.19697232395410538,
      "num_tokens": 9845593.0,
      "step": 4300
    },
    {
      "entropy": 5.402709579467773,
      "epoch": 0.41354466858789624,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.1841,
      "mean_token_accuracy": 0.21107767671346664,
      "num_tokens": 9856763.0,
      "step": 4305
    },
    {
      "entropy": 5.507501029968262,
      "epoch": 0.4140249759846302,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.4395,
      "mean_token_accuracy": 0.18462781310081483,
      "num_tokens": 9869528.0,
      "step": 4310
    },
    {
      "entropy": 5.452263307571411,
      "epoch": 0.41450528338136405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.3162,
      "mean_token_accuracy": 0.19700846821069717,
      "num_tokens": 9881170.0,
      "step": 4315
    },
    {
      "entropy": 5.508918333053589,
      "epoch": 0.414985590778098,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.4062,
      "mean_token_accuracy": 0.18960152417421341,
      "num_tokens": 9893119.0,
      "step": 4320
    },
    {
      "entropy": 5.370126962661743,
      "epoch": 0.41546589817483187,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.2495,
      "mean_token_accuracy": 0.20651557743549348,
      "num_tokens": 9905339.0,
      "step": 4325
    },
    {
      "entropy": 5.523185920715332,
      "epoch": 0.4159462055715658,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.353,
      "mean_token_accuracy": 0.19940332025289537,
      "num_tokens": 9917500.0,
      "step": 4330
    },
    {
      "entropy": 5.49964451789856,
      "epoch": 0.4164265129682997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1904998764395714,
      "num_tokens": 9929667.0,
      "step": 4335
    },
    {
      "entropy": 5.51170916557312,
      "epoch": 0.4169068203650336,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.3248,
      "mean_token_accuracy": 0.19802108108997346,
      "num_tokens": 9941102.0,
      "step": 4340
    },
    {
      "entropy": 5.4384832859039305,
      "epoch": 0.4173871277617675,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2273,
      "mean_token_accuracy": 0.20404383093118666,
      "num_tokens": 9951986.0,
      "step": 4345
    },
    {
      "entropy": 5.39904203414917,
      "epoch": 0.41786743515850144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2485,
      "mean_token_accuracy": 0.1993297517299652,
      "num_tokens": 9964486.0,
      "step": 4350
    },
    {
      "entropy": 5.455586671829224,
      "epoch": 0.4183477425552353,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.343,
      "mean_token_accuracy": 0.19409458339214325,
      "num_tokens": 9975909.0,
      "step": 4355
    },
    {
      "entropy": 5.496417570114136,
      "epoch": 0.41882804995196926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3475,
      "mean_token_accuracy": 0.19777753502130507,
      "num_tokens": 9986951.0,
      "step": 4360
    },
    {
      "entropy": 5.5789155006408695,
      "epoch": 0.41930835734870314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.4507,
      "mean_token_accuracy": 0.1878629818558693,
      "num_tokens": 9999234.0,
      "step": 4365
    },
    {
      "entropy": 5.499477052688599,
      "epoch": 0.4197886647454371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.3196,
      "mean_token_accuracy": 0.20171295255422592,
      "num_tokens": 10010178.0,
      "step": 4370
    },
    {
      "entropy": 5.425793218612671,
      "epoch": 0.420268972142171,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.264,
      "mean_token_accuracy": 0.19647194296121598,
      "num_tokens": 10021183.0,
      "step": 4375
    },
    {
      "entropy": 5.567107248306274,
      "epoch": 0.4207492795389049,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.5645,
      "mean_token_accuracy": 0.18234878778457642,
      "num_tokens": 10033686.0,
      "step": 4380
    },
    {
      "entropy": 5.449434423446656,
      "epoch": 0.42122958693563883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.2652,
      "mean_token_accuracy": 0.1972568526864052,
      "num_tokens": 10045923.0,
      "step": 4385
    },
    {
      "entropy": 5.306301689147949,
      "epoch": 0.4217098943323727,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498801118470758,
      "loss": 5.1858,
      "mean_token_accuracy": 0.19959155470132828,
      "num_tokens": 10057196.0,
      "step": 4390
    },
    {
      "entropy": 5.449332141876221,
      "epoch": 0.42219020172910665,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498797581450484,
      "loss": 5.3584,
      "mean_token_accuracy": 0.19375586211681367,
      "num_tokens": 10069655.0,
      "step": 4395
    },
    {
      "entropy": 5.511579084396362,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2428,
      "mean_token_accuracy": 0.19924264699220656,
      "num_tokens": 10080876.0,
      "step": 4400
    },
    {
      "entropy": 5.482829475402832,
      "epoch": 0.42315081652257447,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.3524,
      "mean_token_accuracy": 0.18975469768047332,
      "num_tokens": 10091986.0,
      "step": 4405
    },
    {
      "entropy": 5.5305768013000485,
      "epoch": 0.42363112391930835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.4321,
      "mean_token_accuracy": 0.19329569041728972,
      "num_tokens": 10104027.0,
      "step": 4410
    },
    {
      "entropy": 5.5070350646972654,
      "epoch": 0.4241114313160423,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3005,
      "mean_token_accuracy": 0.19417518377304077,
      "num_tokens": 10114665.0,
      "step": 4415
    },
    {
      "entropy": 5.434443378448487,
      "epoch": 0.42459173871277617,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.312,
      "mean_token_accuracy": 0.19974145740270616,
      "num_tokens": 10126032.0,
      "step": 4420
    },
    {
      "entropy": 5.545246362686157,
      "epoch": 0.4250720461095101,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.3715,
      "mean_token_accuracy": 0.19160096794366838,
      "num_tokens": 10137256.0,
      "step": 4425
    },
    {
      "entropy": 5.383159017562866,
      "epoch": 0.425552353506244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498772676827552,
      "loss": 5.2353,
      "mean_token_accuracy": 0.1996947318315506,
      "num_tokens": 10149445.0,
      "step": 4430
    },
    {
      "entropy": 5.5012122631073,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.3886,
      "mean_token_accuracy": 0.19004785567522048,
      "num_tokens": 10161607.0,
      "step": 4435
    },
    {
      "entropy": 5.553760147094726,
      "epoch": 0.4265129682997118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.3855,
      "mean_token_accuracy": 0.1945369154214859,
      "num_tokens": 10173184.0,
      "step": 4440
    },
    {
      "entropy": 5.452174377441406,
      "epoch": 0.42699327569644574,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.2738,
      "mean_token_accuracy": 0.2024209216237068,
      "num_tokens": 10184617.0,
      "step": 4445
    },
    {
      "entropy": 5.451268672943115,
      "epoch": 0.4274735830931796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1954964816570282,
      "num_tokens": 10195608.0,
      "step": 4450
    },
    {
      "entropy": 5.481286334991455,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3806,
      "mean_token_accuracy": 0.19195548743009566,
      "num_tokens": 10206763.0,
      "step": 4455
    },
    {
      "entropy": 5.507647132873535,
      "epoch": 0.42843419788664744,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.3352,
      "mean_token_accuracy": 0.19549722224473953,
      "num_tokens": 10218714.0,
      "step": 4460
    },
    {
      "entropy": 5.460297584533691,
      "epoch": 0.4289145052833814,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.3499,
      "mean_token_accuracy": 0.19259987473487855,
      "num_tokens": 10229902.0,
      "step": 4465
    },
    {
      "entropy": 5.434809589385987,
      "epoch": 0.42939481268011526,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.2242,
      "mean_token_accuracy": 0.19963788241147995,
      "num_tokens": 10241578.0,
      "step": 4470
    },
    {
      "entropy": 5.37636399269104,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.1823,
      "mean_token_accuracy": 0.20253763049840928,
      "num_tokens": 10252109.0,
      "step": 4475
    },
    {
      "entropy": 5.490766906738282,
      "epoch": 0.4303554274735831,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.3707,
      "mean_token_accuracy": 0.19313031136989595,
      "num_tokens": 10263357.0,
      "step": 4480
    },
    {
      "entropy": 5.491276741027832,
      "epoch": 0.430835734870317,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.3569,
      "mean_token_accuracy": 0.1924482449889183,
      "num_tokens": 10275536.0,
      "step": 4485
    },
    {
      "entropy": 5.56570987701416,
      "epoch": 0.4313160422670509,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.3442,
      "mean_token_accuracy": 0.19930241852998734,
      "num_tokens": 10286901.0,
      "step": 4490
    },
    {
      "entropy": 5.362359619140625,
      "epoch": 0.43179634966378483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.2413,
      "mean_token_accuracy": 0.19796304106712342,
      "num_tokens": 10298961.0,
      "step": 4495
    },
    {
      "entropy": 5.4794525623321535,
      "epoch": 0.4322766570605187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.3873,
      "mean_token_accuracy": 0.19520987272262574,
      "num_tokens": 10310911.0,
      "step": 4500
    },
    {
      "entropy": 5.569372749328613,
      "epoch": 0.43275696445725265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.372,
      "mean_token_accuracy": 0.19102310240268708,
      "num_tokens": 10322267.0,
      "step": 4505
    },
    {
      "entropy": 5.460135221481323,
      "epoch": 0.4332372718539865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.3853,
      "mean_token_accuracy": 0.19574257284402846,
      "num_tokens": 10335111.0,
      "step": 4510
    },
    {
      "entropy": 5.537425756454468,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.328,
      "mean_token_accuracy": 0.19893868863582612,
      "num_tokens": 10345672.0,
      "step": 4515
    },
    {
      "entropy": 5.535245752334594,
      "epoch": 0.43419788664745435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049870746748679,
      "loss": 5.3078,
      "mean_token_accuracy": 0.19822679609060287,
      "num_tokens": 10357369.0,
      "step": 4520
    },
    {
      "entropy": 5.475355434417724,
      "epoch": 0.4346781940441883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.3352,
      "mean_token_accuracy": 0.19392889589071274,
      "num_tokens": 10368842.0,
      "step": 4525
    },
    {
      "entropy": 5.417094087600708,
      "epoch": 0.43515850144092216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.2853,
      "mean_token_accuracy": 0.19415317475795746,
      "num_tokens": 10380096.0,
      "step": 4530
    },
    {
      "entropy": 5.3792015552520756,
      "epoch": 0.4356388088376561,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.3172,
      "mean_token_accuracy": 0.19834728091955184,
      "num_tokens": 10391996.0,
      "step": 4535
    },
    {
      "entropy": 5.506541299819946,
      "epoch": 0.43611911623439004,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.2618,
      "mean_token_accuracy": 0.1930739849805832,
      "num_tokens": 10403607.0,
      "step": 4540
    },
    {
      "entropy": 5.444096088409424,
      "epoch": 0.4365994236311239,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1976367026567459,
      "num_tokens": 10413580.0,
      "step": 4545
    },
    {
      "entropy": 5.40678129196167,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3125,
      "mean_token_accuracy": 0.1981777012348175,
      "num_tokens": 10426456.0,
      "step": 4550
    },
    {
      "entropy": 5.5582586288452145,
      "epoch": 0.43756003842459174,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.343,
      "mean_token_accuracy": 0.18815538436174392,
      "num_tokens": 10438869.0,
      "step": 4555
    },
    {
      "entropy": 5.448095989227295,
      "epoch": 0.43804034582132567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.3537,
      "mean_token_accuracy": 0.18729300647974015,
      "num_tokens": 10450672.0,
      "step": 4560
    },
    {
      "entropy": 5.459218215942383,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.261,
      "mean_token_accuracy": 0.20304668992757796,
      "num_tokens": 10461890.0,
      "step": 4565
    },
    {
      "entropy": 5.565139579772949,
      "epoch": 0.4390009606147935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.3919,
      "mean_token_accuracy": 0.18603334277868272,
      "num_tokens": 10473866.0,
      "step": 4570
    },
    {
      "entropy": 5.509669589996338,
      "epoch": 0.43948126801152737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.2954,
      "mean_token_accuracy": 0.1933439001441002,
      "num_tokens": 10484889.0,
      "step": 4575
    },
    {
      "entropy": 5.42793641090393,
      "epoch": 0.4399615754082613,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.3419,
      "mean_token_accuracy": 0.19692840725183486,
      "num_tokens": 10496108.0,
      "step": 4580
    },
    {
      "entropy": 5.463746356964111,
      "epoch": 0.4404418828049952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19601958990097046,
      "num_tokens": 10505472.0,
      "step": 4585
    },
    {
      "entropy": 5.345721244812012,
      "epoch": 0.4409221902017291,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2075,
      "mean_token_accuracy": 0.20081722289323806,
      "num_tokens": 10516954.0,
      "step": 4590
    },
    {
      "entropy": 5.5342426776885985,
      "epoch": 0.441402497598463,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1927899092435837,
      "num_tokens": 10528166.0,
      "step": 4595
    },
    {
      "entropy": 5.422142553329468,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19853255301713943,
      "num_tokens": 10538112.0,
      "step": 4600
    },
    {
      "entropy": 5.489729166030884,
      "epoch": 0.4423631123919308,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.3827,
      "mean_token_accuracy": 0.19151799231767655,
      "num_tokens": 10549888.0,
      "step": 4605
    },
    {
      "entropy": 5.626604223251343,
      "epoch": 0.44284341978866476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.4524,
      "mean_token_accuracy": 0.1831491395831108,
      "num_tokens": 10561710.0,
      "step": 4610
    },
    {
      "entropy": 5.424169254302979,
      "epoch": 0.44332372718539864,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.3121,
      "mean_token_accuracy": 0.1889724239706993,
      "num_tokens": 10574564.0,
      "step": 4615
    },
    {
      "entropy": 5.4898937225341795,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1978645905852318,
      "num_tokens": 10586639.0,
      "step": 4620
    },
    {
      "entropy": 5.4653246879577635,
      "epoch": 0.44428434197886646,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19615231901407243,
      "num_tokens": 10598730.0,
      "step": 4625
    },
    {
      "entropy": 5.420555067062378,
      "epoch": 0.4447646493756004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.2016,
      "mean_token_accuracy": 0.21116500198841096,
      "num_tokens": 10610352.0,
      "step": 4630
    },
    {
      "entropy": 5.394788074493408,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.3007,
      "mean_token_accuracy": 0.19723331481218337,
      "num_tokens": 10621951.0,
      "step": 4635
    },
    {
      "entropy": 5.444929742813111,
      "epoch": 0.4457252641690682,
      "grad_norm": 1.078125,
      "learning_rate": 0.000498617904100496,
      "loss": 5.2438,
      "mean_token_accuracy": 0.20000688582658768,
      "num_tokens": 10633207.0,
      "step": 4640
    },
    {
      "entropy": 5.516901731491089,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.3495,
      "mean_token_accuracy": 0.18970742523670198,
      "num_tokens": 10645853.0,
      "step": 4645
    },
    {
      "entropy": 5.386112785339355,
      "epoch": 0.44668587896253603,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.2482,
      "mean_token_accuracy": 0.20123800486326218,
      "num_tokens": 10656892.0,
      "step": 4650
    },
    {
      "entropy": 5.392650985717774,
      "epoch": 0.4471661863592699,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.2359,
      "mean_token_accuracy": 0.19400574266910553,
      "num_tokens": 10670110.0,
      "step": 4655
    },
    {
      "entropy": 5.4809633731842045,
      "epoch": 0.44764649375600385,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.366,
      "mean_token_accuracy": 0.196006740629673,
      "num_tokens": 10681255.0,
      "step": 4660
    },
    {
      "entropy": 5.473309946060181,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.4205,
      "mean_token_accuracy": 0.19332450479269028,
      "num_tokens": 10692631.0,
      "step": 4665
    },
    {
      "entropy": 5.477552318572998,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.2573,
      "mean_token_accuracy": 0.20014705508947372,
      "num_tokens": 10704880.0,
      "step": 4670
    },
    {
      "entropy": 5.45638484954834,
      "epoch": 0.44908741594620555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.2484,
      "mean_token_accuracy": 0.20326420962810515,
      "num_tokens": 10715561.0,
      "step": 4675
    },
    {
      "entropy": 5.385212564468384,
      "epoch": 0.4495677233429395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.2633,
      "mean_token_accuracy": 0.19535834789276124,
      "num_tokens": 10727232.0,
      "step": 4680
    },
    {
      "entropy": 5.416689777374268,
      "epoch": 0.45004803073967337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.2857,
      "mean_token_accuracy": 0.2015791967511177,
      "num_tokens": 10739404.0,
      "step": 4685
    },
    {
      "entropy": 5.449459791183472,
      "epoch": 0.4505283381364073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.3113,
      "mean_token_accuracy": 0.19386572688817977,
      "num_tokens": 10751948.0,
      "step": 4690
    },
    {
      "entropy": 5.434033584594727,
      "epoch": 0.4510086455331412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.274,
      "mean_token_accuracy": 0.19730297327041627,
      "num_tokens": 10764611.0,
      "step": 4695
    },
    {
      "entropy": 5.465791463851929,
      "epoch": 0.4514889529298751,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.3065,
      "mean_token_accuracy": 0.19460007399320603,
      "num_tokens": 10775954.0,
      "step": 4700
    },
    {
      "entropy": 5.45864953994751,
      "epoch": 0.45196926032660906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.319,
      "mean_token_accuracy": 0.18940264582633973,
      "num_tokens": 10788723.0,
      "step": 4705
    },
    {
      "entropy": 5.474675464630127,
      "epoch": 0.45244956772334294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.3058,
      "mean_token_accuracy": 0.19875997453927993,
      "num_tokens": 10799879.0,
      "step": 4710
    },
    {
      "entropy": 5.400967597961426,
      "epoch": 0.4529298751200769,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.2621,
      "mean_token_accuracy": 0.1981558918952942,
      "num_tokens": 10811838.0,
      "step": 4715
    },
    {
      "entropy": 5.49187388420105,
      "epoch": 0.45341018251681076,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.2807,
      "mean_token_accuracy": 0.1971256822347641,
      "num_tokens": 10822910.0,
      "step": 4720
    },
    {
      "entropy": 5.49492335319519,
      "epoch": 0.4538904899135447,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.3446,
      "mean_token_accuracy": 0.19619156718254088,
      "num_tokens": 10835344.0,
      "step": 4725
    },
    {
      "entropy": 5.4535280704498295,
      "epoch": 0.4543707973102786,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.309,
      "mean_token_accuracy": 0.19785993695259094,
      "num_tokens": 10846043.0,
      "step": 4730
    },
    {
      "entropy": 5.456455278396606,
      "epoch": 0.4548511047070125,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498544878867698,
      "loss": 5.2462,
      "mean_token_accuracy": 0.20247492641210557,
      "num_tokens": 10857783.0,
      "step": 4735
    },
    {
      "entropy": 5.465985870361328,
      "epoch": 0.4553314121037464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.3029,
      "mean_token_accuracy": 0.1979512870311737,
      "num_tokens": 10870527.0,
      "step": 4740
    },
    {
      "entropy": 5.441158056259155,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.2907,
      "mean_token_accuracy": 0.2029913693666458,
      "num_tokens": 10882285.0,
      "step": 4745
    },
    {
      "entropy": 5.489823675155639,
      "epoch": 0.4562920268972142,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19783458560705186,
      "num_tokens": 10895319.0,
      "step": 4750
    },
    {
      "entropy": 5.472352361679077,
      "epoch": 0.45677233429394815,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.2957,
      "mean_token_accuracy": 0.19923655539751053,
      "num_tokens": 10906253.0,
      "step": 4755
    },
    {
      "entropy": 5.335994291305542,
      "epoch": 0.457252641690682,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.1855,
      "mean_token_accuracy": 0.2086847722530365,
      "num_tokens": 10918197.0,
      "step": 4760
    },
    {
      "entropy": 5.443401622772217,
      "epoch": 0.45773294908741596,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.3816,
      "mean_token_accuracy": 0.18888968527317046,
      "num_tokens": 10930771.0,
      "step": 4765
    },
    {
      "entropy": 5.4748773097991945,
      "epoch": 0.45821325648414984,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.2834,
      "mean_token_accuracy": 0.1973588064312935,
      "num_tokens": 10942759.0,
      "step": 4770
    },
    {
      "entropy": 5.6522383213043215,
      "epoch": 0.4586935638808838,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.4848,
      "mean_token_accuracy": 0.18645619004964828,
      "num_tokens": 10953557.0,
      "step": 4775
    },
    {
      "entropy": 5.401769304275513,
      "epoch": 0.45917387127761766,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.2411,
      "mean_token_accuracy": 0.20662152916193008,
      "num_tokens": 10963142.0,
      "step": 4780
    },
    {
      "entropy": 5.379817962646484,
      "epoch": 0.4596541786743516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.2184,
      "mean_token_accuracy": 0.20357681661844254,
      "num_tokens": 10974321.0,
      "step": 4785
    },
    {
      "entropy": 5.509200239181519,
      "epoch": 0.4601344860710855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.4133,
      "mean_token_accuracy": 0.19487691521644593,
      "num_tokens": 10985524.0,
      "step": 4790
    },
    {
      "entropy": 5.494694328308105,
      "epoch": 0.4606147934678194,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.3063,
      "mean_token_accuracy": 0.19958257526159287,
      "num_tokens": 10997606.0,
      "step": 4795
    },
    {
      "entropy": 5.412304258346557,
      "epoch": 0.4610951008645533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.3013,
      "mean_token_accuracy": 0.1943564847111702,
      "num_tokens": 11010692.0,
      "step": 4800
    },
    {
      "entropy": 5.417695474624634,
      "epoch": 0.46157540826128723,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.2374,
      "mean_token_accuracy": 0.20364998728036882,
      "num_tokens": 11022517.0,
      "step": 4805
    },
    {
      "entropy": 5.4359543800354,
      "epoch": 0.4620557156580211,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.248,
      "mean_token_accuracy": 0.1995931401848793,
      "num_tokens": 11033141.0,
      "step": 4810
    },
    {
      "entropy": 5.396144914627075,
      "epoch": 0.46253602305475505,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.2408,
      "mean_token_accuracy": 0.19817344099283218,
      "num_tokens": 11044619.0,
      "step": 4815
    },
    {
      "entropy": 5.519974994659424,
      "epoch": 0.46301633045148893,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.2873,
      "mean_token_accuracy": 0.20086241513490677,
      "num_tokens": 11055934.0,
      "step": 4820
    },
    {
      "entropy": 5.440757846832275,
      "epoch": 0.46349663784822287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.1751,
      "mean_token_accuracy": 0.20307007431983948,
      "num_tokens": 11066203.0,
      "step": 4825
    },
    {
      "entropy": 5.272681379318238,
      "epoch": 0.46397694524495675,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.13,
      "mean_token_accuracy": 0.2047649070620537,
      "num_tokens": 11079664.0,
      "step": 4830
    },
    {
      "entropy": 5.312569189071655,
      "epoch": 0.4644572526416907,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498465986095997,
      "loss": 5.1823,
      "mean_token_accuracy": 0.20340044498443605,
      "num_tokens": 11091186.0,
      "step": 4835
    },
    {
      "entropy": 5.447326278686523,
      "epoch": 0.46493756003842457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.3092,
      "mean_token_accuracy": 0.19724527150392532,
      "num_tokens": 11102710.0,
      "step": 4840
    },
    {
      "entropy": 5.387120294570923,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000498457982665893,
      "loss": 5.2046,
      "mean_token_accuracy": 0.2028914585709572,
      "num_tokens": 11114746.0,
      "step": 4845
    },
    {
      "entropy": 5.423678159713745,
      "epoch": 0.4658981748318924,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19754694998264313,
      "num_tokens": 11126572.0,
      "step": 4850
    },
    {
      "entropy": 5.399871730804444,
      "epoch": 0.4663784822286263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.1642,
      "mean_token_accuracy": 0.20589766055345535,
      "num_tokens": 11137830.0,
      "step": 4855
    },
    {
      "entropy": 5.381054067611695,
      "epoch": 0.4668587896253602,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498445938610739,
      "loss": 5.1978,
      "mean_token_accuracy": 0.2022456184029579,
      "num_tokens": 11148860.0,
      "step": 4860
    },
    {
      "entropy": 5.438592147827149,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.3228,
      "mean_token_accuracy": 0.19238598942756652,
      "num_tokens": 11161311.0,
      "step": 4865
    },
    {
      "entropy": 5.425858068466186,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.2317,
      "mean_token_accuracy": 0.19774759262800218,
      "num_tokens": 11173124.0,
      "step": 4870
    },
    {
      "entropy": 5.395693683624268,
      "epoch": 0.46829971181556196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3059,
      "mean_token_accuracy": 0.1966658964753151,
      "num_tokens": 11184879.0,
      "step": 4875
    },
    {
      "entropy": 5.4837006568908695,
      "epoch": 0.4687800192122959,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1932589590549469,
      "num_tokens": 11196243.0,
      "step": 4880
    },
    {
      "entropy": 5.432374286651611,
      "epoch": 0.4692603266090298,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498425761431707,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20986756533384324,
      "num_tokens": 11207485.0,
      "step": 4885
    },
    {
      "entropy": 5.388256454467774,
      "epoch": 0.4697406340057637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498421710433677,
      "loss": 5.2099,
      "mean_token_accuracy": 0.19842060953378676,
      "num_tokens": 11219891.0,
      "step": 4890
    },
    {
      "entropy": 5.479468727111817,
      "epoch": 0.4702209414024976,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3116,
      "mean_token_accuracy": 0.19674236327409744,
      "num_tokens": 11231329.0,
      "step": 4895
    },
    {
      "entropy": 5.403811025619507,
      "epoch": 0.47070124879923153,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.1932,
      "mean_token_accuracy": 0.20610021948814392,
      "num_tokens": 11241367.0,
      "step": 4900
    },
    {
      "entropy": 5.443480825424194,
      "epoch": 0.4711815561959654,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19220978170633315,
      "num_tokens": 11254532.0,
      "step": 4905
    },
    {
      "entropy": 5.511846876144409,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.3603,
      "mean_token_accuracy": 0.18875951766967775,
      "num_tokens": 11265223.0,
      "step": 4910
    },
    {
      "entropy": 5.373107099533081,
      "epoch": 0.47214217098943323,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3021,
      "mean_token_accuracy": 0.19637542366981506,
      "num_tokens": 11278228.0,
      "step": 4915
    },
    {
      "entropy": 5.564932346343994,
      "epoch": 0.47262247838616717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498397295517835,
      "loss": 5.3949,
      "mean_token_accuracy": 0.1917308121919632,
      "num_tokens": 11289654.0,
      "step": 4920
    },
    {
      "entropy": 5.421516752243042,
      "epoch": 0.47310278578290105,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.1726,
      "mean_token_accuracy": 0.2053057834506035,
      "num_tokens": 11301911.0,
      "step": 4925
    },
    {
      "entropy": 5.5105547428131105,
      "epoch": 0.473583093179635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.3574,
      "mean_token_accuracy": 0.191430202126503,
      "num_tokens": 11312945.0,
      "step": 4930
    },
    {
      "entropy": 5.598050594329834,
      "epoch": 0.47406340057636887,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.4025,
      "mean_token_accuracy": 0.18814169615507126,
      "num_tokens": 11324126.0,
      "step": 4935
    },
    {
      "entropy": 5.4626157760620115,
      "epoch": 0.4745437079731028,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.3348,
      "mean_token_accuracy": 0.1932709753513336,
      "num_tokens": 11336395.0,
      "step": 4940
    },
    {
      "entropy": 5.45054349899292,
      "epoch": 0.4750240153698367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.2371,
      "mean_token_accuracy": 0.2002416417002678,
      "num_tokens": 11347387.0,
      "step": 4945
    },
    {
      "entropy": 5.3878782272338865,
      "epoch": 0.4755043227665706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.2352,
      "mean_token_accuracy": 0.20208385437726975,
      "num_tokens": 11358467.0,
      "step": 4950
    },
    {
      "entropy": 5.491963720321655,
      "epoch": 0.4759846301633045,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.4037,
      "mean_token_accuracy": 0.1896727964282036,
      "num_tokens": 11370322.0,
      "step": 4955
    },
    {
      "entropy": 5.352085447311401,
      "epoch": 0.47646493756003844,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.2004,
      "mean_token_accuracy": 0.20571735352277756,
      "num_tokens": 11381719.0,
      "step": 4960
    },
    {
      "entropy": 5.39383020401001,
      "epoch": 0.4769452449567723,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19686990678310395,
      "num_tokens": 11393561.0,
      "step": 4965
    },
    {
      "entropy": 5.495372152328491,
      "epoch": 0.47742555235350626,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.3062,
      "mean_token_accuracy": 0.19655826836824417,
      "num_tokens": 11405411.0,
      "step": 4970
    },
    {
      "entropy": 5.526172113418579,
      "epoch": 0.47790585975024014,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498352049904511,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18656282126903534,
      "num_tokens": 11417419.0,
      "step": 4975
    },
    {
      "entropy": 5.416673946380615,
      "epoch": 0.4783861671469741,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19893506020307541,
      "num_tokens": 11428145.0,
      "step": 4980
    },
    {
      "entropy": 5.351204967498779,
      "epoch": 0.47886647454370795,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.1681,
      "mean_token_accuracy": 0.20236289799213408,
      "num_tokens": 11439224.0,
      "step": 4985
    },
    {
      "entropy": 5.425208616256714,
      "epoch": 0.4793467819404419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.272,
      "mean_token_accuracy": 0.19770564287900924,
      "num_tokens": 11451731.0,
      "step": 4990
    },
    {
      "entropy": 5.458701133728027,
      "epoch": 0.47982708933717577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.2414,
      "mean_token_accuracy": 0.20528912395238877,
      "num_tokens": 11463058.0,
      "step": 4995
    },
    {
      "entropy": 5.543280839920044,
      "epoch": 0.4803073967339097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.3528,
      "mean_token_accuracy": 0.19466368705034257,
      "num_tokens": 11472618.0,
      "step": 5000
    },
    {
      "entropy": 5.399996662139893,
      "epoch": 0.4807877041306436,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1996273174881935,
      "num_tokens": 11483114.0,
      "step": 5005
    },
    {
      "entropy": 5.363938045501709,
      "epoch": 0.4812680115273775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20535318106412886,
      "num_tokens": 11494281.0,
      "step": 5010
    },
    {
      "entropy": 5.440830373764038,
      "epoch": 0.4817483189241114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.2907,
      "mean_token_accuracy": 0.19693202525377274,
      "num_tokens": 11505962.0,
      "step": 5015
    },
    {
      "entropy": 5.44342532157898,
      "epoch": 0.48222862632084534,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.3485,
      "mean_token_accuracy": 0.18968168497085572,
      "num_tokens": 11517558.0,
      "step": 5020
    },
    {
      "entropy": 5.45598030090332,
      "epoch": 0.4827089337175792,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19084265679121018,
      "num_tokens": 11529184.0,
      "step": 5025
    },
    {
      "entropy": 5.452455759048462,
      "epoch": 0.48318924111431316,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498306176837196,
      "loss": 5.257,
      "mean_token_accuracy": 0.1939294755458832,
      "num_tokens": 11540727.0,
      "step": 5030
    },
    {
      "entropy": 5.3856573581695555,
      "epoch": 0.48366954851104704,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.1863,
      "mean_token_accuracy": 0.20147235542535782,
      "num_tokens": 11551510.0,
      "step": 5035
    },
    {
      "entropy": 5.441668701171875,
      "epoch": 0.484149855907781,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.3278,
      "mean_token_accuracy": 0.20095289647579193,
      "num_tokens": 11563203.0,
      "step": 5040
    },
    {
      "entropy": 5.4081573486328125,
      "epoch": 0.4846301633045149,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.2104,
      "mean_token_accuracy": 0.20497542768716812,
      "num_tokens": 11576013.0,
      "step": 5045
    },
    {
      "entropy": 5.323009014129639,
      "epoch": 0.4851104707012488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.1553,
      "mean_token_accuracy": 0.21624571532011033,
      "num_tokens": 11587535.0,
      "step": 5050
    },
    {
      "entropy": 5.364218139648438,
      "epoch": 0.48559077809798273,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.2053,
      "mean_token_accuracy": 0.20111326426267623,
      "num_tokens": 11598487.0,
      "step": 5055
    },
    {
      "entropy": 5.324197387695312,
      "epoch": 0.4860710854947166,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.091,
      "mean_token_accuracy": 0.22296716421842575,
      "num_tokens": 11609457.0,
      "step": 5060
    },
    {
      "entropy": 5.3482780933380125,
      "epoch": 0.48655139289145055,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.1526,
      "mean_token_accuracy": 0.2081102043390274,
      "num_tokens": 11620251.0,
      "step": 5065
    },
    {
      "entropy": 5.358762311935425,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20766099393367768,
      "num_tokens": 11630956.0,
      "step": 5070
    },
    {
      "entropy": 5.420529079437256,
      "epoch": 0.48751200768491837,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.2004,
      "mean_token_accuracy": 0.20143322944641112,
      "num_tokens": 11642560.0,
      "step": 5075
    },
    {
      "entropy": 5.312218856811524,
      "epoch": 0.48799231508165225,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.1756,
      "mean_token_accuracy": 0.20298496782779693,
      "num_tokens": 11654050.0,
      "step": 5080
    },
    {
      "entropy": 5.3660177230834964,
      "epoch": 0.4884726224783862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.3127,
      "mean_token_accuracy": 0.19897778779268266,
      "num_tokens": 11664947.0,
      "step": 5085
    },
    {
      "entropy": 5.4209997177124025,
      "epoch": 0.48895292987512007,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.2324,
      "mean_token_accuracy": 0.20139615535736083,
      "num_tokens": 11676927.0,
      "step": 5090
    },
    {
      "entropy": 5.530186462402344,
      "epoch": 0.489433237271854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3406,
      "mean_token_accuracy": 0.19309221357107162,
      "num_tokens": 11688573.0,
      "step": 5095
    },
    {
      "entropy": 5.338684368133545,
      "epoch": 0.4899135446685879,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.2421,
      "mean_token_accuracy": 0.1984547719359398,
      "num_tokens": 11698704.0,
      "step": 5100
    },
    {
      "entropy": 5.307417392730713,
      "epoch": 0.4903938520653218,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20310387909412383,
      "num_tokens": 11709466.0,
      "step": 5105
    },
    {
      "entropy": 5.384412813186645,
      "epoch": 0.4908741594620557,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.2733,
      "mean_token_accuracy": 0.19685922116041182,
      "num_tokens": 11721120.0,
      "step": 5110
    },
    {
      "entropy": 5.509656810760498,
      "epoch": 0.49135446685878964,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.2637,
      "mean_token_accuracy": 0.1910458594560623,
      "num_tokens": 11732120.0,
      "step": 5115
    },
    {
      "entropy": 5.373399448394776,
      "epoch": 0.4918347742555235,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20690827518701554,
      "num_tokens": 11743501.0,
      "step": 5120
    },
    {
      "entropy": 5.366362190246582,
      "epoch": 0.49231508165225746,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.2242,
      "mean_token_accuracy": 0.20181894153356553,
      "num_tokens": 11755051.0,
      "step": 5125
    },
    {
      "entropy": 5.41463942527771,
      "epoch": 0.49279538904899134,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.3751,
      "mean_token_accuracy": 0.1938163235783577,
      "num_tokens": 11767663.0,
      "step": 5130
    },
    {
      "entropy": 5.473185110092163,
      "epoch": 0.4932756964457253,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.2429,
      "mean_token_accuracy": 0.20339857041835785,
      "num_tokens": 11778828.0,
      "step": 5135
    },
    {
      "entropy": 5.507049894332885,
      "epoch": 0.49375600384245916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.4109,
      "mean_token_accuracy": 0.19114813357591628,
      "num_tokens": 11790654.0,
      "step": 5140
    },
    {
      "entropy": 5.4209589004516605,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19969180077314377,
      "num_tokens": 11801489.0,
      "step": 5145
    },
    {
      "entropy": 5.366079807281494,
      "epoch": 0.494716618635927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.2152,
      "mean_token_accuracy": 0.20027052909135817,
      "num_tokens": 11813818.0,
      "step": 5150
    },
    {
      "entropy": 5.347721862792969,
      "epoch": 0.4951969260326609,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20178951025009156,
      "num_tokens": 11824814.0,
      "step": 5155
    },
    {
      "entropy": 5.463880491256714,
      "epoch": 0.4956772334293948,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19569255411624908,
      "num_tokens": 11836252.0,
      "step": 5160
    },
    {
      "entropy": 5.381760025024414,
      "epoch": 0.49615754082612873,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.1709,
      "mean_token_accuracy": 0.20152270197868347,
      "num_tokens": 11847715.0,
      "step": 5165
    },
    {
      "entropy": 5.327801704406738,
      "epoch": 0.4966378482228626,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.1758,
      "mean_token_accuracy": 0.20665301084518434,
      "num_tokens": 11860309.0,
      "step": 5170
    },
    {
      "entropy": 5.362015056610107,
      "epoch": 0.49711815561959655,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.1851,
      "mean_token_accuracy": 0.20086576491594316,
      "num_tokens": 11872569.0,
      "step": 5175
    },
    {
      "entropy": 5.359728670120239,
      "epoch": 0.49759846301633043,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.1747,
      "mean_token_accuracy": 0.20249261260032653,
      "num_tokens": 11884778.0,
      "step": 5180
    },
    {
      "entropy": 5.437938070297241,
      "epoch": 0.49807877041306436,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.2555,
      "mean_token_accuracy": 0.20191961973905564,
      "num_tokens": 11897324.0,
      "step": 5185
    },
    {
      "entropy": 5.447718381881714,
      "epoch": 0.49855907780979825,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.2865,
      "mean_token_accuracy": 0.19850171953439713,
      "num_tokens": 11909145.0,
      "step": 5190
    },
    {
      "entropy": 5.35249171257019,
      "epoch": 0.4990393852065322,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20171126276254653,
      "num_tokens": 11921021.0,
      "step": 5195
    },
    {
      "entropy": 5.519689655303955,
      "epoch": 0.49951969260326606,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.3469,
      "mean_token_accuracy": 0.1942518636584282,
      "num_tokens": 11932565.0,
      "step": 5200
    },
    {
      "entropy": 5.374987077713013,
      "epoch": 0.5,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498156044185496,
      "loss": 5.262,
      "mean_token_accuracy": 0.19995938688516618,
      "num_tokens": 11943225.0,
      "step": 5205
    },
    {
      "entropy": 5.364169502258301,
      "epoch": 0.5004803073967339,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.2485,
      "mean_token_accuracy": 0.2035745918750763,
      "num_tokens": 11954821.0,
      "step": 5210
    },
    {
      "entropy": 5.43224196434021,
      "epoch": 0.5009606147934679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.1909,
      "mean_token_accuracy": 0.20336654484272004,
      "num_tokens": 11966090.0,
      "step": 5215
    },
    {
      "entropy": 5.41751708984375,
      "epoch": 0.5014409221902018,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.2322,
      "mean_token_accuracy": 0.19972468316555023,
      "num_tokens": 11977410.0,
      "step": 5220
    },
    {
      "entropy": 5.4079958438873295,
      "epoch": 0.5019212295869356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19654285460710524,
      "num_tokens": 11989119.0,
      "step": 5225
    },
    {
      "entropy": 5.499818134307861,
      "epoch": 0.5024015369836695,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.3181,
      "mean_token_accuracy": 0.19646679908037185,
      "num_tokens": 12000165.0,
      "step": 5230
    },
    {
      "entropy": 5.40293173789978,
      "epoch": 0.5028818443804035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.2831,
      "mean_token_accuracy": 0.1927013948559761,
      "num_tokens": 12012118.0,
      "step": 5235
    },
    {
      "entropy": 5.442527198791504,
      "epoch": 0.5033621517771374,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.3205,
      "mean_token_accuracy": 0.19446006715297698,
      "num_tokens": 12023432.0,
      "step": 5240
    },
    {
      "entropy": 5.4181389808654785,
      "epoch": 0.5038424591738713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.2141,
      "mean_token_accuracy": 0.20718919187784196,
      "num_tokens": 12034323.0,
      "step": 5245
    },
    {
      "entropy": 5.35970139503479,
      "epoch": 0.5043227665706052,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19909979403018951,
      "num_tokens": 12045532.0,
      "step": 5250
    },
    {
      "entropy": 5.451566362380982,
      "epoch": 0.5048030739673391,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.2431,
      "mean_token_accuracy": 0.19895632565021515,
      "num_tokens": 12057346.0,
      "step": 5255
    },
    {
      "entropy": 5.349082517623901,
      "epoch": 0.505283381364073,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20472613275051116,
      "num_tokens": 12068670.0,
      "step": 5260
    },
    {
      "entropy": 5.4529791355133055,
      "epoch": 0.5057636887608069,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.3272,
      "mean_token_accuracy": 0.19437083154916762,
      "num_tokens": 12079664.0,
      "step": 5265
    },
    {
      "entropy": 5.469378185272217,
      "epoch": 0.5062439961575408,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.2632,
      "mean_token_accuracy": 0.2042248487472534,
      "num_tokens": 12089984.0,
      "step": 5270
    },
    {
      "entropy": 5.319522666931152,
      "epoch": 0.5067243035542748,
      "grad_norm": 1.0,
      "learning_rate": 0.000498094213793741,
      "loss": 5.1899,
      "mean_token_accuracy": 0.20940718650817872,
      "num_tokens": 12101182.0,
      "step": 5275
    },
    {
      "entropy": 5.4405517578125,
      "epoch": 0.5072046109510087,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498089758486952,
      "loss": 5.2895,
      "mean_token_accuracy": 0.19249950349330902,
      "num_tokens": 12112002.0,
      "step": 5280
    },
    {
      "entropy": 5.489940357208252,
      "epoch": 0.5076849183477425,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.356,
      "mean_token_accuracy": 0.1845449000597,
      "num_tokens": 12124194.0,
      "step": 5285
    },
    {
      "entropy": 5.358373641967773,
      "epoch": 0.5081652257444764,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.2607,
      "mean_token_accuracy": 0.20721071660518647,
      "num_tokens": 12133966.0,
      "step": 5290
    },
    {
      "entropy": 5.400069761276245,
      "epoch": 0.5086455331412104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.2166,
      "mean_token_accuracy": 0.2029334396123886,
      "num_tokens": 12145643.0,
      "step": 5295
    },
    {
      "entropy": 5.388980054855347,
      "epoch": 0.5091258405379443,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.2638,
      "mean_token_accuracy": 0.19128015339374543,
      "num_tokens": 12156804.0,
      "step": 5300
    },
    {
      "entropy": 5.410366010665894,
      "epoch": 0.5096061479346782,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.2201,
      "mean_token_accuracy": 0.19946327358484267,
      "num_tokens": 12168700.0,
      "step": 5305
    },
    {
      "entropy": 5.441403770446778,
      "epoch": 0.5100864553314121,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00049806291788073,
      "loss": 5.2344,
      "mean_token_accuracy": 0.19958012998104097,
      "num_tokens": 12181050.0,
      "step": 5310
    },
    {
      "entropy": 5.441706371307373,
      "epoch": 0.510566762728146,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.321,
      "mean_token_accuracy": 0.1931978940963745,
      "num_tokens": 12192001.0,
      "step": 5315
    },
    {
      "entropy": 5.559450435638428,
      "epoch": 0.5110470701248799,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.4211,
      "mean_token_accuracy": 0.19310695230960845,
      "num_tokens": 12202436.0,
      "step": 5320
    },
    {
      "entropy": 5.374323940277099,
      "epoch": 0.5115273775216138,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.2081,
      "mean_token_accuracy": 0.20342611223459245,
      "num_tokens": 12214454.0,
      "step": 5325
    },
    {
      "entropy": 5.385732889175415,
      "epoch": 0.5120076849183477,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.2272,
      "mean_token_accuracy": 0.20538058280944824,
      "num_tokens": 12225924.0,
      "step": 5330
    },
    {
      "entropy": 5.471169757843017,
      "epoch": 0.5124879923150817,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049804040828932,
      "loss": 5.2273,
      "mean_token_accuracy": 0.20452706813812255,
      "num_tokens": 12236456.0,
      "step": 5335
    },
    {
      "entropy": 5.417616176605224,
      "epoch": 0.5129682997118156,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.2598,
      "mean_token_accuracy": 0.20173069089651108,
      "num_tokens": 12247719.0,
      "step": 5340
    },
    {
      "entropy": 5.375013160705566,
      "epoch": 0.5134486071085494,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.3013,
      "mean_token_accuracy": 0.1980268254876137,
      "num_tokens": 12259141.0,
      "step": 5345
    },
    {
      "entropy": 5.52657642364502,
      "epoch": 0.5139289145052833,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.342,
      "mean_token_accuracy": 0.19116934835910798,
      "num_tokens": 12269748.0,
      "step": 5350
    },
    {
      "entropy": 5.426257944107055,
      "epoch": 0.5144092219020173,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.297,
      "mean_token_accuracy": 0.2044803574681282,
      "num_tokens": 12281456.0,
      "step": 5355
    },
    {
      "entropy": 5.440986442565918,
      "epoch": 0.5148895292987512,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.2796,
      "mean_token_accuracy": 0.1947730675339699,
      "num_tokens": 12293763.0,
      "step": 5360
    },
    {
      "entropy": 5.409850406646728,
      "epoch": 0.5153698366954851,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.2026,
      "mean_token_accuracy": 0.20839707404375077,
      "num_tokens": 12305398.0,
      "step": 5365
    },
    {
      "entropy": 5.35020489692688,
      "epoch": 0.515850144092219,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.1854,
      "mean_token_accuracy": 0.199608738720417,
      "num_tokens": 12316582.0,
      "step": 5370
    },
    {
      "entropy": 5.461497783660889,
      "epoch": 0.516330451488953,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.3115,
      "mean_token_accuracy": 0.1998691514134407,
      "num_tokens": 12328463.0,
      "step": 5375
    },
    {
      "entropy": 5.471179628372193,
      "epoch": 0.5168107588856868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.291,
      "mean_token_accuracy": 0.2032223239541054,
      "num_tokens": 12341040.0,
      "step": 5380
    },
    {
      "entropy": 5.324481773376465,
      "epoch": 0.5172910662824207,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.1905,
      "mean_token_accuracy": 0.2048783466219902,
      "num_tokens": 12353024.0,
      "step": 5385
    },
    {
      "entropy": 5.357087421417236,
      "epoch": 0.5177713736791547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.206,
      "mean_token_accuracy": 0.20590624809265137,
      "num_tokens": 12366100.0,
      "step": 5390
    },
    {
      "entropy": 5.353571033477783,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.1981,
      "mean_token_accuracy": 0.20849017649888993,
      "num_tokens": 12377829.0,
      "step": 5395
    },
    {
      "entropy": 5.511527395248413,
      "epoch": 0.5187319884726225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.315,
      "mean_token_accuracy": 0.19967973232269287,
      "num_tokens": 12390830.0,
      "step": 5400
    },
    {
      "entropy": 5.343885278701782,
      "epoch": 0.5192122958693564,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.1545,
      "mean_token_accuracy": 0.20680712759494782,
      "num_tokens": 12401642.0,
      "step": 5405
    },
    {
      "entropy": 5.35053563117981,
      "epoch": 0.5196926032660903,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2792,
      "mean_token_accuracy": 0.20148641914129256,
      "num_tokens": 12411653.0,
      "step": 5410
    },
    {
      "entropy": 5.414511108398438,
      "epoch": 0.5201729106628242,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497967507701269,
      "loss": 5.1717,
      "mean_token_accuracy": 0.2088325709104538,
      "num_tokens": 12422891.0,
      "step": 5415
    },
    {
      "entropy": 5.461643075942993,
      "epoch": 0.5206532180595581,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.2585,
      "mean_token_accuracy": 0.20017822086811066,
      "num_tokens": 12434190.0,
      "step": 5420
    },
    {
      "entropy": 5.425161838531494,
      "epoch": 0.521133525456292,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.3223,
      "mean_token_accuracy": 0.19771393835544587,
      "num_tokens": 12445796.0,
      "step": 5425
    },
    {
      "entropy": 5.515705680847168,
      "epoch": 0.521613832853026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.3871,
      "mean_token_accuracy": 0.19371298253536223,
      "num_tokens": 12456212.0,
      "step": 5430
    },
    {
      "entropy": 5.369207715988159,
      "epoch": 0.5220941402497599,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497949075471522,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20107672214508057,
      "num_tokens": 12467871.0,
      "step": 5435
    },
    {
      "entropy": 5.394469690322876,
      "epoch": 0.5225744476464937,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1949324682354927,
      "num_tokens": 12478626.0,
      "step": 5440
    },
    {
      "entropy": 5.365805816650391,
      "epoch": 0.5230547550432276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497939828298927,
      "loss": 5.2886,
      "mean_token_accuracy": 0.19763265550136566,
      "num_tokens": 12491487.0,
      "step": 5445
    },
    {
      "entropy": 5.461965608596802,
      "epoch": 0.5235350624399616,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19547561854124068,
      "num_tokens": 12503240.0,
      "step": 5450
    },
    {
      "entropy": 5.455387878417969,
      "epoch": 0.5240153698366955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20037689208984374,
      "num_tokens": 12513860.0,
      "step": 5455
    },
    {
      "entropy": 5.3650209426879885,
      "epoch": 0.5244956772334294,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.2683,
      "mean_token_accuracy": 0.20282870829105376,
      "num_tokens": 12525884.0,
      "step": 5460
    },
    {
      "entropy": 5.4195245742797855,
      "epoch": 0.5249759846301633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.1955,
      "mean_token_accuracy": 0.20026529282331468,
      "num_tokens": 12536709.0,
      "step": 5465
    },
    {
      "entropy": 5.410429239273071,
      "epoch": 0.5254562920268973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.1746,
      "mean_token_accuracy": 0.20506655275821686,
      "num_tokens": 12549727.0,
      "step": 5470
    },
    {
      "entropy": 5.3270786762237545,
      "epoch": 0.5259365994236311,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.2661,
      "mean_token_accuracy": 0.19516700655221939,
      "num_tokens": 12562053.0,
      "step": 5475
    },
    {
      "entropy": 5.345616006851197,
      "epoch": 0.526416906820365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20667382925748826,
      "num_tokens": 12574096.0,
      "step": 5480
    },
    {
      "entropy": 5.460164356231689,
      "epoch": 0.5268972142170989,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.3212,
      "mean_token_accuracy": 0.20060795843601226,
      "num_tokens": 12585416.0,
      "step": 5485
    },
    {
      "entropy": 5.413971662521362,
      "epoch": 0.5273775216138329,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20576133579015732,
      "num_tokens": 12596861.0,
      "step": 5490
    },
    {
      "entropy": 5.333009815216064,
      "epoch": 0.5278578290105668,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.1927,
      "mean_token_accuracy": 0.20716883689165116,
      "num_tokens": 12607761.0,
      "step": 5495
    },
    {
      "entropy": 5.443383312225341,
      "epoch": 0.5283381364073007,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3051,
      "mean_token_accuracy": 0.20227987468242645,
      "num_tokens": 12619889.0,
      "step": 5500
    },
    {
      "entropy": 5.509159994125366,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19618211537599564,
      "num_tokens": 12630604.0,
      "step": 5505
    },
    {
      "entropy": 5.314917039871216,
      "epoch": 0.5292987512007685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20746731758117676,
      "num_tokens": 12641172.0,
      "step": 5510
    },
    {
      "entropy": 5.312827301025391,
      "epoch": 0.5297790585975024,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.0923,
      "mean_token_accuracy": 0.21527302861213685,
      "num_tokens": 12651731.0,
      "step": 5515
    },
    {
      "entropy": 5.386341571807861,
      "epoch": 0.5302593659942363,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.2144,
      "mean_token_accuracy": 0.20662128329277038,
      "num_tokens": 12663050.0,
      "step": 5520
    },
    {
      "entropy": 5.479044103622437,
      "epoch": 0.5307396733909702,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.3325,
      "mean_token_accuracy": 0.19395910054445267,
      "num_tokens": 12674732.0,
      "step": 5525
    },
    {
      "entropy": 5.313097810745239,
      "epoch": 0.5312199807877042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000497860391494935,
      "loss": 5.1422,
      "mean_token_accuracy": 0.2039245158433914,
      "num_tokens": 12685981.0,
      "step": 5530
    },
    {
      "entropy": 5.447102403640747,
      "epoch": 0.531700288184438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.2773,
      "mean_token_accuracy": 0.19787385314702988,
      "num_tokens": 12697139.0,
      "step": 5535
    },
    {
      "entropy": 5.3786839008331295,
      "epoch": 0.5321805955811719,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.2202,
      "mean_token_accuracy": 0.19959738850593567,
      "num_tokens": 12708268.0,
      "step": 5540
    },
    {
      "entropy": 5.375888729095459,
      "epoch": 0.5326609029779059,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.2049,
      "mean_token_accuracy": 0.20851173847913743,
      "num_tokens": 12720715.0,
      "step": 5545
    },
    {
      "entropy": 5.367966890335083,
      "epoch": 0.5331412103746398,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.1717,
      "mean_token_accuracy": 0.20525323897600173,
      "num_tokens": 12732409.0,
      "step": 5550
    },
    {
      "entropy": 5.346772003173828,
      "epoch": 0.5336215177713737,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.3142,
      "mean_token_accuracy": 0.1936624139547348,
      "num_tokens": 12745174.0,
      "step": 5555
    },
    {
      "entropy": 5.437586259841919,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.2593,
      "mean_token_accuracy": 0.20328891277313232,
      "num_tokens": 12756116.0,
      "step": 5560
    },
    {
      "entropy": 5.431155967712402,
      "epoch": 0.5345821325648416,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.1951,
      "mean_token_accuracy": 0.20230283141136168,
      "num_tokens": 12768183.0,
      "step": 5565
    },
    {
      "entropy": 5.306485891342163,
      "epoch": 0.5350624399615754,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.1953,
      "mean_token_accuracy": 0.19942874312400818,
      "num_tokens": 12778537.0,
      "step": 5570
    },
    {
      "entropy": 5.288288307189942,
      "epoch": 0.5355427473583093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.2401,
      "mean_token_accuracy": 0.2020668715238571,
      "num_tokens": 12789691.0,
      "step": 5575
    },
    {
      "entropy": 5.450948762893677,
      "epoch": 0.5360230547550432,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19842252731323243,
      "num_tokens": 12801438.0,
      "step": 5580
    },
    {
      "entropy": 5.403745031356811,
      "epoch": 0.5365033621517772,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19873054474592208,
      "num_tokens": 12812425.0,
      "step": 5585
    },
    {
      "entropy": 5.376599311828613,
      "epoch": 0.5369836695485111,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.1988,
      "mean_token_accuracy": 0.19666872173547745,
      "num_tokens": 12824735.0,
      "step": 5590
    },
    {
      "entropy": 5.448877477645874,
      "epoch": 0.537463976945245,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.2731,
      "mean_token_accuracy": 0.2031214103102684,
      "num_tokens": 12835454.0,
      "step": 5595
    },
    {
      "entropy": 5.4336474418640135,
      "epoch": 0.5379442843419788,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.3367,
      "mean_token_accuracy": 0.1929567873477936,
      "num_tokens": 12847086.0,
      "step": 5600
    },
    {
      "entropy": 5.420505571365356,
      "epoch": 0.5384245917387128,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19988568872213364,
      "num_tokens": 12857713.0,
      "step": 5605
    },
    {
      "entropy": 5.375241994857788,
      "epoch": 0.5389048991354467,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1993047297000885,
      "num_tokens": 12869967.0,
      "step": 5610
    },
    {
      "entropy": 5.490482473373413,
      "epoch": 0.5393852065321806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.2826,
      "mean_token_accuracy": 0.19814034998416902,
      "num_tokens": 12881230.0,
      "step": 5615
    },
    {
      "entropy": 5.39257779121399,
      "epoch": 0.5398655139289145,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497774652090923,
      "loss": 5.2414,
      "mean_token_accuracy": 0.19702692776918412,
      "num_tokens": 12892376.0,
      "step": 5620
    },
    {
      "entropy": 5.413067626953125,
      "epoch": 0.5403458213256485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20418344736099242,
      "num_tokens": 12903709.0,
      "step": 5625
    },
    {
      "entropy": 5.319581651687622,
      "epoch": 0.5408261287223823,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.0824,
      "mean_token_accuracy": 0.20584484487771987,
      "num_tokens": 12914958.0,
      "step": 5630
    },
    {
      "entropy": 5.376936054229736,
      "epoch": 0.5413064361191162,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.2056,
      "mean_token_accuracy": 0.20099034607410432,
      "num_tokens": 12927418.0,
      "step": 5635
    },
    {
      "entropy": 5.412064933776856,
      "epoch": 0.5417867435158501,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.2503,
      "mean_token_accuracy": 0.20160660594701768,
      "num_tokens": 12939874.0,
      "step": 5640
    },
    {
      "entropy": 5.416967535018921,
      "epoch": 0.5422670509125841,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.2681,
      "mean_token_accuracy": 0.19486918598413466,
      "num_tokens": 12951113.0,
      "step": 5645
    },
    {
      "entropy": 5.315142440795898,
      "epoch": 0.542747358309318,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.1594,
      "mean_token_accuracy": 0.2023612678050995,
      "num_tokens": 12961940.0,
      "step": 5650
    },
    {
      "entropy": 5.284940385818482,
      "epoch": 0.5432276657060519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.1674,
      "mean_token_accuracy": 0.19892300814390182,
      "num_tokens": 12973938.0,
      "step": 5655
    },
    {
      "entropy": 5.412263488769531,
      "epoch": 0.5437079731027857,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.1788,
      "mean_token_accuracy": 0.20297273993492126,
      "num_tokens": 12985400.0,
      "step": 5660
    },
    {
      "entropy": 5.385592126846314,
      "epoch": 0.5441882804995197,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.2286,
      "mean_token_accuracy": 0.20850644558668135,
      "num_tokens": 12996402.0,
      "step": 5665
    },
    {
      "entropy": 5.293762063980102,
      "epoch": 0.5446685878962536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.1092,
      "mean_token_accuracy": 0.21470348685979843,
      "num_tokens": 13006991.0,
      "step": 5670
    },
    {
      "entropy": 5.397123908996582,
      "epoch": 0.5451488952929875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.1999,
      "mean_token_accuracy": 0.20438773483037947,
      "num_tokens": 13016969.0,
      "step": 5675
    },
    {
      "entropy": 5.392855453491211,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497716561458634,
      "loss": 5.1437,
      "mean_token_accuracy": 0.20777873545885087,
      "num_tokens": 13027759.0,
      "step": 5680
    },
    {
      "entropy": 5.486040258407593,
      "epoch": 0.5461095100864554,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.3619,
      "mean_token_accuracy": 0.1973151370882988,
      "num_tokens": 13039881.0,
      "step": 5685
    },
    {
      "entropy": 5.378542852401734,
      "epoch": 0.5465898174831892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.1963,
      "mean_token_accuracy": 0.20589208602905273,
      "num_tokens": 13051443.0,
      "step": 5690
    },
    {
      "entropy": 5.4433074474334715,
      "epoch": 0.5470701248799231,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1966116562485695,
      "num_tokens": 13063474.0,
      "step": 5695
    },
    {
      "entropy": 5.370102453231811,
      "epoch": 0.547550432276657,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.1416,
      "mean_token_accuracy": 0.2106616035103798,
      "num_tokens": 13074365.0,
      "step": 5700
    },
    {
      "entropy": 5.358544445037841,
      "epoch": 0.548030739673391,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.2212,
      "mean_token_accuracy": 0.20368915945291519,
      "num_tokens": 13087354.0,
      "step": 5705
    },
    {
      "entropy": 5.370021200180053,
      "epoch": 0.5485110470701249,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.1967,
      "mean_token_accuracy": 0.19930920600891114,
      "num_tokens": 13099306.0,
      "step": 5710
    },
    {
      "entropy": 5.408384513854981,
      "epoch": 0.5489913544668588,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.3329,
      "mean_token_accuracy": 0.19190380424261094,
      "num_tokens": 13111259.0,
      "step": 5715
    },
    {
      "entropy": 5.45887770652771,
      "epoch": 0.5494716618635928,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.2471,
      "mean_token_accuracy": 0.19341034889221193,
      "num_tokens": 13123589.0,
      "step": 5720
    },
    {
      "entropy": 5.377495574951172,
      "epoch": 0.5499519692603266,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.2444,
      "mean_token_accuracy": 0.19959345310926438,
      "num_tokens": 13136485.0,
      "step": 5725
    },
    {
      "entropy": 5.409855890274048,
      "epoch": 0.5504322766570605,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19824812114238738,
      "num_tokens": 13148067.0,
      "step": 5730
    },
    {
      "entropy": 5.3532360076904295,
      "epoch": 0.5509125840537944,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.1943,
      "mean_token_accuracy": 0.19984031617641448,
      "num_tokens": 13159813.0,
      "step": 5735
    },
    {
      "entropy": 5.414771270751953,
      "epoch": 0.5513928914505284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.2672,
      "mean_token_accuracy": 0.19723765850067138,
      "num_tokens": 13170828.0,
      "step": 5740
    },
    {
      "entropy": 5.348534822463989,
      "epoch": 0.5518731988472623,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.0988,
      "mean_token_accuracy": 0.21700155436992646,
      "num_tokens": 13181883.0,
      "step": 5745
    },
    {
      "entropy": 5.332348966598511,
      "epoch": 0.5523535062439962,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20436936616897583,
      "num_tokens": 13193530.0,
      "step": 5750
    },
    {
      "entropy": 5.461158657073975,
      "epoch": 0.55283381364073,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.3936,
      "mean_token_accuracy": 0.1911667287349701,
      "num_tokens": 13205822.0,
      "step": 5755
    },
    {
      "entropy": 5.377006673812867,
      "epoch": 0.553314121037464,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.1143,
      "mean_token_accuracy": 0.210676147043705,
      "num_tokens": 13216698.0,
      "step": 5760
    },
    {
      "entropy": 5.346554327011108,
      "epoch": 0.5537944284341979,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.2225,
      "mean_token_accuracy": 0.2032114490866661,
      "num_tokens": 13228389.0,
      "step": 5765
    },
    {
      "entropy": 5.437006664276123,
      "epoch": 0.5542747358309318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.2655,
      "mean_token_accuracy": 0.19674807637929917,
      "num_tokens": 13239628.0,
      "step": 5770
    },
    {
      "entropy": 5.438928031921387,
      "epoch": 0.5547550432276657,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.2726,
      "mean_token_accuracy": 0.20090280920267106,
      "num_tokens": 13251149.0,
      "step": 5775
    },
    {
      "entropy": 5.4733867168426515,
      "epoch": 0.5552353506243997,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497618089174132,
      "loss": 5.3166,
      "mean_token_accuracy": 0.19709511697292328,
      "num_tokens": 13264846.0,
      "step": 5780
    },
    {
      "entropy": 5.324634456634522,
      "epoch": 0.5557156580211335,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.1012,
      "mean_token_accuracy": 0.2113727256655693,
      "num_tokens": 13275409.0,
      "step": 5785
    },
    {
      "entropy": 5.231151103973389,
      "epoch": 0.5561959654178674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.0786,
      "mean_token_accuracy": 0.20907817929983138,
      "num_tokens": 13287173.0,
      "step": 5790
    },
    {
      "entropy": 5.402900981903076,
      "epoch": 0.5566762728146013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000497603139994675,
      "loss": 5.1658,
      "mean_token_accuracy": 0.2003258064389229,
      "num_tokens": 13298225.0,
      "step": 5795
    },
    {
      "entropy": 5.381881332397461,
      "epoch": 0.5571565802113353,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20382763296365738,
      "num_tokens": 13309685.0,
      "step": 5800
    },
    {
      "entropy": 5.313303709030151,
      "epoch": 0.5576368876080692,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.1777,
      "mean_token_accuracy": 0.20920694172382354,
      "num_tokens": 13320837.0,
      "step": 5805
    },
    {
      "entropy": 5.373168277740478,
      "epoch": 0.5581171950048031,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.279,
      "mean_token_accuracy": 0.20333101153373717,
      "num_tokens": 13333512.0,
      "step": 5810
    },
    {
      "entropy": 5.478464984893799,
      "epoch": 0.5585975024015369,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19963099658489228,
      "num_tokens": 13345189.0,
      "step": 5815
    },
    {
      "entropy": 5.362034654617309,
      "epoch": 0.5590778097982709,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.1646,
      "mean_token_accuracy": 0.2086104243993759,
      "num_tokens": 13356123.0,
      "step": 5820
    },
    {
      "entropy": 5.3880220413208,
      "epoch": 0.5595581171950048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.2036,
      "mean_token_accuracy": 0.20239470899105072,
      "num_tokens": 13366902.0,
      "step": 5825
    },
    {
      "entropy": 5.459943771362305,
      "epoch": 0.5600384245917387,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.2605,
      "mean_token_accuracy": 0.200407475233078,
      "num_tokens": 13377567.0,
      "step": 5830
    },
    {
      "entropy": 5.400122451782226,
      "epoch": 0.5605187319884726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.2056,
      "mean_token_accuracy": 0.2129899948835373,
      "num_tokens": 13387297.0,
      "step": 5835
    },
    {
      "entropy": 5.408231210708618,
      "epoch": 0.5609990393852066,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.2405,
      "mean_token_accuracy": 0.20631884038448334,
      "num_tokens": 13397723.0,
      "step": 5840
    },
    {
      "entropy": 5.416802930831909,
      "epoch": 0.5614793467819404,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.2988,
      "mean_token_accuracy": 0.20144724398851394,
      "num_tokens": 13409875.0,
      "step": 5845
    },
    {
      "entropy": 5.36104941368103,
      "epoch": 0.5619596541786743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.1902,
      "mean_token_accuracy": 0.20576498061418533,
      "num_tokens": 13421338.0,
      "step": 5850
    },
    {
      "entropy": 5.426533842086792,
      "epoch": 0.5624399615754082,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.2212,
      "mean_token_accuracy": 0.20507221519947053,
      "num_tokens": 13431373.0,
      "step": 5855
    },
    {
      "entropy": 5.410734415054321,
      "epoch": 0.5629202689721422,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.2411,
      "mean_token_accuracy": 0.20476576089859008,
      "num_tokens": 13443158.0,
      "step": 5860
    },
    {
      "entropy": 5.4595184326171875,
      "epoch": 0.5634005763688761,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.3079,
      "mean_token_accuracy": 0.1933220088481903,
      "num_tokens": 13454559.0,
      "step": 5865
    },
    {
      "entropy": 5.514381456375122,
      "epoch": 0.56388088376561,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.3656,
      "mean_token_accuracy": 0.19711938351392747,
      "num_tokens": 13466934.0,
      "step": 5870
    },
    {
      "entropy": 5.40661940574646,
      "epoch": 0.5643611911623438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.1231,
      "mean_token_accuracy": 0.21357636600732804,
      "num_tokens": 13477770.0,
      "step": 5875
    },
    {
      "entropy": 5.451413154602051,
      "epoch": 0.5648414985590778,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1998008131980896,
      "num_tokens": 13490093.0,
      "step": 5880
    },
    {
      "entropy": 5.453090238571167,
      "epoch": 0.5653218059558117,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.182,
      "mean_token_accuracy": 0.20121294111013413,
      "num_tokens": 13500663.0,
      "step": 5885
    },
    {
      "entropy": 5.405351638793945,
      "epoch": 0.5658021133525456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.2896,
      "mean_token_accuracy": 0.2002902314066887,
      "num_tokens": 13512369.0,
      "step": 5890
    },
    {
      "entropy": 5.3891054630279545,
      "epoch": 0.5662824207492796,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20277402997016908,
      "num_tokens": 13523181.0,
      "step": 5895
    },
    {
      "entropy": 5.361138439178466,
      "epoch": 0.5667627281460135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20029759407043457,
      "num_tokens": 13534113.0,
      "step": 5900
    },
    {
      "entropy": 5.327734184265137,
      "epoch": 0.5672430355427474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20367970913648606,
      "num_tokens": 13545037.0,
      "step": 5905
    },
    {
      "entropy": 5.405796194076538,
      "epoch": 0.5677233429394812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.2249,
      "mean_token_accuracy": 0.20131197422742844,
      "num_tokens": 13555541.0,
      "step": 5910
    },
    {
      "entropy": 5.352486181259155,
      "epoch": 0.5682036503362152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.1964,
      "mean_token_accuracy": 0.20470297634601592,
      "num_tokens": 13566951.0,
      "step": 5915
    },
    {
      "entropy": 5.4211790561676025,
      "epoch": 0.5686839577329491,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.2705,
      "mean_token_accuracy": 0.2000730812549591,
      "num_tokens": 13578492.0,
      "step": 5920
    },
    {
      "entropy": 5.4369526386260985,
      "epoch": 0.569164265129683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.1561,
      "mean_token_accuracy": 0.21389784663915634,
      "num_tokens": 13589183.0,
      "step": 5925
    },
    {
      "entropy": 5.404523801803589,
      "epoch": 0.5696445725264169,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.2862,
      "mean_token_accuracy": 0.202628256380558,
      "num_tokens": 13600588.0,
      "step": 5930
    },
    {
      "entropy": 5.343457460403442,
      "epoch": 0.5701248799231509,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.1779,
      "mean_token_accuracy": 0.20688795298337936,
      "num_tokens": 13612107.0,
      "step": 5935
    },
    {
      "entropy": 5.510196113586426,
      "epoch": 0.5706051873198847,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.3502,
      "mean_token_accuracy": 0.1963939294219017,
      "num_tokens": 13623973.0,
      "step": 5940
    },
    {
      "entropy": 5.374628925323487,
      "epoch": 0.5710854947166186,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.146,
      "mean_token_accuracy": 0.20711901485919954,
      "num_tokens": 13634994.0,
      "step": 5945
    },
    {
      "entropy": 5.426691341400146,
      "epoch": 0.5715658021133525,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.3057,
      "mean_token_accuracy": 0.20290742069482803,
      "num_tokens": 13646361.0,
      "step": 5950
    },
    {
      "entropy": 5.406249523162842,
      "epoch": 0.5720461095100865,
      "grad_norm": 1.0,
      "learning_rate": 0.000497440788529968,
      "loss": 5.1866,
      "mean_token_accuracy": 0.20263564288616182,
      "num_tokens": 13656975.0,
      "step": 5955
    },
    {
      "entropy": 5.3955769538879395,
      "epoch": 0.5725264169068204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19841890335083007,
      "num_tokens": 13668434.0,
      "step": 5960
    },
    {
      "entropy": 5.416602754592896,
      "epoch": 0.5730067243035543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.2127,
      "mean_token_accuracy": 0.19981901794672013,
      "num_tokens": 13679266.0,
      "step": 5965
    },
    {
      "entropy": 5.355310297012329,
      "epoch": 0.5734870317002881,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.2388,
      "mean_token_accuracy": 0.1961240813136101,
      "num_tokens": 13690921.0,
      "step": 5970
    },
    {
      "entropy": 5.414062929153443,
      "epoch": 0.5739673390970221,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497420122679239,
      "loss": 5.1833,
      "mean_token_accuracy": 0.20539622157812118,
      "num_tokens": 13702329.0,
      "step": 5975
    },
    {
      "entropy": 5.3050618171691895,
      "epoch": 0.574447646493756,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.1471,
      "mean_token_accuracy": 0.2154633417725563,
      "num_tokens": 13713356.0,
      "step": 5980
    },
    {
      "entropy": 5.2760419845581055,
      "epoch": 0.5749279538904899,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.1355,
      "mean_token_accuracy": 0.20253586918115615,
      "num_tokens": 13724718.0,
      "step": 5985
    },
    {
      "entropy": 5.281339883804321,
      "epoch": 0.5754082612872238,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.1596,
      "mean_token_accuracy": 0.2093519538640976,
      "num_tokens": 13736113.0,
      "step": 5990
    },
    {
      "entropy": 5.454013776779175,
      "epoch": 0.5758885686839578,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.1872,
      "mean_token_accuracy": 0.2056156277656555,
      "num_tokens": 13748322.0,
      "step": 5995
    },
    {
      "entropy": 5.352500104904175,
      "epoch": 0.5763688760806917,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.1919,
      "mean_token_accuracy": 0.19695847779512404,
      "num_tokens": 13759433.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.17189993351509,
      "eval_loss": 5.2146196365356445,
      "eval_mean_token_accuracy": 0.2106242566784959,
      "eval_num_tokens": 13759433.0,
      "eval_runtime": 20.5416,
      "eval_samples_per_second": 1597.487,
      "eval_steps_per_second": 199.692,
      "step": 6000
    },
    {
      "entropy": 5.306978511810303,
      "epoch": 0.5768491834774255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973889689875135,
      "loss": 5.1126,
      "mean_token_accuracy": 0.2072141870856285,
      "num_tokens": 13770181.0,
      "step": 6005
    },
    {
      "entropy": 5.3306817531585695,
      "epoch": 0.5773294908741594,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004973837586336147,
      "loss": 5.1477,
      "mean_token_accuracy": 0.2056859627366066,
      "num_tokens": 13781792.0,
      "step": 6010
    },
    {
      "entropy": 5.371387672424317,
      "epoch": 0.5778097982708934,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004973785431166254,
      "loss": 5.2165,
      "mean_token_accuracy": 0.1999555066227913,
      "num_tokens": 13792101.0,
      "step": 6015
    },
    {
      "entropy": 5.355870532989502,
      "epoch": 0.5782901056676273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973733224366666,
      "loss": 5.1967,
      "mean_token_accuracy": 0.2045416697859764,
      "num_tokens": 13803640.0,
      "step": 6020
    },
    {
      "entropy": 5.396602439880371,
      "epoch": 0.5787704130643612,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973680965938597,
      "loss": 5.1339,
      "mean_token_accuracy": 0.20828541219234467,
      "num_tokens": 13815017.0,
      "step": 6025
    },
    {
      "entropy": 5.471268939971924,
      "epoch": 0.579250720461095,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973628655883258,
      "loss": 5.3498,
      "mean_token_accuracy": 0.1929504841566086,
      "num_tokens": 13826119.0,
      "step": 6030
    },
    {
      "entropy": 5.325129604339599,
      "epoch": 0.579731027857829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973576294201865,
      "loss": 5.171,
      "mean_token_accuracy": 0.20424154549837112,
      "num_tokens": 13837869.0,
      "step": 6035
    },
    {
      "entropy": 5.520558834075928,
      "epoch": 0.5802113352545629,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973523880895633,
      "loss": 5.3258,
      "mean_token_accuracy": 0.189467091858387,
      "num_tokens": 13849333.0,
      "step": 6040
    },
    {
      "entropy": 5.431898689270019,
      "epoch": 0.5806916426512968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004973471415965779,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20163828879594803,
      "num_tokens": 13860648.0,
      "step": 6045
    },
    {
      "entropy": 5.487684631347657,
      "epoch": 0.5811719500480308,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497341889941352,
      "loss": 5.3207,
      "mean_token_accuracy": 0.19369055032730104,
      "num_tokens": 13872979.0,
      "step": 6050
    },
    {
      "entropy": 5.381687974929809,
      "epoch": 0.5816522574447647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973366331240078,
      "loss": 5.22,
      "mean_token_accuracy": 0.19978900402784347,
      "num_tokens": 13884363.0,
      "step": 6055
    },
    {
      "entropy": 5.368413972854614,
      "epoch": 0.5821325648414986,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497331371144667,
      "loss": 5.1937,
      "mean_token_accuracy": 0.19701047241687775,
      "num_tokens": 13895424.0,
      "step": 6060
    },
    {
      "entropy": 5.420424938201904,
      "epoch": 0.5826128722382324,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973261040034521,
      "loss": 5.3199,
      "mean_token_accuracy": 0.1924986258149147,
      "num_tokens": 13907319.0,
      "step": 6065
    },
    {
      "entropy": 5.286451292037964,
      "epoch": 0.5830931796349664,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973208317004852,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20716721713542938,
      "num_tokens": 13920013.0,
      "step": 6070
    },
    {
      "entropy": 5.358293342590332,
      "epoch": 0.5835734870317003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973155542358889,
      "loss": 5.1224,
      "mean_token_accuracy": 0.20192718803882598,
      "num_tokens": 13932033.0,
      "step": 6075
    },
    {
      "entropy": 5.3162446975708,
      "epoch": 0.5840537944284342,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004973102716097853,
      "loss": 5.136,
      "mean_token_accuracy": 0.21228280961513518,
      "num_tokens": 13943324.0,
      "step": 6080
    },
    {
      "entropy": 5.365341424942017,
      "epoch": 0.5845341018251681,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973049838222973,
      "loss": 5.2169,
      "mean_token_accuracy": 0.21387782841920852,
      "num_tokens": 13954291.0,
      "step": 6085
    },
    {
      "entropy": 5.43151888847351,
      "epoch": 0.5850144092219021,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004972996908735479,
      "loss": 5.3151,
      "mean_token_accuracy": 0.19576199650764464,
      "num_tokens": 13966264.0,
      "step": 6090
    },
    {
      "entropy": 5.4537965774536135,
      "epoch": 0.585494716618636,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972943927636597,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19616830497980117,
      "num_tokens": 13977785.0,
      "step": 6095
    },
    {
      "entropy": 5.328124332427978,
      "epoch": 0.5859750240153698,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972890894927558,
      "loss": 5.1833,
      "mean_token_accuracy": 0.2037198066711426,
      "num_tokens": 13989704.0,
      "step": 6100
    },
    {
      "entropy": 5.399294757843018,
      "epoch": 0.5864553314121037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004972837810609592,
      "loss": 5.1665,
      "mean_token_accuracy": 0.20562218874692917,
      "num_tokens": 14000565.0,
      "step": 6105
    },
    {
      "entropy": 5.290426778793335,
      "epoch": 0.5869356388088377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004972784674683933,
      "loss": 5.0835,
      "mean_token_accuracy": 0.21431584507226945,
      "num_tokens": 14011521.0,
      "step": 6110
    },
    {
      "entropy": 5.309748888015747,
      "epoch": 0.5874159462055716,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972731487151815,
      "loss": 5.1299,
      "mean_token_accuracy": 0.21178604513406754,
      "num_tokens": 14022966.0,
      "step": 6115
    },
    {
      "entropy": 5.457789134979248,
      "epoch": 0.5878962536023055,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004972678248014471,
      "loss": 5.2662,
      "mean_token_accuracy": 0.20152803361415864,
      "num_tokens": 14034905.0,
      "step": 6120
    },
    {
      "entropy": 5.296006965637207,
      "epoch": 0.5883765609990393,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972624957273139,
      "loss": 5.1128,
      "mean_token_accuracy": 0.20513343960046768,
      "num_tokens": 14045816.0,
      "step": 6125
    },
    {
      "entropy": 5.307597351074219,
      "epoch": 0.5888568683957733,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004972571614929055,
      "loss": 5.0906,
      "mean_token_accuracy": 0.204602187871933,
      "num_tokens": 14057316.0,
      "step": 6130
    },
    {
      "entropy": 5.402098369598389,
      "epoch": 0.5893371757925072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972518220983457,
      "loss": 5.2213,
      "mean_token_accuracy": 0.19972933381795882,
      "num_tokens": 14067542.0,
      "step": 6135
    },
    {
      "entropy": 5.303103017807007,
      "epoch": 0.5898174831892411,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972464775437586,
      "loss": 5.092,
      "mean_token_accuracy": 0.20539776235818863,
      "num_tokens": 14079467.0,
      "step": 6140
    },
    {
      "entropy": 5.288284063339233,
      "epoch": 0.590297790585975,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972411278292683,
      "loss": 5.0369,
      "mean_token_accuracy": 0.214233922958374,
      "num_tokens": 14090695.0,
      "step": 6145
    },
    {
      "entropy": 5.334277677536011,
      "epoch": 0.590778097982709,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004972357729549988,
      "loss": 5.0882,
      "mean_token_accuracy": 0.21341056525707244,
      "num_tokens": 14101380.0,
      "step": 6150
    },
    {
      "entropy": 5.30261263847351,
      "epoch": 0.5912584053794429,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972304129210746,
      "loss": 5.1416,
      "mean_token_accuracy": 0.20822506695985793,
      "num_tokens": 14112265.0,
      "step": 6155
    },
    {
      "entropy": 5.358154821395874,
      "epoch": 0.5917387127761767,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972250477276202,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20382460057735444,
      "num_tokens": 14124294.0,
      "step": 6160
    },
    {
      "entropy": 5.372543287277222,
      "epoch": 0.5922190201729106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972196773747599,
      "loss": 5.171,
      "mean_token_accuracy": 0.20522281378507615,
      "num_tokens": 14135230.0,
      "step": 6165
    },
    {
      "entropy": 5.3968939781188965,
      "epoch": 0.5926993275696446,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972143018626186,
      "loss": 5.2534,
      "mean_token_accuracy": 0.19905537217855454,
      "num_tokens": 14147576.0,
      "step": 6170
    },
    {
      "entropy": 5.329421806335449,
      "epoch": 0.5931796349663785,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004972089211913211,
      "loss": 5.1287,
      "mean_token_accuracy": 0.2102440357208252,
      "num_tokens": 14158834.0,
      "step": 6175
    },
    {
      "entropy": 5.39668607711792,
      "epoch": 0.5936599423631124,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972035353609923,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19556810408830644,
      "num_tokens": 14170694.0,
      "step": 6180
    },
    {
      "entropy": 5.349567079544068,
      "epoch": 0.5941402497598463,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971981443717572,
      "loss": 5.1747,
      "mean_token_accuracy": 0.2067970633506775,
      "num_tokens": 14183184.0,
      "step": 6185
    },
    {
      "entropy": 5.46701717376709,
      "epoch": 0.5946205571565802,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971927482237409,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20054755210876465,
      "num_tokens": 14194761.0,
      "step": 6190
    },
    {
      "entropy": 5.333970355987549,
      "epoch": 0.5951008645533141,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971873469170689,
      "loss": 5.077,
      "mean_token_accuracy": 0.20996096581220627,
      "num_tokens": 14205820.0,
      "step": 6195
    },
    {
      "entropy": 5.245835304260254,
      "epoch": 0.595581171950048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004971819404518664,
      "loss": 5.0945,
      "mean_token_accuracy": 0.20841328501701356,
      "num_tokens": 14217826.0,
      "step": 6200
    },
    {
      "entropy": 5.3001384258270265,
      "epoch": 0.5960614793467819,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000497176528828259,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21605950444936753,
      "num_tokens": 14228632.0,
      "step": 6205
    },
    {
      "entropy": 5.387531566619873,
      "epoch": 0.5965417867435159,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971711120463722,
      "loss": 5.296,
      "mean_token_accuracy": 0.1957363396883011,
      "num_tokens": 14240231.0,
      "step": 6210
    },
    {
      "entropy": 5.439971780776977,
      "epoch": 0.5970220941402498,
      "grad_norm": 1.0,
      "learning_rate": 0.000497165690106332,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20782527029514314,
      "num_tokens": 14251718.0,
      "step": 6215
    },
    {
      "entropy": 5.36957859992981,
      "epoch": 0.5975024015369836,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004971602630082642,
      "loss": 5.1783,
      "mean_token_accuracy": 0.2077389195561409,
      "num_tokens": 14263129.0,
      "step": 6220
    },
    {
      "entropy": 5.275402593612671,
      "epoch": 0.5979827089337176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971548307522947,
      "loss": 5.0912,
      "mean_token_accuracy": 0.20676200836896896,
      "num_tokens": 14274465.0,
      "step": 6225
    },
    {
      "entropy": 5.432278680801391,
      "epoch": 0.5984630163304515,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971493933385498,
      "loss": 5.2675,
      "mean_token_accuracy": 0.2048093259334564,
      "num_tokens": 14286190.0,
      "step": 6230
    },
    {
      "entropy": 5.438970947265625,
      "epoch": 0.5989433237271854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004971439507671556,
      "loss": 5.2192,
      "mean_token_accuracy": 0.20013955533504485,
      "num_tokens": 14297638.0,
      "step": 6235
    },
    {
      "entropy": 5.327042865753174,
      "epoch": 0.5994236311239193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971385030382384,
      "loss": 5.1522,
      "mean_token_accuracy": 0.2074238583445549,
      "num_tokens": 14309773.0,
      "step": 6240
    },
    {
      "entropy": 5.352311897277832,
      "epoch": 0.5999039385206533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971330501519248,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20862484276294707,
      "num_tokens": 14320543.0,
      "step": 6245
    },
    {
      "entropy": 5.395197534561158,
      "epoch": 0.6003842459173871,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971275921083414,
      "loss": 5.2093,
      "mean_token_accuracy": 0.20144042670726775,
      "num_tokens": 14332076.0,
      "step": 6250
    },
    {
      "entropy": 5.317584037780762,
      "epoch": 0.600864553314121,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497122128907615,
      "loss": 5.129,
      "mean_token_accuracy": 0.20920995026826858,
      "num_tokens": 14343768.0,
      "step": 6255
    },
    {
      "entropy": 5.310052871704102,
      "epoch": 0.6013448607108549,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971166605498722,
      "loss": 5.1646,
      "mean_token_accuracy": 0.2052968069911003,
      "num_tokens": 14357051.0,
      "step": 6260
    },
    {
      "entropy": 5.353819084167481,
      "epoch": 0.6018251681075889,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971111870352401,
      "loss": 5.1891,
      "mean_token_accuracy": 0.20196061283349992,
      "num_tokens": 14368800.0,
      "step": 6265
    },
    {
      "entropy": 5.3648130893707275,
      "epoch": 0.6023054755043228,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971057083638458,
      "loss": 5.172,
      "mean_token_accuracy": 0.20868575274944307,
      "num_tokens": 14379617.0,
      "step": 6270
    },
    {
      "entropy": 5.339131259918213,
      "epoch": 0.6027857829010567,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004971002245358166,
      "loss": 5.1452,
      "mean_token_accuracy": 0.213275907933712,
      "num_tokens": 14391454.0,
      "step": 6275
    },
    {
      "entropy": 5.36895980834961,
      "epoch": 0.6032660902977905,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004970947355512795,
      "loss": 5.1412,
      "mean_token_accuracy": 0.2128308355808258,
      "num_tokens": 14402379.0,
      "step": 6280
    },
    {
      "entropy": 5.328395509719849,
      "epoch": 0.6037463976945245,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970892414103622,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20732113867998123,
      "num_tokens": 14415040.0,
      "step": 6285
    },
    {
      "entropy": 5.362541007995605,
      "epoch": 0.6042267050912584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970837421131921,
      "loss": 5.2517,
      "mean_token_accuracy": 0.19479774683713913,
      "num_tokens": 14426677.0,
      "step": 6290
    },
    {
      "entropy": 5.357956600189209,
      "epoch": 0.6047070124879923,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970782376598972,
      "loss": 5.1422,
      "mean_token_accuracy": 0.206345434486866,
      "num_tokens": 14436676.0,
      "step": 6295
    },
    {
      "entropy": 5.376431560516357,
      "epoch": 0.6051873198847262,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970727280506048,
      "loss": 5.1763,
      "mean_token_accuracy": 0.2027482271194458,
      "num_tokens": 14448294.0,
      "step": 6300
    },
    {
      "entropy": 5.328565835952759,
      "epoch": 0.6056676272814602,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004970672132854431,
      "loss": 5.156,
      "mean_token_accuracy": 0.20563995242118835,
      "num_tokens": 14460642.0,
      "step": 6305
    },
    {
      "entropy": 5.29191780090332,
      "epoch": 0.6061479346781941,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970616933645403,
      "loss": 5.0616,
      "mean_token_accuracy": 0.21846782863140107,
      "num_tokens": 14471370.0,
      "step": 6310
    },
    {
      "entropy": 5.412834596633911,
      "epoch": 0.6066282420749279,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970561682880242,
      "loss": 5.2184,
      "mean_token_accuracy": 0.20499045699834822,
      "num_tokens": 14482358.0,
      "step": 6315
    },
    {
      "entropy": 5.385143852233886,
      "epoch": 0.6071085494716618,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970506380560233,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20311239063739778,
      "num_tokens": 14494413.0,
      "step": 6320
    },
    {
      "entropy": 5.395890903472901,
      "epoch": 0.6075888568683958,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004970451026686659,
      "loss": 5.1607,
      "mean_token_accuracy": 0.21031228005886077,
      "num_tokens": 14506370.0,
      "step": 6325
    },
    {
      "entropy": 5.366713953018189,
      "epoch": 0.6080691642651297,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970395621260806,
      "loss": 5.1524,
      "mean_token_accuracy": 0.21305836588144303,
      "num_tokens": 14517198.0,
      "step": 6330
    },
    {
      "entropy": 5.424633312225342,
      "epoch": 0.6085494716618636,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497034016428396,
      "loss": 5.2514,
      "mean_token_accuracy": 0.1964514061808586,
      "num_tokens": 14529434.0,
      "step": 6335
    },
    {
      "entropy": 5.307683801651001,
      "epoch": 0.6090297790585975,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970284655757409,
      "loss": 5.1426,
      "mean_token_accuracy": 0.2103516861796379,
      "num_tokens": 14540857.0,
      "step": 6340
    },
    {
      "entropy": 5.319919157028198,
      "epoch": 0.6095100864553314,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970229095682439,
      "loss": 5.1554,
      "mean_token_accuracy": 0.20560289323329925,
      "num_tokens": 14552594.0,
      "step": 6345
    },
    {
      "entropy": 5.375809860229492,
      "epoch": 0.6099903938520653,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970173484060344,
      "loss": 5.2158,
      "mean_token_accuracy": 0.2014269471168518,
      "num_tokens": 14563599.0,
      "step": 6350
    },
    {
      "entropy": 5.400948429107666,
      "epoch": 0.6104707012487992,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004970117820892414,
      "loss": 5.1483,
      "mean_token_accuracy": 0.20366021990776062,
      "num_tokens": 14575905.0,
      "step": 6355
    },
    {
      "entropy": 5.380741024017334,
      "epoch": 0.6109510086455331,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970062106179939,
      "loss": 5.1825,
      "mean_token_accuracy": 0.21184269338846207,
      "num_tokens": 14587800.0,
      "step": 6360
    },
    {
      "entropy": 5.326152706146241,
      "epoch": 0.6114313160422671,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970006339924214,
      "loss": 5.134,
      "mean_token_accuracy": 0.21315738260746003,
      "num_tokens": 14600654.0,
      "step": 6365
    },
    {
      "entropy": 5.393792629241943,
      "epoch": 0.611911623439001,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969950522126534,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20788408666849137,
      "num_tokens": 14611985.0,
      "step": 6370
    },
    {
      "entropy": 5.295234346389771,
      "epoch": 0.6123919308357348,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969894652788196,
      "loss": 5.1226,
      "mean_token_accuracy": 0.21155203580856324,
      "num_tokens": 14625004.0,
      "step": 6375
    },
    {
      "entropy": 5.367025756835938,
      "epoch": 0.6128722382324687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969838731910494,
      "loss": 5.1756,
      "mean_token_accuracy": 0.20544978678226472,
      "num_tokens": 14635381.0,
      "step": 6380
    },
    {
      "entropy": 5.317724943161011,
      "epoch": 0.6133525456292027,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004969782759494729,
      "loss": 5.1284,
      "mean_token_accuracy": 0.21093843430280684,
      "num_tokens": 14646582.0,
      "step": 6385
    },
    {
      "entropy": 5.312413883209229,
      "epoch": 0.6138328530259366,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049697267355422,
      "loss": 5.2048,
      "mean_token_accuracy": 0.21181879341602325,
      "num_tokens": 14658024.0,
      "step": 6390
    },
    {
      "entropy": 5.485131025314331,
      "epoch": 0.6143131604226705,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969670660054208,
      "loss": 5.2383,
      "mean_token_accuracy": 0.19830967038869857,
      "num_tokens": 14669933.0,
      "step": 6395
    },
    {
      "entropy": 5.379800939559937,
      "epoch": 0.6147934678194045,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969614533032054,
      "loss": 5.1814,
      "mean_token_accuracy": 0.20320038497447968,
      "num_tokens": 14681331.0,
      "step": 6400
    },
    {
      "entropy": 5.237337875366211,
      "epoch": 0.6152737752161384,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969558354477041,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21839836984872818,
      "num_tokens": 14691361.0,
      "step": 6405
    },
    {
      "entropy": 5.315851306915283,
      "epoch": 0.6157540826128722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969502124390474,
      "loss": 5.139,
      "mean_token_accuracy": 0.20834049731492996,
      "num_tokens": 14701791.0,
      "step": 6410
    },
    {
      "entropy": 5.390240621566773,
      "epoch": 0.6162343900096061,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004969445842773658,
      "loss": 5.206,
      "mean_token_accuracy": 0.19944172948598862,
      "num_tokens": 14713672.0,
      "step": 6415
    },
    {
      "entropy": 5.303598499298095,
      "epoch": 0.6167146974063401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049693895096279,
      "loss": 5.0381,
      "mean_token_accuracy": 0.2155421108007431,
      "num_tokens": 14724789.0,
      "step": 6420
    },
    {
      "entropy": 5.286634874343872,
      "epoch": 0.617195004803074,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969333124954508,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20550125539302827,
      "num_tokens": 14737212.0,
      "step": 6425
    },
    {
      "entropy": 5.292817258834839,
      "epoch": 0.6176753121998079,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004969276688754791,
      "loss": 5.1368,
      "mean_token_accuracy": 0.20599037855863572,
      "num_tokens": 14748387.0,
      "step": 6430
    },
    {
      "entropy": 5.363704776763916,
      "epoch": 0.6181556195965417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969220201030059,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20211084485054015,
      "num_tokens": 14758477.0,
      "step": 6435
    },
    {
      "entropy": 5.370852279663086,
      "epoch": 0.6186359269932757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969163661781624,
      "loss": 5.2579,
      "mean_token_accuracy": 0.20346666276454925,
      "num_tokens": 14769650.0,
      "step": 6440
    },
    {
      "entropy": 5.252781963348388,
      "epoch": 0.6191162343900096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969107071010798,
      "loss": 5.0366,
      "mean_token_accuracy": 0.21276762336492538,
      "num_tokens": 14780988.0,
      "step": 6445
    },
    {
      "entropy": 5.361257648468017,
      "epoch": 0.6195965417867435,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004969050428718895,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20762884467840195,
      "num_tokens": 14792458.0,
      "step": 6450
    },
    {
      "entropy": 5.369265222549439,
      "epoch": 0.6200768491834774,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496899373490723,
      "loss": 5.2092,
      "mean_token_accuracy": 0.20686816722154616,
      "num_tokens": 14804772.0,
      "step": 6455
    },
    {
      "entropy": 5.3651207447052,
      "epoch": 0.6205571565802114,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496893698957712,
      "loss": 5.1392,
      "mean_token_accuracy": 0.20470530688762664,
      "num_tokens": 14816126.0,
      "step": 6460
    },
    {
      "entropy": 5.397782659530639,
      "epoch": 0.6210374639769453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968880192729882,
      "loss": 5.2598,
      "mean_token_accuracy": 0.20079650729894638,
      "num_tokens": 14829376.0,
      "step": 6465
    },
    {
      "entropy": 5.312540054321289,
      "epoch": 0.6215177713736791,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968823344366835,
      "loss": 5.071,
      "mean_token_accuracy": 0.2153061196208,
      "num_tokens": 14841098.0,
      "step": 6470
    },
    {
      "entropy": 5.303245687484742,
      "epoch": 0.621998078770413,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968766444489298,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20885020941495896,
      "num_tokens": 14852690.0,
      "step": 6475
    },
    {
      "entropy": 5.3362432479858395,
      "epoch": 0.622478386167147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004968709493098593,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20867665559053422,
      "num_tokens": 14863327.0,
      "step": 6480
    },
    {
      "entropy": 5.387118530273438,
      "epoch": 0.6229586935638809,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004968652490196041,
      "loss": 5.2583,
      "mean_token_accuracy": 0.20046956092119217,
      "num_tokens": 14875213.0,
      "step": 6485
    },
    {
      "entropy": 5.348819875717163,
      "epoch": 0.6234390009606148,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004968595435782967,
      "loss": 5.1514,
      "mean_token_accuracy": 0.22016421258449553,
      "num_tokens": 14886129.0,
      "step": 6490
    },
    {
      "entropy": 5.426475858688354,
      "epoch": 0.6239193083573487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968538329860695,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19709108769893646,
      "num_tokens": 14897217.0,
      "step": 6495
    },
    {
      "entropy": 5.355967044830322,
      "epoch": 0.6243996157540826,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004968481172430549,
      "loss": 5.1945,
      "mean_token_accuracy": 0.20492311269044877,
      "num_tokens": 14908438.0,
      "step": 6500
    },
    {
      "entropy": 5.276374959945679,
      "epoch": 0.6248799231508165,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000496842396349386,
      "loss": 5.0919,
      "mean_token_accuracy": 0.2137288123369217,
      "num_tokens": 14920472.0,
      "step": 6505
    },
    {
      "entropy": 5.299534654617309,
      "epoch": 0.6253602305475504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968366703051952,
      "loss": 5.1441,
      "mean_token_accuracy": 0.209611739218235,
      "num_tokens": 14932075.0,
      "step": 6510
    },
    {
      "entropy": 5.330506467819214,
      "epoch": 0.6258405379442843,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004968309391106157,
      "loss": 5.1724,
      "mean_token_accuracy": 0.21013996005058289,
      "num_tokens": 14942971.0,
      "step": 6515
    },
    {
      "entropy": 5.315490245819092,
      "epoch": 0.6263208453410183,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968252027657806,
      "loss": 5.041,
      "mean_token_accuracy": 0.2154913440346718,
      "num_tokens": 14954288.0,
      "step": 6520
    },
    {
      "entropy": 5.256664657592774,
      "epoch": 0.6268011527377522,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968194612708229,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20832486301660538,
      "num_tokens": 14966017.0,
      "step": 6525
    },
    {
      "entropy": 5.360432958602905,
      "epoch": 0.627281460134486,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968137146258759,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20783136337995528,
      "num_tokens": 14978022.0,
      "step": 6530
    },
    {
      "entropy": 5.269650602340699,
      "epoch": 0.6277617675312199,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968079628310732,
      "loss": 5.0852,
      "mean_token_accuracy": 0.21249632388353348,
      "num_tokens": 14990370.0,
      "step": 6535
    },
    {
      "entropy": 5.248974943161011,
      "epoch": 0.6282420749279539,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968022058865482,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21422789245843887,
      "num_tokens": 15001535.0,
      "step": 6540
    },
    {
      "entropy": 5.416878795623779,
      "epoch": 0.6287223823246878,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967964437924346,
      "loss": 5.1645,
      "mean_token_accuracy": 0.21044517308473587,
      "num_tokens": 15012560.0,
      "step": 6545
    },
    {
      "entropy": 5.257675266265869,
      "epoch": 0.6292026897214217,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967906765488662,
      "loss": 5.0835,
      "mean_token_accuracy": 0.21322489976882936,
      "num_tokens": 15024594.0,
      "step": 6550
    },
    {
      "entropy": 5.326331043243409,
      "epoch": 0.6296829971181557,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967849041559769,
      "loss": 5.1057,
      "mean_token_accuracy": 0.21285986304283142,
      "num_tokens": 15034735.0,
      "step": 6555
    },
    {
      "entropy": 5.396697187423706,
      "epoch": 0.6301633045148896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004967791266139006,
      "loss": 5.2209,
      "mean_token_accuracy": 0.20284678786993027,
      "num_tokens": 15046955.0,
      "step": 6560
    },
    {
      "entropy": 5.3369475364685055,
      "epoch": 0.6306436119116234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967733439227716,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20890088230371476,
      "num_tokens": 15058250.0,
      "step": 6565
    },
    {
      "entropy": 5.359533166885376,
      "epoch": 0.6311239193083573,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496767556082724,
      "loss": 5.213,
      "mean_token_accuracy": 0.20450974106788636,
      "num_tokens": 15069451.0,
      "step": 6570
    },
    {
      "entropy": 5.3008757591247555,
      "epoch": 0.6316042267050913,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004967617630938924,
      "loss": 5.171,
      "mean_token_accuracy": 0.21107075065374375,
      "num_tokens": 15080920.0,
      "step": 6575
    },
    {
      "entropy": 5.284978914260864,
      "epoch": 0.6320845341018252,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496755964956411,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20591130405664443,
      "num_tokens": 15091579.0,
      "step": 6580
    },
    {
      "entropy": 5.308378839492798,
      "epoch": 0.6325648414985591,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004967501616704147,
      "loss": 5.212,
      "mean_token_accuracy": 0.20694429576396942,
      "num_tokens": 15103097.0,
      "step": 6585
    },
    {
      "entropy": 5.305421543121338,
      "epoch": 0.633045148895293,
      "grad_norm": 0.9375,
      "learning_rate": 0.000496744353236038,
      "loss": 5.094,
      "mean_token_accuracy": 0.21201552748680114,
      "num_tokens": 15114646.0,
      "step": 6590
    },
    {
      "entropy": 5.406478595733643,
      "epoch": 0.633525456292027,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496738539653416,
      "loss": 5.1546,
      "mean_token_accuracy": 0.2054781585931778,
      "num_tokens": 15126019.0,
      "step": 6595
    },
    {
      "entropy": 5.364590787887574,
      "epoch": 0.6340057636887608,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967327209226835,
      "loss": 5.1486,
      "mean_token_accuracy": 0.21009373515844346,
      "num_tokens": 15136132.0,
      "step": 6600
    },
    {
      "entropy": 5.3448058605194095,
      "epoch": 0.6344860710854947,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004967268970439755,
      "loss": 5.2203,
      "mean_token_accuracy": 0.20390277355909348,
      "num_tokens": 15149016.0,
      "step": 6605
    },
    {
      "entropy": 5.399990177154541,
      "epoch": 0.6349663784822286,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967210680174274,
      "loss": 5.1985,
      "mean_token_accuracy": 0.2116396740078926,
      "num_tokens": 15160302.0,
      "step": 6610
    },
    {
      "entropy": 5.350642824172974,
      "epoch": 0.6354466858789626,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967152338431745,
      "loss": 5.1716,
      "mean_token_accuracy": 0.20768017321825027,
      "num_tokens": 15171770.0,
      "step": 6615
    },
    {
      "entropy": 5.319510459899902,
      "epoch": 0.6359269932756965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967093945213523,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21049544364213943,
      "num_tokens": 15182283.0,
      "step": 6620
    },
    {
      "entropy": 5.277123641967774,
      "epoch": 0.6364073006724303,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967035500520962,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20742365568876267,
      "num_tokens": 15193917.0,
      "step": 6625
    },
    {
      "entropy": 5.453793668746949,
      "epoch": 0.6368876080691642,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966977004355421,
      "loss": 5.2382,
      "mean_token_accuracy": 0.20050635039806367,
      "num_tokens": 15205854.0,
      "step": 6630
    },
    {
      "entropy": 5.287836170196533,
      "epoch": 0.6373679154658982,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004966918456718256,
      "loss": 5.0962,
      "mean_token_accuracy": 0.21320753097534179,
      "num_tokens": 15217884.0,
      "step": 6635
    },
    {
      "entropy": 5.347429323196411,
      "epoch": 0.6378482228626321,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966859857610828,
      "loss": 5.223,
      "mean_token_accuracy": 0.20406776517629624,
      "num_tokens": 15229493.0,
      "step": 6640
    },
    {
      "entropy": 5.344024515151977,
      "epoch": 0.638328530259366,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004966801207034497,
      "loss": 5.1791,
      "mean_token_accuracy": 0.20313767045736314,
      "num_tokens": 15241402.0,
      "step": 6645
    },
    {
      "entropy": 5.389021968841552,
      "epoch": 0.6388088376560999,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966742504990624,
      "loss": 5.1781,
      "mean_token_accuracy": 0.20621784329414367,
      "num_tokens": 15252981.0,
      "step": 6650
    },
    {
      "entropy": 5.3174539566040036,
      "epoch": 0.6392891450528339,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004966683751480572,
      "loss": 5.1175,
      "mean_token_accuracy": 0.21280963718891144,
      "num_tokens": 15264171.0,
      "step": 6655
    },
    {
      "entropy": 5.3186562061309814,
      "epoch": 0.6397694524495677,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966624946505706,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20979261100292207,
      "num_tokens": 15275595.0,
      "step": 6660
    },
    {
      "entropy": 5.387149906158447,
      "epoch": 0.6402497598463016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966566090067391,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20688546895980836,
      "num_tokens": 15286074.0,
      "step": 6665
    },
    {
      "entropy": 5.379571866989136,
      "epoch": 0.6407300672430355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966507182166993,
      "loss": 5.1504,
      "mean_token_accuracy": 0.2026302382349968,
      "num_tokens": 15296230.0,
      "step": 6670
    },
    {
      "entropy": 5.316071510314941,
      "epoch": 0.6412103746397695,
      "grad_norm": 0.90625,
      "learning_rate": 0.000496644822280588,
      "loss": 5.1324,
      "mean_token_accuracy": 0.20843050479888917,
      "num_tokens": 15307428.0,
      "step": 6675
    },
    {
      "entropy": 5.408726644515991,
      "epoch": 0.6416906820365034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000496638921198542,
      "loss": 5.2482,
      "mean_token_accuracy": 0.19868980199098588,
      "num_tokens": 15318796.0,
      "step": 6680
    },
    {
      "entropy": 5.310985040664673,
      "epoch": 0.6421709894332372,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966330149706984,
      "loss": 5.1461,
      "mean_token_accuracy": 0.2115129992365837,
      "num_tokens": 15330914.0,
      "step": 6685
    },
    {
      "entropy": 5.282835483551025,
      "epoch": 0.6426512968299711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004966271035971944,
      "loss": 5.113,
      "mean_token_accuracy": 0.20557762086391448,
      "num_tokens": 15341756.0,
      "step": 6690
    },
    {
      "entropy": 5.40358681678772,
      "epoch": 0.6431316042267051,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004966211870781671,
      "loss": 5.2095,
      "mean_token_accuracy": 0.20149461328983306,
      "num_tokens": 15353217.0,
      "step": 6695
    },
    {
      "entropy": 5.3673028469085695,
      "epoch": 0.643611911623439,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496615265413754,
      "loss": 5.1021,
      "mean_token_accuracy": 0.2055822029709816,
      "num_tokens": 15363813.0,
      "step": 6700
    },
    {
      "entropy": 5.288237810134888,
      "epoch": 0.6440922190201729,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966093386040923,
      "loss": 5.1151,
      "mean_token_accuracy": 0.2095430463552475,
      "num_tokens": 15375223.0,
      "step": 6705
    },
    {
      "entropy": 5.326645851135254,
      "epoch": 0.6445725264169068,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049660340664932,
      "loss": 5.1194,
      "mean_token_accuracy": 0.21088357269763947,
      "num_tokens": 15385523.0,
      "step": 6710
    },
    {
      "entropy": 5.3942523956298825,
      "epoch": 0.6450528338136408,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965974695495746,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20370138734579085,
      "num_tokens": 15397262.0,
      "step": 6715
    },
    {
      "entropy": 5.243142366409302,
      "epoch": 0.6455331412103746,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965915273049938,
      "loss": 5.1069,
      "mean_token_accuracy": 0.21149415522813797,
      "num_tokens": 15409043.0,
      "step": 6720
    },
    {
      "entropy": 5.320618915557861,
      "epoch": 0.6460134486071085,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965855799157158,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20918045789003373,
      "num_tokens": 15420859.0,
      "step": 6725
    },
    {
      "entropy": 5.36569995880127,
      "epoch": 0.6464937560038425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965796273818787,
      "loss": 5.1815,
      "mean_token_accuracy": 0.20250088423490525,
      "num_tokens": 15433650.0,
      "step": 6730
    },
    {
      "entropy": 5.275715637207031,
      "epoch": 0.6469740634005764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965736697036206,
      "loss": 5.0704,
      "mean_token_accuracy": 0.21408701539039612,
      "num_tokens": 15445158.0,
      "step": 6735
    },
    {
      "entropy": 5.418146705627441,
      "epoch": 0.6474543707973103,
      "grad_norm": 0.875,
      "learning_rate": 0.0004965677068810798,
      "loss": 5.3321,
      "mean_token_accuracy": 0.19387490600347518,
      "num_tokens": 15456974.0,
      "step": 6740
    },
    {
      "entropy": 5.41830039024353,
      "epoch": 0.6479346781940442,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965617389143947,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20549649000167847,
      "num_tokens": 15467395.0,
      "step": 6745
    },
    {
      "entropy": 5.293714284896851,
      "epoch": 0.6484149855907781,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000496555765803704,
      "loss": 5.193,
      "mean_token_accuracy": 0.2080518677830696,
      "num_tokens": 15479396.0,
      "step": 6750
    },
    {
      "entropy": 5.310993337631226,
      "epoch": 0.648895292987512,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965497875491462,
      "loss": 5.1267,
      "mean_token_accuracy": 0.21137345135211943,
      "num_tokens": 15490992.0,
      "step": 6755
    },
    {
      "entropy": 5.4591199398040775,
      "epoch": 0.6493756003842459,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004965438041508604,
      "loss": 5.2087,
      "mean_token_accuracy": 0.19662567228078842,
      "num_tokens": 15502413.0,
      "step": 6760
    },
    {
      "entropy": 5.374146842956543,
      "epoch": 0.6498559077809798,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965378156089851,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20146453827619554,
      "num_tokens": 15512588.0,
      "step": 6765
    },
    {
      "entropy": 5.319894742965698,
      "epoch": 0.6503362151777138,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965318219236597,
      "loss": 5.1252,
      "mean_token_accuracy": 0.2045357957482338,
      "num_tokens": 15523252.0,
      "step": 6770
    },
    {
      "entropy": 5.309787845611572,
      "epoch": 0.6508165225744477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965258230950232,
      "loss": 5.222,
      "mean_token_accuracy": 0.2023079663515091,
      "num_tokens": 15534975.0,
      "step": 6775
    },
    {
      "entropy": 5.322136354446411,
      "epoch": 0.6512968299711815,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965198191232148,
      "loss": 5.091,
      "mean_token_accuracy": 0.2112494647502899,
      "num_tokens": 15545709.0,
      "step": 6780
    },
    {
      "entropy": 5.382352638244629,
      "epoch": 0.6517771373679154,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496513810008374,
      "loss": 5.1852,
      "mean_token_accuracy": 0.21332192420959473,
      "num_tokens": 15557270.0,
      "step": 6785
    },
    {
      "entropy": 5.391598176956177,
      "epoch": 0.6522574447646494,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965077957506403,
      "loss": 5.1415,
      "mean_token_accuracy": 0.20827943235635757,
      "num_tokens": 15569156.0,
      "step": 6790
    },
    {
      "entropy": 5.294995450973511,
      "epoch": 0.6527377521613833,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965017763501533,
      "loss": 5.0965,
      "mean_token_accuracy": 0.20894810557365417,
      "num_tokens": 15579270.0,
      "step": 6795
    },
    {
      "entropy": 5.2520428657531735,
      "epoch": 0.6532180595581172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964957518070529,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21518850773572923,
      "num_tokens": 15589912.0,
      "step": 6800
    },
    {
      "entropy": 5.378147077560425,
      "epoch": 0.6536983669548511,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964897221214788,
      "loss": 5.1614,
      "mean_token_accuracy": 0.20349239856004714,
      "num_tokens": 15601088.0,
      "step": 6805
    },
    {
      "entropy": 5.376770877838135,
      "epoch": 0.654178674351585,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496483687293571,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20756430178880692,
      "num_tokens": 15612520.0,
      "step": 6810
    },
    {
      "entropy": 5.302024173736572,
      "epoch": 0.6546589817483189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964776473234696,
      "loss": 5.0783,
      "mean_token_accuracy": 0.21862724870443345,
      "num_tokens": 15623991.0,
      "step": 6815
    },
    {
      "entropy": 5.218190479278564,
      "epoch": 0.6551392891450528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496471602211315,
      "loss": 5.0159,
      "mean_token_accuracy": 0.21332069784402846,
      "num_tokens": 15635389.0,
      "step": 6820
    },
    {
      "entropy": 5.326016759872436,
      "epoch": 0.6556195965417867,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004964655519572475,
      "loss": 5.1179,
      "mean_token_accuracy": 0.2181401401758194,
      "num_tokens": 15646427.0,
      "step": 6825
    },
    {
      "entropy": 5.2474266529083256,
      "epoch": 0.6560999039385207,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004964594965614072,
      "loss": 5.1114,
      "mean_token_accuracy": 0.2103773534297943,
      "num_tokens": 15657518.0,
      "step": 6830
    },
    {
      "entropy": 5.351989126205444,
      "epoch": 0.6565802113352546,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964534360239353,
      "loss": 5.142,
      "mean_token_accuracy": 0.20573293566703796,
      "num_tokens": 15669775.0,
      "step": 6835
    },
    {
      "entropy": 5.312732362747193,
      "epoch": 0.6570605187319885,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496447370344972,
      "loss": 5.1466,
      "mean_token_accuracy": 0.20142304003238679,
      "num_tokens": 15682566.0,
      "step": 6840
    },
    {
      "entropy": 5.272958993911743,
      "epoch": 0.6575408261287223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004964412995246584,
      "loss": 5.1212,
      "mean_token_accuracy": 0.2153865084052086,
      "num_tokens": 15693168.0,
      "step": 6845
    },
    {
      "entropy": 5.333586740493774,
      "epoch": 0.6580211335254563,
      "grad_norm": 1.0,
      "learning_rate": 0.0004964352235631354,
      "loss": 5.1441,
      "mean_token_accuracy": 0.2094581201672554,
      "num_tokens": 15703879.0,
      "step": 6850
    },
    {
      "entropy": 5.400853109359741,
      "epoch": 0.6585014409221902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000496429142460544,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20100296437740325,
      "num_tokens": 15716015.0,
      "step": 6855
    },
    {
      "entropy": 5.411762380599976,
      "epoch": 0.6589817483189241,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004964230562170254,
      "loss": 5.1781,
      "mean_token_accuracy": 0.2079763263463974,
      "num_tokens": 15728254.0,
      "step": 6860
    },
    {
      "entropy": 5.31914324760437,
      "epoch": 0.659462055715658,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004964169648327209,
      "loss": 5.15,
      "mean_token_accuracy": 0.20646823644638063,
      "num_tokens": 15738778.0,
      "step": 6865
    },
    {
      "entropy": 5.350067758560181,
      "epoch": 0.659942363112392,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496410868307772,
      "loss": 5.1459,
      "mean_token_accuracy": 0.2045704886317253,
      "num_tokens": 15750305.0,
      "step": 6870
    },
    {
      "entropy": 5.372748517990113,
      "epoch": 0.6604226705091258,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004964047666423203,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20822612941265106,
      "num_tokens": 15761303.0,
      "step": 6875
    },
    {
      "entropy": 5.299829530715942,
      "epoch": 0.6609029779058597,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963986598365072,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21532647013664247,
      "num_tokens": 15773095.0,
      "step": 6880
    },
    {
      "entropy": 5.351714468002319,
      "epoch": 0.6613832853025937,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963925478904746,
      "loss": 5.1633,
      "mean_token_accuracy": 0.21195616871118544,
      "num_tokens": 15784405.0,
      "step": 6885
    },
    {
      "entropy": 5.235157918930054,
      "epoch": 0.6618635926993276,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963864308043645,
      "loss": 5.0892,
      "mean_token_accuracy": 0.20711841732263564,
      "num_tokens": 15795178.0,
      "step": 6890
    },
    {
      "entropy": 5.290901565551758,
      "epoch": 0.6623439000960615,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963803085783189,
      "loss": 5.1197,
      "mean_token_accuracy": 0.2067147433757782,
      "num_tokens": 15806205.0,
      "step": 6895
    },
    {
      "entropy": 5.313950729370117,
      "epoch": 0.6628242074927954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963741812124799,
      "loss": 5.0362,
      "mean_token_accuracy": 0.21753399223089218,
      "num_tokens": 15817474.0,
      "step": 6900
    },
    {
      "entropy": 5.268809366226196,
      "epoch": 0.6633045148895294,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963680487069898,
      "loss": 5.1482,
      "mean_token_accuracy": 0.2037314921617508,
      "num_tokens": 15829728.0,
      "step": 6905
    },
    {
      "entropy": 5.385620164871216,
      "epoch": 0.6637848222862632,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963619110619908,
      "loss": 5.2262,
      "mean_token_accuracy": 0.20655222982168198,
      "num_tokens": 15840082.0,
      "step": 6910
    },
    {
      "entropy": 5.380099678039551,
      "epoch": 0.6642651296829971,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963557682776256,
      "loss": 5.115,
      "mean_token_accuracy": 0.20173939764499665,
      "num_tokens": 15851450.0,
      "step": 6915
    },
    {
      "entropy": 5.400586414337158,
      "epoch": 0.664745437079731,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963496203540368,
      "loss": 5.2625,
      "mean_token_accuracy": 0.20122259259223937,
      "num_tokens": 15864168.0,
      "step": 6920
    },
    {
      "entropy": 5.3505774974823,
      "epoch": 0.665225744476465,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963434672913671,
      "loss": 5.1837,
      "mean_token_accuracy": 0.2109785258769989,
      "num_tokens": 15875634.0,
      "step": 6925
    },
    {
      "entropy": 5.464935636520385,
      "epoch": 0.6657060518731989,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963373090897592,
      "loss": 5.2124,
      "mean_token_accuracy": 0.20236394703388214,
      "num_tokens": 15888411.0,
      "step": 6930
    },
    {
      "entropy": 5.358474016189575,
      "epoch": 0.6661863592699327,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963311457493563,
      "loss": 5.1856,
      "mean_token_accuracy": 0.20183018147945403,
      "num_tokens": 15901084.0,
      "step": 6935
    },
    {
      "entropy": 5.398572540283203,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004963249772703015,
      "loss": 5.2033,
      "mean_token_accuracy": 0.20429981648921966,
      "num_tokens": 15912061.0,
      "step": 6940
    },
    {
      "entropy": 5.398046255111694,
      "epoch": 0.6671469740634006,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963188036527378,
      "loss": 5.2802,
      "mean_token_accuracy": 0.20116444826126098,
      "num_tokens": 15925439.0,
      "step": 6945
    },
    {
      "entropy": 5.292640972137451,
      "epoch": 0.6676272814601345,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963126248968087,
      "loss": 5.0105,
      "mean_token_accuracy": 0.22056220322847367,
      "num_tokens": 15937762.0,
      "step": 6950
    },
    {
      "entropy": 5.226414442062378,
      "epoch": 0.6681075888568684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963064410026577,
      "loss": 5.036,
      "mean_token_accuracy": 0.21784193813800812,
      "num_tokens": 15948656.0,
      "step": 6955
    },
    {
      "entropy": 5.370252323150635,
      "epoch": 0.6685878962536023,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963002519704281,
      "loss": 5.114,
      "mean_token_accuracy": 0.20235869139432908,
      "num_tokens": 15960376.0,
      "step": 6960
    },
    {
      "entropy": 5.392440366744995,
      "epoch": 0.6690682036503363,
      "grad_norm": 1.0,
      "learning_rate": 0.000496294057800264,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20715708136558533,
      "num_tokens": 15971913.0,
      "step": 6965
    },
    {
      "entropy": 5.336017179489136,
      "epoch": 0.6695485110470701,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962878584923089,
      "loss": 5.1875,
      "mean_token_accuracy": 0.20061488151550294,
      "num_tokens": 15984775.0,
      "step": 6970
    },
    {
      "entropy": 5.324187469482422,
      "epoch": 0.670028818443804,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962816540467068,
      "loss": 5.1132,
      "mean_token_accuracy": 0.20296642184257507,
      "num_tokens": 15996717.0,
      "step": 6975
    },
    {
      "entropy": 5.31804633140564,
      "epoch": 0.6705091258405379,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004962754444636017,
      "loss": 5.1132,
      "mean_token_accuracy": 0.21299038529396058,
      "num_tokens": 16007964.0,
      "step": 6980
    },
    {
      "entropy": 5.302518272399903,
      "epoch": 0.6709894332372719,
      "grad_norm": 0.953125,
      "learning_rate": 0.000496269229743138,
      "loss": 5.12,
      "mean_token_accuracy": 0.2090325817465782,
      "num_tokens": 16019185.0,
      "step": 6985
    },
    {
      "entropy": 5.34990234375,
      "epoch": 0.6714697406340058,
      "grad_norm": 1.0,
      "learning_rate": 0.0004962630098854597,
      "loss": 5.1156,
      "mean_token_accuracy": 0.21032043546438217,
      "num_tokens": 16029983.0,
      "step": 6990
    },
    {
      "entropy": 5.316616630554199,
      "epoch": 0.6719500480307397,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004962567848907113,
      "loss": 5.0699,
      "mean_token_accuracy": 0.20968452394008635,
      "num_tokens": 16040574.0,
      "step": 6995
    },
    {
      "entropy": 5.368011045455932,
      "epoch": 0.6724303554274735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962505547590374,
      "loss": 5.2296,
      "mean_token_accuracy": 0.20728860199451446,
      "num_tokens": 16052037.0,
      "step": 7000
    },
    {
      "entropy": 5.393733358383178,
      "epoch": 0.6729106628242075,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004962443194905826,
      "loss": 5.196,
      "mean_token_accuracy": 0.2026839643716812,
      "num_tokens": 16063413.0,
      "step": 7005
    },
    {
      "entropy": 5.387011575698852,
      "epoch": 0.6733909702209414,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004962380790854916,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20336944609880447,
      "num_tokens": 16074373.0,
      "step": 7010
    },
    {
      "entropy": 5.370492839813233,
      "epoch": 0.6738712776176753,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004962318335439094,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19826406985521317,
      "num_tokens": 16086575.0,
      "step": 7015
    },
    {
      "entropy": 5.306420516967774,
      "epoch": 0.6743515850144092,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004962255828659809,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21166469156742096,
      "num_tokens": 16098529.0,
      "step": 7020
    },
    {
      "entropy": 5.295602798461914,
      "epoch": 0.6748318924111432,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962193270518513,
      "loss": 5.1335,
      "mean_token_accuracy": 0.21224457025527954,
      "num_tokens": 16110085.0,
      "step": 7025
    },
    {
      "entropy": 5.29178900718689,
      "epoch": 0.675312199807877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004962130661016659,
      "loss": 5.0076,
      "mean_token_accuracy": 0.21021574437618257,
      "num_tokens": 16120249.0,
      "step": 7030
    },
    {
      "entropy": 5.408480501174926,
      "epoch": 0.6757925072046109,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004962068000155699,
      "loss": 5.295,
      "mean_token_accuracy": 0.19446971565485,
      "num_tokens": 16132645.0,
      "step": 7035
    },
    {
      "entropy": 5.334974527359009,
      "epoch": 0.6762728146013448,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962005287937088,
      "loss": 5.1675,
      "mean_token_accuracy": 0.21253057569265366,
      "num_tokens": 16143808.0,
      "step": 7040
    },
    {
      "entropy": 5.309336805343628,
      "epoch": 0.6767531219980788,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961942524362283,
      "loss": 5.1691,
      "mean_token_accuracy": 0.21257366091012955,
      "num_tokens": 16154309.0,
      "step": 7045
    },
    {
      "entropy": 5.348855829238891,
      "epoch": 0.6772334293948127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961879709432741,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20713894218206405,
      "num_tokens": 16164654.0,
      "step": 7050
    },
    {
      "entropy": 5.409746599197388,
      "epoch": 0.6777137367915466,
      "grad_norm": 1.0,
      "learning_rate": 0.000496181684314992,
      "loss": 5.1632,
      "mean_token_accuracy": 0.20302171260118484,
      "num_tokens": 16177837.0,
      "step": 7055
    },
    {
      "entropy": 5.29151873588562,
      "epoch": 0.6781940441882806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004961753925515279,
      "loss": 5.0592,
      "mean_token_accuracy": 0.21328389942646026,
      "num_tokens": 16189073.0,
      "step": 7060
    },
    {
      "entropy": 5.369131755828858,
      "epoch": 0.6786743515850144,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496169095653028,
      "loss": 5.2359,
      "mean_token_accuracy": 0.2068806990981102,
      "num_tokens": 16200371.0,
      "step": 7065
    },
    {
      "entropy": 5.326678657531739,
      "epoch": 0.6791546589817483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961627936196384,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21685010492801665,
      "num_tokens": 16210526.0,
      "step": 7070
    },
    {
      "entropy": 5.347667503356933,
      "epoch": 0.6796349663784822,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961564864515055,
      "loss": 5.1319,
      "mean_token_accuracy": 0.21133444607257842,
      "num_tokens": 16221687.0,
      "step": 7075
    },
    {
      "entropy": 5.2318473815917965,
      "epoch": 0.6801152737752162,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004961501741487757,
      "loss": 5.0217,
      "mean_token_accuracy": 0.22075052410364152,
      "num_tokens": 16233828.0,
      "step": 7080
    },
    {
      "entropy": 5.2553661346435545,
      "epoch": 0.6805955811719501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961438567115955,
      "loss": 5.0146,
      "mean_token_accuracy": 0.22390031367540358,
      "num_tokens": 16243900.0,
      "step": 7085
    },
    {
      "entropy": 5.280510330200196,
      "epoch": 0.681075888568684,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961375341401116,
      "loss": 5.105,
      "mean_token_accuracy": 0.2133181780576706,
      "num_tokens": 16256347.0,
      "step": 7090
    },
    {
      "entropy": 5.319049263000489,
      "epoch": 0.6815561959654178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961312064344708,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21245002895593643,
      "num_tokens": 16267743.0,
      "step": 7095
    },
    {
      "entropy": 5.236669540405273,
      "epoch": 0.6820365033621518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00049612487359482,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21554897278547286,
      "num_tokens": 16278280.0,
      "step": 7100
    },
    {
      "entropy": 5.234877729415894,
      "epoch": 0.6825168107588857,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961185356213062,
      "loss": 5.0452,
      "mean_token_accuracy": 0.21809980124235154,
      "num_tokens": 16288568.0,
      "step": 7105
    },
    {
      "entropy": 5.294752979278565,
      "epoch": 0.6829971181556196,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004961121925140767,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21069593280553817,
      "num_tokens": 16300730.0,
      "step": 7110
    },
    {
      "entropy": 5.304313468933105,
      "epoch": 0.6834774255523535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961058442732786,
      "loss": 5.1308,
      "mean_token_accuracy": 0.20534905791282654,
      "num_tokens": 16311789.0,
      "step": 7115
    },
    {
      "entropy": 5.305329036712647,
      "epoch": 0.6839577329490875,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960994908990594,
      "loss": 5.101,
      "mean_token_accuracy": 0.20463583320379258,
      "num_tokens": 16324439.0,
      "step": 7120
    },
    {
      "entropy": 5.416294384002685,
      "epoch": 0.6844380403458213,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960931323915665,
      "loss": 5.2352,
      "mean_token_accuracy": 0.21601954698562623,
      "num_tokens": 16335344.0,
      "step": 7125
    },
    {
      "entropy": 5.329084968566894,
      "epoch": 0.6849183477425552,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004960867687509475,
      "loss": 5.0828,
      "mean_token_accuracy": 0.21148615926504136,
      "num_tokens": 16349018.0,
      "step": 7130
    },
    {
      "entropy": 5.325907421112061,
      "epoch": 0.6853986551392891,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960803999773504,
      "loss": 5.1641,
      "mean_token_accuracy": 0.20542858988046647,
      "num_tokens": 16360137.0,
      "step": 7135
    },
    {
      "entropy": 5.354696941375733,
      "epoch": 0.6858789625360231,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960740260709228,
      "loss": 5.1637,
      "mean_token_accuracy": 0.21040215939283372,
      "num_tokens": 16372277.0,
      "step": 7140
    },
    {
      "entropy": 5.36764178276062,
      "epoch": 0.686359269932757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960676470318128,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20619126707315444,
      "num_tokens": 16383440.0,
      "step": 7145
    },
    {
      "entropy": 5.336434745788575,
      "epoch": 0.6868395773294909,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004960612628601683,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20799643397331238,
      "num_tokens": 16394330.0,
      "step": 7150
    },
    {
      "entropy": 5.259572553634643,
      "epoch": 0.6873198847262247,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960548735561379,
      "loss": 5.0537,
      "mean_token_accuracy": 0.21908190548419954,
      "num_tokens": 16405734.0,
      "step": 7155
    },
    {
      "entropy": 5.3112781047821045,
      "epoch": 0.6878001921229587,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960484791198697,
      "loss": 5.0736,
      "mean_token_accuracy": 0.20679262578487395,
      "num_tokens": 16416025.0,
      "step": 7160
    },
    {
      "entropy": 5.296441125869751,
      "epoch": 0.6882804995196926,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004960420795515121,
      "loss": 5.0933,
      "mean_token_accuracy": 0.21095090508460998,
      "num_tokens": 16427416.0,
      "step": 7165
    },
    {
      "entropy": 5.288483285903931,
      "epoch": 0.6887608069164265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960356748512138,
      "loss": 5.0884,
      "mean_token_accuracy": 0.213330939412117,
      "num_tokens": 16438073.0,
      "step": 7170
    },
    {
      "entropy": 5.276486349105835,
      "epoch": 0.6892411143131604,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004960292650191236,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2134612113237381,
      "num_tokens": 16449994.0,
      "step": 7175
    },
    {
      "entropy": 5.2832695007324215,
      "epoch": 0.6897214217098944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004960228500553899,
      "loss": 5.0432,
      "mean_token_accuracy": 0.22136877477169037,
      "num_tokens": 16460355.0,
      "step": 7180
    },
    {
      "entropy": 5.358346033096313,
      "epoch": 0.6902017291066282,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960164299601623,
      "loss": 5.1235,
      "mean_token_accuracy": 0.20528924614191055,
      "num_tokens": 16472274.0,
      "step": 7185
    },
    {
      "entropy": 5.3691483497619625,
      "epoch": 0.6906820365033621,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004960100047335892,
      "loss": 5.1447,
      "mean_token_accuracy": 0.21010650247335433,
      "num_tokens": 16482959.0,
      "step": 7190
    },
    {
      "entropy": 5.345746898651123,
      "epoch": 0.691162343900096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960035743758202,
      "loss": 5.1897,
      "mean_token_accuracy": 0.2034367188811302,
      "num_tokens": 16494533.0,
      "step": 7195
    },
    {
      "entropy": 5.212039041519165,
      "epoch": 0.69164265129683,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004959971388870044,
      "loss": 5.0475,
      "mean_token_accuracy": 0.21525924652814865,
      "num_tokens": 16506124.0,
      "step": 7200
    },
    {
      "entropy": 5.305228567123413,
      "epoch": 0.6921229586935639,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004959906982672912,
      "loss": 5.0656,
      "mean_token_accuracy": 0.2091907739639282,
      "num_tokens": 16517867.0,
      "step": 7205
    },
    {
      "entropy": 5.256267166137695,
      "epoch": 0.6926032660902978,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004959842525168302,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21432278901338578,
      "num_tokens": 16529075.0,
      "step": 7210
    },
    {
      "entropy": 5.317261743545532,
      "epoch": 0.6930835734870316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959778016357712,
      "loss": 5.0689,
      "mean_token_accuracy": 0.2152673825621605,
      "num_tokens": 16540326.0,
      "step": 7215
    },
    {
      "entropy": 5.318550968170166,
      "epoch": 0.6935638808837656,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959713456242637,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21104695349931718,
      "num_tokens": 16551570.0,
      "step": 7220
    },
    {
      "entropy": 5.385565185546875,
      "epoch": 0.6940441882804995,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004959648844824576,
      "loss": 5.1361,
      "mean_token_accuracy": 0.21007074266672135,
      "num_tokens": 16562636.0,
      "step": 7225
    },
    {
      "entropy": 5.319220018386841,
      "epoch": 0.6945244956772334,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959584182105032,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21816612184047698,
      "num_tokens": 16573867.0,
      "step": 7230
    },
    {
      "entropy": 5.206090497970581,
      "epoch": 0.6950048030739674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004959519468085504,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2207987681031227,
      "num_tokens": 16584533.0,
      "step": 7235
    },
    {
      "entropy": 5.289491128921509,
      "epoch": 0.6954851104707013,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959454702767494,
      "loss": 5.2381,
      "mean_token_accuracy": 0.2078647419810295,
      "num_tokens": 16596562.0,
      "step": 7240
    },
    {
      "entropy": 5.348395824432373,
      "epoch": 0.6959654178674352,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959389886152507,
      "loss": 5.0863,
      "mean_token_accuracy": 0.2130254775285721,
      "num_tokens": 16607508.0,
      "step": 7245
    },
    {
      "entropy": 5.350907850265503,
      "epoch": 0.696445725264169,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959325018242048,
      "loss": 5.1689,
      "mean_token_accuracy": 0.20744186639785767,
      "num_tokens": 16617737.0,
      "step": 7250
    },
    {
      "entropy": 5.369990825653076,
      "epoch": 0.696926032660903,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959260099037622,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20871317833662034,
      "num_tokens": 16628518.0,
      "step": 7255
    },
    {
      "entropy": 5.323066663742066,
      "epoch": 0.6974063400576369,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959195128540737,
      "loss": 5.1696,
      "mean_token_accuracy": 0.20603234469890594,
      "num_tokens": 16639644.0,
      "step": 7260
    },
    {
      "entropy": 5.429886150360107,
      "epoch": 0.6978866474543708,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049591301067529,
      "loss": 5.219,
      "mean_token_accuracy": 0.19961978644132614,
      "num_tokens": 16650882.0,
      "step": 7265
    },
    {
      "entropy": 5.371072101593017,
      "epoch": 0.6983669548511047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959065033675623,
      "loss": 5.1271,
      "mean_token_accuracy": 0.20525182038545609,
      "num_tokens": 16662464.0,
      "step": 7270
    },
    {
      "entropy": 5.361147451400757,
      "epoch": 0.6988472622478387,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958999909310414,
      "loss": 5.1558,
      "mean_token_accuracy": 0.20310764908790588,
      "num_tokens": 16673718.0,
      "step": 7275
    },
    {
      "entropy": 5.267883825302124,
      "epoch": 0.6993275696445725,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958934733658788,
      "loss": 5.094,
      "mean_token_accuracy": 0.20379672348499298,
      "num_tokens": 16684957.0,
      "step": 7280
    },
    {
      "entropy": 5.354166889190674,
      "epoch": 0.6998078770413064,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958869506722256,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20695476233959198,
      "num_tokens": 16695782.0,
      "step": 7285
    },
    {
      "entropy": 5.318038415908814,
      "epoch": 0.7002881844380403,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004958804228502332,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21745894253253936,
      "num_tokens": 16707448.0,
      "step": 7290
    },
    {
      "entropy": 5.206328392028809,
      "epoch": 0.7007684918347743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958738899000534,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20685766786336898,
      "num_tokens": 16718074.0,
      "step": 7295
    },
    {
      "entropy": 5.296359872817993,
      "epoch": 0.7012487992315082,
      "grad_norm": 1.125,
      "learning_rate": 0.0004958673518218377,
      "loss": 5.0889,
      "mean_token_accuracy": 0.2153885081410408,
      "num_tokens": 16728656.0,
      "step": 7300
    },
    {
      "entropy": 5.316540145874024,
      "epoch": 0.7017291066282421,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958608086157379,
      "loss": 5.157,
      "mean_token_accuracy": 0.2120751053094864,
      "num_tokens": 16740703.0,
      "step": 7305
    },
    {
      "entropy": 5.307763719558716,
      "epoch": 0.7022094140249759,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495854260281906,
      "loss": 5.0171,
      "mean_token_accuracy": 0.22452596575021744,
      "num_tokens": 16753320.0,
      "step": 7310
    },
    {
      "entropy": 5.209990787506103,
      "epoch": 0.7026897214217099,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004958477068204941,
      "loss": 5.096,
      "mean_token_accuracy": 0.2123555824160576,
      "num_tokens": 16764889.0,
      "step": 7315
    },
    {
      "entropy": 5.362256240844727,
      "epoch": 0.7031700288184438,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000495841148231654,
      "loss": 5.1306,
      "mean_token_accuracy": 0.21161444634199142,
      "num_tokens": 16776301.0,
      "step": 7320
    },
    {
      "entropy": 5.346973085403443,
      "epoch": 0.7036503362151777,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004958345845155383,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2196694180369377,
      "num_tokens": 16786935.0,
      "step": 7325
    },
    {
      "entropy": 5.265995645523072,
      "epoch": 0.7041306436119116,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958280156722992,
      "loss": 5.1155,
      "mean_token_accuracy": 0.20905097723007202,
      "num_tokens": 16799335.0,
      "step": 7330
    },
    {
      "entropy": 5.230437231063843,
      "epoch": 0.7046109510086456,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958214417020894,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20976387858390808,
      "num_tokens": 16811376.0,
      "step": 7335
    },
    {
      "entropy": 5.359433364868164,
      "epoch": 0.7050912584053795,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958148626050614,
      "loss": 5.1172,
      "mean_token_accuracy": 0.2076278418302536,
      "num_tokens": 16822859.0,
      "step": 7340
    },
    {
      "entropy": 5.313623666763306,
      "epoch": 0.7055715658021133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495808278381368,
      "loss": 5.0985,
      "mean_token_accuracy": 0.21256341636180878,
      "num_tokens": 16833831.0,
      "step": 7345
    },
    {
      "entropy": 5.299200201034546,
      "epoch": 0.7060518731988472,
      "grad_norm": 0.875,
      "learning_rate": 0.000495801689031162,
      "loss": 5.1175,
      "mean_token_accuracy": 0.21099364310503005,
      "num_tokens": 16846827.0,
      "step": 7350
    },
    {
      "entropy": 5.329308938980103,
      "epoch": 0.7065321805955812,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957950945545965,
      "loss": 5.1,
      "mean_token_accuracy": 0.2063808485865593,
      "num_tokens": 16858166.0,
      "step": 7355
    },
    {
      "entropy": 5.376455068588257,
      "epoch": 0.7070124879923151,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957884949518246,
      "loss": 5.234,
      "mean_token_accuracy": 0.19484158605337143,
      "num_tokens": 16870201.0,
      "step": 7360
    },
    {
      "entropy": 5.372378778457642,
      "epoch": 0.707492795389049,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004957818902229992,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21045285314321518,
      "num_tokens": 16880891.0,
      "step": 7365
    },
    {
      "entropy": 5.342394495010376,
      "epoch": 0.7079731027857828,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957752803682741,
      "loss": 5.1578,
      "mean_token_accuracy": 0.20331737101078035,
      "num_tokens": 16893498.0,
      "step": 7370
    },
    {
      "entropy": 5.355802011489868,
      "epoch": 0.7084534101825168,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004957686653878024,
      "loss": 5.127,
      "mean_token_accuracy": 0.2090783879160881,
      "num_tokens": 16904959.0,
      "step": 7375
    },
    {
      "entropy": 5.3680956840515135,
      "epoch": 0.7089337175792507,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000495762045281738,
      "loss": 5.1641,
      "mean_token_accuracy": 0.20913613587617874,
      "num_tokens": 16916547.0,
      "step": 7380
    },
    {
      "entropy": 5.297506475448609,
      "epoch": 0.7094140249759846,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004957554200502344,
      "loss": 5.0991,
      "mean_token_accuracy": 0.21234960705041886,
      "num_tokens": 16928580.0,
      "step": 7385
    },
    {
      "entropy": 5.319205904006958,
      "epoch": 0.7098943323727186,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004957487896934454,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20790536105632781,
      "num_tokens": 16941247.0,
      "step": 7390
    },
    {
      "entropy": 5.383007049560547,
      "epoch": 0.7103746397694525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000495742154211525,
      "loss": 5.1568,
      "mean_token_accuracy": 0.2079955071210861,
      "num_tokens": 16951119.0,
      "step": 7395
    },
    {
      "entropy": 5.295486354827881,
      "epoch": 0.7108549471661864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957355136046272,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21048038601875305,
      "num_tokens": 16962608.0,
      "step": 7400
    },
    {
      "entropy": 5.264986562728882,
      "epoch": 0.7113352545629202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957288678729064,
      "loss": 5.1119,
      "mean_token_accuracy": 0.21038804352283477,
      "num_tokens": 16973291.0,
      "step": 7405
    },
    {
      "entropy": 5.36090874671936,
      "epoch": 0.7118155619596542,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957222170165166,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20317897647619249,
      "num_tokens": 16984895.0,
      "step": 7410
    },
    {
      "entropy": 5.349123573303222,
      "epoch": 0.7122958693563881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957155610356124,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20322143286466599,
      "num_tokens": 16997109.0,
      "step": 7415
    },
    {
      "entropy": 5.303277158737183,
      "epoch": 0.712776176753122,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957088999303481,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2074026107788086,
      "num_tokens": 17008640.0,
      "step": 7420
    },
    {
      "entropy": 5.281453371047974,
      "epoch": 0.7132564841498559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957022337008787,
      "loss": 5.0543,
      "mean_token_accuracy": 0.21270692795515062,
      "num_tokens": 17018924.0,
      "step": 7425
    },
    {
      "entropy": 5.27117953300476,
      "epoch": 0.7137367915465899,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004956955623473587,
      "loss": 5.0808,
      "mean_token_accuracy": 0.2093396544456482,
      "num_tokens": 17029932.0,
      "step": 7430
    },
    {
      "entropy": 5.348640775680542,
      "epoch": 0.7142170989433237,
      "grad_norm": 1.046875,
      "learning_rate": 0.000495688885869943,
      "loss": 5.1856,
      "mean_token_accuracy": 0.20419342815876007,
      "num_tokens": 17042164.0,
      "step": 7435
    },
    {
      "entropy": 5.344086837768555,
      "epoch": 0.7146974063400576,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004956822042687868,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19928455352783203,
      "num_tokens": 17054729.0,
      "step": 7440
    },
    {
      "entropy": 5.251951551437378,
      "epoch": 0.7151777137367915,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956755175440451,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2236065834760666,
      "num_tokens": 17066537.0,
      "step": 7445
    },
    {
      "entropy": 5.29123477935791,
      "epoch": 0.7156580211335255,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956688256958731,
      "loss": 5.0822,
      "mean_token_accuracy": 0.21617083251476288,
      "num_tokens": 17077787.0,
      "step": 7450
    },
    {
      "entropy": 5.267868709564209,
      "epoch": 0.7161383285302594,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004956621287244262,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20687348544597625,
      "num_tokens": 17089555.0,
      "step": 7455
    },
    {
      "entropy": 5.373935079574585,
      "epoch": 0.7166186359269933,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956554266298599,
      "loss": 5.0856,
      "mean_token_accuracy": 0.21658468544483184,
      "num_tokens": 17099911.0,
      "step": 7460
    },
    {
      "entropy": 5.261654949188232,
      "epoch": 0.7170989433237271,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956487194123298,
      "loss": 5.0611,
      "mean_token_accuracy": 0.21106222569942473,
      "num_tokens": 17110880.0,
      "step": 7465
    },
    {
      "entropy": 5.261746025085449,
      "epoch": 0.7175792507204611,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004956420070719918,
      "loss": 5.0196,
      "mean_token_accuracy": 0.21547937095165254,
      "num_tokens": 17122272.0,
      "step": 7470
    },
    {
      "entropy": 5.31106014251709,
      "epoch": 0.718059558117195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956352896090014,
      "loss": 5.1652,
      "mean_token_accuracy": 0.20728080123662948,
      "num_tokens": 17133995.0,
      "step": 7475
    },
    {
      "entropy": 5.331283950805664,
      "epoch": 0.7185398655139289,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956285670235147,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21416814774274825,
      "num_tokens": 17145970.0,
      "step": 7480
    },
    {
      "entropy": 5.3921637535095215,
      "epoch": 0.7190201729106628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004956218393156879,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19899100363254546,
      "num_tokens": 17157747.0,
      "step": 7485
    },
    {
      "entropy": 5.292384815216065,
      "epoch": 0.7195004803073968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956151064856772,
      "loss": 5.1784,
      "mean_token_accuracy": 0.21398458033800125,
      "num_tokens": 17168357.0,
      "step": 7490
    },
    {
      "entropy": 5.332787752151489,
      "epoch": 0.7199807877041307,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004956083685336386,
      "loss": 5.1195,
      "mean_token_accuracy": 0.2166551887989044,
      "num_tokens": 17179871.0,
      "step": 7495
    },
    {
      "entropy": 5.333356714248657,
      "epoch": 0.7204610951008645,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956016254597289,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21713766753673552,
      "num_tokens": 17190456.0,
      "step": 7500
    },
    {
      "entropy": 5.292677927017212,
      "epoch": 0.7209414024975984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955948772641044,
      "loss": 5.1586,
      "mean_token_accuracy": 0.20351035445928572,
      "num_tokens": 17201623.0,
      "step": 7505
    },
    {
      "entropy": 5.273524570465088,
      "epoch": 0.7214217098943324,
      "grad_norm": 1.0,
      "learning_rate": 0.000495588123946922,
      "loss": 5.0593,
      "mean_token_accuracy": 0.21409086734056473,
      "num_tokens": 17212891.0,
      "step": 7510
    },
    {
      "entropy": 5.306383371353149,
      "epoch": 0.7219020172910663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955813655083384,
      "loss": 5.1116,
      "mean_token_accuracy": 0.21043863296508789,
      "num_tokens": 17223983.0,
      "step": 7515
    },
    {
      "entropy": 5.348400354385376,
      "epoch": 0.7223823246878002,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004955746019485104,
      "loss": 5.1401,
      "mean_token_accuracy": 0.21320433765649796,
      "num_tokens": 17233701.0,
      "step": 7520
    },
    {
      "entropy": 5.291280460357666,
      "epoch": 0.722862632084534,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000495567833267595,
      "loss": 5.1367,
      "mean_token_accuracy": 0.2105158671736717,
      "num_tokens": 17245958.0,
      "step": 7525
    },
    {
      "entropy": 5.397392225265503,
      "epoch": 0.723342939481268,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955610594657496,
      "loss": 5.1429,
      "mean_token_accuracy": 0.1996593102812767,
      "num_tokens": 17257687.0,
      "step": 7530
    },
    {
      "entropy": 5.378547048568725,
      "epoch": 0.7238232468780019,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955542805431313,
      "loss": 5.1959,
      "mean_token_accuracy": 0.20859490633010863,
      "num_tokens": 17270247.0,
      "step": 7535
    },
    {
      "entropy": 5.287229585647583,
      "epoch": 0.7243035542747358,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004955474964998976,
      "loss": 5.109,
      "mean_token_accuracy": 0.21423844248056412,
      "num_tokens": 17281920.0,
      "step": 7540
    },
    {
      "entropy": 5.405027866363525,
      "epoch": 0.7247838616714697,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004955407073362058,
      "loss": 5.2695,
      "mean_token_accuracy": 0.19586165845394135,
      "num_tokens": 17293602.0,
      "step": 7545
    },
    {
      "entropy": 5.391187238693237,
      "epoch": 0.7252641690682037,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955339130522136,
      "loss": 5.056,
      "mean_token_accuracy": 0.2137158066034317,
      "num_tokens": 17304484.0,
      "step": 7550
    },
    {
      "entropy": 5.19770622253418,
      "epoch": 0.7257444764649376,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495527113648079,
      "loss": 5.0034,
      "mean_token_accuracy": 0.21465377360582352,
      "num_tokens": 17314953.0,
      "step": 7555
    },
    {
      "entropy": 5.306738376617432,
      "epoch": 0.7262247838616714,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955203091239596,
      "loss": 5.13,
      "mean_token_accuracy": 0.2061687335371971,
      "num_tokens": 17326813.0,
      "step": 7560
    },
    {
      "entropy": 5.413270282745361,
      "epoch": 0.7267050912584054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004955134994800134,
      "loss": 5.195,
      "mean_token_accuracy": 0.2014137849211693,
      "num_tokens": 17337968.0,
      "step": 7565
    },
    {
      "entropy": 5.325583982467651,
      "epoch": 0.7271853986551393,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004955066847163984,
      "loss": 5.0992,
      "mean_token_accuracy": 0.21064977049827577,
      "num_tokens": 17350406.0,
      "step": 7570
    },
    {
      "entropy": 5.234203243255616,
      "epoch": 0.7276657060518732,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954998648332731,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21830987334251403,
      "num_tokens": 17361888.0,
      "step": 7575
    },
    {
      "entropy": 5.269831943511963,
      "epoch": 0.7281460134486071,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954930398307956,
      "loss": 5.1326,
      "mean_token_accuracy": 0.2094791740179062,
      "num_tokens": 17374047.0,
      "step": 7580
    },
    {
      "entropy": 5.331007862091065,
      "epoch": 0.7286263208453411,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004954862097091245,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20073632299900054,
      "num_tokens": 17386175.0,
      "step": 7585
    },
    {
      "entropy": 5.310300588607788,
      "epoch": 0.729106628242075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954793744684184,
      "loss": 5.1267,
      "mean_token_accuracy": 0.22059799134731292,
      "num_tokens": 17398168.0,
      "step": 7590
    },
    {
      "entropy": 5.2579484462738035,
      "epoch": 0.7295869356388088,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954725341088358,
      "loss": 5.0158,
      "mean_token_accuracy": 0.21232549101114273,
      "num_tokens": 17408825.0,
      "step": 7595
    },
    {
      "entropy": 5.305180788040161,
      "epoch": 0.7300672430355427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954656886305356,
      "loss": 5.1106,
      "mean_token_accuracy": 0.2107972264289856,
      "num_tokens": 17419813.0,
      "step": 7600
    },
    {
      "entropy": 5.37008376121521,
      "epoch": 0.7305475504322767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954588380336768,
      "loss": 5.1518,
      "mean_token_accuracy": 0.20207465142011644,
      "num_tokens": 17431134.0,
      "step": 7605
    },
    {
      "entropy": 5.261743974685669,
      "epoch": 0.7310278578290106,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004954519823184184,
      "loss": 5.0647,
      "mean_token_accuracy": 0.21424833685159683,
      "num_tokens": 17442614.0,
      "step": 7610
    },
    {
      "entropy": 5.327200174331665,
      "epoch": 0.7315081652257445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954451214849196,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21629141122102738,
      "num_tokens": 17454615.0,
      "step": 7615
    },
    {
      "entropy": 5.336834573745728,
      "epoch": 0.7319884726224783,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954382555333397,
      "loss": 5.1111,
      "mean_token_accuracy": 0.20628802478313446,
      "num_tokens": 17467025.0,
      "step": 7620
    },
    {
      "entropy": 5.264048147201538,
      "epoch": 0.7324687800192123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495431384463838,
      "loss": 5.0859,
      "mean_token_accuracy": 0.21480168998241425,
      "num_tokens": 17477681.0,
      "step": 7625
    },
    {
      "entropy": 5.285310077667236,
      "epoch": 0.7329490874159462,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004954245082765741,
      "loss": 5.1157,
      "mean_token_accuracy": 0.21148568838834764,
      "num_tokens": 17489814.0,
      "step": 7630
    },
    {
      "entropy": 5.346252346038819,
      "epoch": 0.7334293948126801,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954176269717077,
      "loss": 5.1202,
      "mean_token_accuracy": 0.2132637083530426,
      "num_tokens": 17501701.0,
      "step": 7635
    },
    {
      "entropy": 5.3037293434143065,
      "epoch": 0.733909702209414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954107405493984,
      "loss": 5.0333,
      "mean_token_accuracy": 0.214559967815876,
      "num_tokens": 17513585.0,
      "step": 7640
    },
    {
      "entropy": 5.3021446704864506,
      "epoch": 0.734390009606148,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954038490098064,
      "loss": 5.1618,
      "mean_token_accuracy": 0.2128313183784485,
      "num_tokens": 17525557.0,
      "step": 7645
    },
    {
      "entropy": 5.2387083053588865,
      "epoch": 0.7348703170028819,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004953969523530914,
      "loss": 5.0275,
      "mean_token_accuracy": 0.20948038697242738,
      "num_tokens": 17538312.0,
      "step": 7650
    },
    {
      "entropy": 5.418973731994629,
      "epoch": 0.7353506243996157,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953900505794136,
      "loss": 5.162,
      "mean_token_accuracy": 0.20908292829990388,
      "num_tokens": 17550248.0,
      "step": 7655
    },
    {
      "entropy": 5.292871332168579,
      "epoch": 0.7358309317963496,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953831436889332,
      "loss": 5.0346,
      "mean_token_accuracy": 0.20809418261051177,
      "num_tokens": 17560624.0,
      "step": 7660
    },
    {
      "entropy": 5.309444332122803,
      "epoch": 0.7363112391930836,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953762316818106,
      "loss": 5.0887,
      "mean_token_accuracy": 0.20811459124088288,
      "num_tokens": 17572439.0,
      "step": 7665
    },
    {
      "entropy": 5.343560075759887,
      "epoch": 0.7367915465898175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953693145582064,
      "loss": 5.1339,
      "mean_token_accuracy": 0.20418035835027695,
      "num_tokens": 17583120.0,
      "step": 7670
    },
    {
      "entropy": 5.212920808792115,
      "epoch": 0.7372718539865514,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000495362392318281,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2187417581677437,
      "num_tokens": 17593677.0,
      "step": 7675
    },
    {
      "entropy": 5.2843059539794925,
      "epoch": 0.7377521613832853,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004953554649621951,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20589945167303086,
      "num_tokens": 17605180.0,
      "step": 7680
    },
    {
      "entropy": 5.331402111053467,
      "epoch": 0.7382324687800192,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004953485324901098,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21452029049396515,
      "num_tokens": 17617459.0,
      "step": 7685
    },
    {
      "entropy": 5.290313196182251,
      "epoch": 0.7387127761767531,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953415949021857,
      "loss": 5.0781,
      "mean_token_accuracy": 0.2178971976041794,
      "num_tokens": 17628024.0,
      "step": 7690
    },
    {
      "entropy": 5.239890813827515,
      "epoch": 0.739193083573487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953346521985843,
      "loss": 5.088,
      "mean_token_accuracy": 0.20925626009702683,
      "num_tokens": 17639833.0,
      "step": 7695
    },
    {
      "entropy": 5.369693613052368,
      "epoch": 0.7396733909702209,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953277043794663,
      "loss": 5.1818,
      "mean_token_accuracy": 0.1988198459148407,
      "num_tokens": 17651057.0,
      "step": 7700
    },
    {
      "entropy": 5.329780864715576,
      "epoch": 0.7401536983669549,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004953207514449933,
      "loss": 5.0765,
      "mean_token_accuracy": 0.21758732348680496,
      "num_tokens": 17662288.0,
      "step": 7705
    },
    {
      "entropy": 5.29347710609436,
      "epoch": 0.7406340057636888,
      "grad_norm": 0.875,
      "learning_rate": 0.0004953137933953267,
      "loss": 5.081,
      "mean_token_accuracy": 0.21002478748559952,
      "num_tokens": 17673885.0,
      "step": 7710
    },
    {
      "entropy": 5.31736159324646,
      "epoch": 0.7411143131604226,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495306830230628,
      "loss": 5.1114,
      "mean_token_accuracy": 0.21368523836135864,
      "num_tokens": 17685030.0,
      "step": 7715
    },
    {
      "entropy": 5.289226102828979,
      "epoch": 0.7415946205571565,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952998619510589,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2157010018825531,
      "num_tokens": 17696624.0,
      "step": 7720
    },
    {
      "entropy": 5.273167467117309,
      "epoch": 0.7420749279538905,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952928885567811,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21482881605625154,
      "num_tokens": 17707386.0,
      "step": 7725
    },
    {
      "entropy": 5.425893783569336,
      "epoch": 0.7425552353506244,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952859100479566,
      "loss": 5.2825,
      "mean_token_accuracy": 0.20461898297071457,
      "num_tokens": 17718605.0,
      "step": 7730
    },
    {
      "entropy": 5.3197441577911375,
      "epoch": 0.7430355427473583,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952789264247474,
      "loss": 5.0686,
      "mean_token_accuracy": 0.2129740059375763,
      "num_tokens": 17730275.0,
      "step": 7735
    },
    {
      "entropy": 5.359417963027954,
      "epoch": 0.7435158501440923,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952719376873156,
      "loss": 5.1666,
      "mean_token_accuracy": 0.20866380780935287,
      "num_tokens": 17741390.0,
      "step": 7740
    },
    {
      "entropy": 5.301401042938233,
      "epoch": 0.7439961575408262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004952649438358234,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21666279733181,
      "num_tokens": 17752354.0,
      "step": 7745
    },
    {
      "entropy": 5.350618362426758,
      "epoch": 0.74447646493756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952579448704334,
      "loss": 5.0566,
      "mean_token_accuracy": 0.21065741777420044,
      "num_tokens": 17762839.0,
      "step": 7750
    },
    {
      "entropy": 5.312939834594727,
      "epoch": 0.7449567723342939,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000495250940791308,
      "loss": 5.1569,
      "mean_token_accuracy": 0.20553001314401625,
      "num_tokens": 17775800.0,
      "step": 7755
    },
    {
      "entropy": 5.317015075683594,
      "epoch": 0.7454370797310279,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004952439315986096,
      "loss": 5.0771,
      "mean_token_accuracy": 0.21331654787063598,
      "num_tokens": 17787804.0,
      "step": 7760
    },
    {
      "entropy": 5.355747652053833,
      "epoch": 0.7459173871277618,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952369172925012,
      "loss": 5.1738,
      "mean_token_accuracy": 0.21383114010095597,
      "num_tokens": 17800291.0,
      "step": 7765
    },
    {
      "entropy": 5.299292850494385,
      "epoch": 0.7463976945244957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004952298978731454,
      "loss": 5.0425,
      "mean_token_accuracy": 0.2166093558073044,
      "num_tokens": 17811548.0,
      "step": 7770
    },
    {
      "entropy": 5.3169831275939945,
      "epoch": 0.7468780019212296,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952228733407055,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22119933664798735,
      "num_tokens": 17822589.0,
      "step": 7775
    },
    {
      "entropy": 5.22701244354248,
      "epoch": 0.7473583093179635,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952158436953444,
      "loss": 5.0147,
      "mean_token_accuracy": 0.21899669468402863,
      "num_tokens": 17834203.0,
      "step": 7780
    },
    {
      "entropy": 5.295480871200562,
      "epoch": 0.7478386167146974,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952088089372252,
      "loss": 5.1013,
      "mean_token_accuracy": 0.20918091088533403,
      "num_tokens": 17846238.0,
      "step": 7785
    },
    {
      "entropy": 5.289197826385498,
      "epoch": 0.7483189241114313,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952017690665114,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21453914940357208,
      "num_tokens": 17857640.0,
      "step": 7790
    },
    {
      "entropy": 5.2739338874816895,
      "epoch": 0.7487992315081652,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004951947240833664,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21228243559598922,
      "num_tokens": 17868755.0,
      "step": 7795
    },
    {
      "entropy": 5.256406688690186,
      "epoch": 0.7492795389048992,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951876739879537,
      "loss": 5.1,
      "mean_token_accuracy": 0.20885979682207106,
      "num_tokens": 17881078.0,
      "step": 7800
    },
    {
      "entropy": 5.248721313476563,
      "epoch": 0.7497598463016331,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951806187804371,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2142539069056511,
      "num_tokens": 17893888.0,
      "step": 7805
    },
    {
      "entropy": 5.3160608291625975,
      "epoch": 0.7502401536983669,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951735584609804,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20865219682455063,
      "num_tokens": 17904443.0,
      "step": 7810
    },
    {
      "entropy": 5.25682487487793,
      "epoch": 0.7507204610951008,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951664930297474,
      "loss": 5.1071,
      "mean_token_accuracy": 0.21454736590385437,
      "num_tokens": 17918090.0,
      "step": 7815
    },
    {
      "entropy": 5.233595657348633,
      "epoch": 0.7512007684918348,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495159422486902,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2176428869366646,
      "num_tokens": 17929233.0,
      "step": 7820
    },
    {
      "entropy": 5.296094274520874,
      "epoch": 0.7516810758885687,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004951523468326088,
      "loss": 5.076,
      "mean_token_accuracy": 0.209262515604496,
      "num_tokens": 17940580.0,
      "step": 7825
    },
    {
      "entropy": 5.33233232498169,
      "epoch": 0.7521613832853026,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004951452660670317,
      "loss": 5.1637,
      "mean_token_accuracy": 0.20420581698417664,
      "num_tokens": 17953993.0,
      "step": 7830
    },
    {
      "entropy": 5.270310974121093,
      "epoch": 0.7526416906820365,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004951381801903352,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21062700897455217,
      "num_tokens": 17966033.0,
      "step": 7835
    },
    {
      "entropy": 5.194505786895752,
      "epoch": 0.7531219980787704,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004951310892026839,
      "loss": 4.9804,
      "mean_token_accuracy": 0.222573122382164,
      "num_tokens": 17977943.0,
      "step": 7840
    },
    {
      "entropy": 5.292294883728028,
      "epoch": 0.7536023054755043,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004951239931042424,
      "loss": 5.0517,
      "mean_token_accuracy": 0.2184312343597412,
      "num_tokens": 17990135.0,
      "step": 7845
    },
    {
      "entropy": 5.232913684844971,
      "epoch": 0.7540826128722382,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951168918951753,
      "loss": 5.0832,
      "mean_token_accuracy": 0.21639539897441865,
      "num_tokens": 18002126.0,
      "step": 7850
    },
    {
      "entropy": 5.200091791152954,
      "epoch": 0.7545629202689721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004951097855756476,
      "loss": 4.9794,
      "mean_token_accuracy": 0.22055791765451432,
      "num_tokens": 18013147.0,
      "step": 7855
    },
    {
      "entropy": 5.337146377563476,
      "epoch": 0.7550432276657061,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004951026741458243,
      "loss": 5.0909,
      "mean_token_accuracy": 0.215498249232769,
      "num_tokens": 18025146.0,
      "step": 7860
    },
    {
      "entropy": 5.257822465896607,
      "epoch": 0.75552353506244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950955576058705,
      "loss": 5.032,
      "mean_token_accuracy": 0.21024022549390792,
      "num_tokens": 18036412.0,
      "step": 7865
    },
    {
      "entropy": 5.274294996261597,
      "epoch": 0.7560038424591738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950884359559513,
      "loss": 5.1129,
      "mean_token_accuracy": 0.21416025012731552,
      "num_tokens": 18048041.0,
      "step": 7870
    },
    {
      "entropy": 5.3436744689941404,
      "epoch": 0.7564841498559077,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004950813091962324,
      "loss": 5.1536,
      "mean_token_accuracy": 0.2061038300395012,
      "num_tokens": 18060163.0,
      "step": 7875
    },
    {
      "entropy": 5.331613254547119,
      "epoch": 0.7569644572526417,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004950741773268788,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20520974397659303,
      "num_tokens": 18071628.0,
      "step": 7880
    },
    {
      "entropy": 5.216393280029297,
      "epoch": 0.7574447646493756,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950670403480562,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21191180646419525,
      "num_tokens": 18082979.0,
      "step": 7885
    },
    {
      "entropy": 5.271268558502197,
      "epoch": 0.7579250720461095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950598982599306,
      "loss": 5.0618,
      "mean_token_accuracy": 0.21403386443853378,
      "num_tokens": 18093889.0,
      "step": 7890
    },
    {
      "entropy": 5.256098937988281,
      "epoch": 0.7584053794428435,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004950527510626675,
      "loss": 5.0995,
      "mean_token_accuracy": 0.2129449024796486,
      "num_tokens": 18105559.0,
      "step": 7895
    },
    {
      "entropy": 5.318772459030152,
      "epoch": 0.7588856868395774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950455987564329,
      "loss": 5.1806,
      "mean_token_accuracy": 0.20708672255277633,
      "num_tokens": 18116316.0,
      "step": 7900
    },
    {
      "entropy": 5.3347306728363035,
      "epoch": 0.7593659942363112,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004950384413413931,
      "loss": 5.0708,
      "mean_token_accuracy": 0.21373462826013565,
      "num_tokens": 18126851.0,
      "step": 7905
    },
    {
      "entropy": 5.3427056789398195,
      "epoch": 0.7598463016330451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950312788177139,
      "loss": 5.1577,
      "mean_token_accuracy": 0.20313034504652022,
      "num_tokens": 18139571.0,
      "step": 7910
    },
    {
      "entropy": 5.376320457458496,
      "epoch": 0.7603266090297791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950241111855618,
      "loss": 5.1459,
      "mean_token_accuracy": 0.20792322903871535,
      "num_tokens": 18150680.0,
      "step": 7915
    },
    {
      "entropy": 5.326610422134399,
      "epoch": 0.760806916426513,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004950169384451031,
      "loss": 5.0837,
      "mean_token_accuracy": 0.21604117900133132,
      "num_tokens": 18161911.0,
      "step": 7920
    },
    {
      "entropy": 5.296671390533447,
      "epoch": 0.7612872238232469,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950097605965045,
      "loss": 5.078,
      "mean_token_accuracy": 0.2198405459523201,
      "num_tokens": 18172714.0,
      "step": 7925
    },
    {
      "entropy": 5.313606548309326,
      "epoch": 0.7617675312199808,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950025776399326,
      "loss": 5.0674,
      "mean_token_accuracy": 0.21653930395841597,
      "num_tokens": 18184250.0,
      "step": 7930
    },
    {
      "entropy": 5.385619831085205,
      "epoch": 0.7622478386167147,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494995389575554,
      "loss": 5.1912,
      "mean_token_accuracy": 0.2073808193206787,
      "num_tokens": 18195299.0,
      "step": 7935
    },
    {
      "entropy": 5.40221004486084,
      "epoch": 0.7627281460134486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949881964035357,
      "loss": 5.2403,
      "mean_token_accuracy": 0.20646310448646546,
      "num_tokens": 18206863.0,
      "step": 7940
    },
    {
      "entropy": 5.363748502731323,
      "epoch": 0.7632084534101825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004949809981240448,
      "loss": 5.1684,
      "mean_token_accuracy": 0.20585118383169174,
      "num_tokens": 18219643.0,
      "step": 7945
    },
    {
      "entropy": 5.275132751464843,
      "epoch": 0.7636887608069164,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949737947372483,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21818018108606338,
      "num_tokens": 18230461.0,
      "step": 7950
    },
    {
      "entropy": 5.316145420074463,
      "epoch": 0.7641690682036504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004949665862433134,
      "loss": 5.1546,
      "mean_token_accuracy": 0.20263040512800218,
      "num_tokens": 18243768.0,
      "step": 7955
    },
    {
      "entropy": 5.322182703018188,
      "epoch": 0.7646493756003843,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949593726424077,
      "loss": 5.0557,
      "mean_token_accuracy": 0.2150597706437111,
      "num_tokens": 18255322.0,
      "step": 7960
    },
    {
      "entropy": 5.315861701965332,
      "epoch": 0.7651296829971181,
      "grad_norm": 1.0,
      "learning_rate": 0.0004949521539346985,
      "loss": 5.0966,
      "mean_token_accuracy": 0.21947134137153626,
      "num_tokens": 18267262.0,
      "step": 7965
    },
    {
      "entropy": 5.276798105239868,
      "epoch": 0.765609990393852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004949449301203533,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21508977562189102,
      "num_tokens": 18277332.0,
      "step": 7970
    },
    {
      "entropy": 5.286798334121704,
      "epoch": 0.766090297790586,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949377011995399,
      "loss": 5.1305,
      "mean_token_accuracy": 0.2131699487566948,
      "num_tokens": 18289547.0,
      "step": 7975
    },
    {
      "entropy": 5.307200813293457,
      "epoch": 0.7665706051873199,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949304671724263,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21467551290988923,
      "num_tokens": 18302097.0,
      "step": 7980
    },
    {
      "entropy": 5.323423957824707,
      "epoch": 0.7670509125840538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004949232280391802,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21219189018011092,
      "num_tokens": 18314151.0,
      "step": 7985
    },
    {
      "entropy": 5.2811822414398195,
      "epoch": 0.7675312199807877,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949159837999698,
      "loss": 5.0631,
      "mean_token_accuracy": 0.20877229869365693,
      "num_tokens": 18326085.0,
      "step": 7990
    },
    {
      "entropy": 5.306690454483032,
      "epoch": 0.7680115273775217,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949087344549633,
      "loss": 5.042,
      "mean_token_accuracy": 0.2084398701786995,
      "num_tokens": 18338158.0,
      "step": 7995
    },
    {
      "entropy": 5.289818334579468,
      "epoch": 0.7684918347742555,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000494901480004329,
      "loss": 5.0515,
      "mean_token_accuracy": 0.21258990615606307,
      "num_tokens": 18349418.0,
      "step": 8000
    },
    {
      "entropy": 5.309824275970459,
      "epoch": 0.7689721421709894,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004948942204482351,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21151256561279297,
      "num_tokens": 18361964.0,
      "step": 8005
    },
    {
      "entropy": 5.252948999404907,
      "epoch": 0.7694524495677233,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948869557868506,
      "loss": 4.9762,
      "mean_token_accuracy": 0.22491969764232636,
      "num_tokens": 18373108.0,
      "step": 8010
    },
    {
      "entropy": 5.176406383514404,
      "epoch": 0.7699327569644573,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948796860203439,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21920865774154663,
      "num_tokens": 18385310.0,
      "step": 8015
    },
    {
      "entropy": 5.328154754638672,
      "epoch": 0.7704130643611912,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948724111488838,
      "loss": 5.1854,
      "mean_token_accuracy": 0.20643505156040193,
      "num_tokens": 18396132.0,
      "step": 8020
    },
    {
      "entropy": 5.468338966369629,
      "epoch": 0.770893371757925,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004948651311726391,
      "loss": 5.2852,
      "mean_token_accuracy": 0.20121830999851226,
      "num_tokens": 18409930.0,
      "step": 8025
    },
    {
      "entropy": 5.381359672546386,
      "epoch": 0.7713736791546589,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004948578460917789,
      "loss": 5.158,
      "mean_token_accuracy": 0.20594588071107864,
      "num_tokens": 18421204.0,
      "step": 8030
    },
    {
      "entropy": 5.324843168258667,
      "epoch": 0.7718539865513929,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948505559064725,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21968425065279007,
      "num_tokens": 18433194.0,
      "step": 8035
    },
    {
      "entropy": 5.23973913192749,
      "epoch": 0.7723342939481268,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004948432606168889,
      "loss": 5.0226,
      "mean_token_accuracy": 0.21483880132436753,
      "num_tokens": 18445282.0,
      "step": 8040
    },
    {
      "entropy": 5.287118530273437,
      "epoch": 0.7728146013448607,
      "grad_norm": 1.0,
      "learning_rate": 0.0004948359602231976,
      "loss": 5.0515,
      "mean_token_accuracy": 0.2178490787744522,
      "num_tokens": 18456264.0,
      "step": 8045
    },
    {
      "entropy": 5.248097944259643,
      "epoch": 0.7732949087415946,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948286547255681,
      "loss": 4.9906,
      "mean_token_accuracy": 0.21612643152475358,
      "num_tokens": 18467573.0,
      "step": 8050
    },
    {
      "entropy": 5.284157228469849,
      "epoch": 0.7737752161383286,
      "grad_norm": 1.0,
      "learning_rate": 0.00049482134412417,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22036667317152023,
      "num_tokens": 18478333.0,
      "step": 8055
    },
    {
      "entropy": 5.258845043182373,
      "epoch": 0.7742555235350624,
      "grad_norm": 0.96875,
      "learning_rate": 0.000494814028419173,
      "loss": 4.9873,
      "mean_token_accuracy": 0.2183872029185295,
      "num_tokens": 18489106.0,
      "step": 8060
    },
    {
      "entropy": 5.31499285697937,
      "epoch": 0.7747358309317963,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494806707610747,
      "loss": 5.1268,
      "mean_token_accuracy": 0.21177464425563813,
      "num_tokens": 18500090.0,
      "step": 8065
    },
    {
      "entropy": 5.348161792755127,
      "epoch": 0.7752161383285303,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494799381699062,
      "loss": 5.0982,
      "mean_token_accuracy": 0.21690022945404053,
      "num_tokens": 18510305.0,
      "step": 8070
    },
    {
      "entropy": 5.342441606521606,
      "epoch": 0.7756964457252642,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947920506842879,
      "loss": 5.0927,
      "mean_token_accuracy": 0.2126712679862976,
      "num_tokens": 18523048.0,
      "step": 8075
    },
    {
      "entropy": 5.220316600799561,
      "epoch": 0.7761767531219981,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947847145665951,
      "loss": 5.0074,
      "mean_token_accuracy": 0.22062323242425919,
      "num_tokens": 18534145.0,
      "step": 8080
    },
    {
      "entropy": 5.297750473022461,
      "epoch": 0.776657060518732,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947773733461539,
      "loss": 5.054,
      "mean_token_accuracy": 0.21516508013010024,
      "num_tokens": 18545045.0,
      "step": 8085
    },
    {
      "entropy": 5.289766788482666,
      "epoch": 0.777137367915466,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947700270231347,
      "loss": 5.0534,
      "mean_token_accuracy": 0.21543944478034974,
      "num_tokens": 18557531.0,
      "step": 8090
    },
    {
      "entropy": 5.29792013168335,
      "epoch": 0.7776176753121998,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004947626755977079,
      "loss": 5.0852,
      "mean_token_accuracy": 0.211107274889946,
      "num_tokens": 18569127.0,
      "step": 8095
    },
    {
      "entropy": 5.334991550445556,
      "epoch": 0.7780979827089337,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004947553190700444,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2164621651172638,
      "num_tokens": 18580606.0,
      "step": 8100
    },
    {
      "entropy": 5.38695068359375,
      "epoch": 0.7785782901056676,
      "grad_norm": 0.921875,
      "learning_rate": 0.000494747957440315,
      "loss": 5.2227,
      "mean_token_accuracy": 0.20214432328939438,
      "num_tokens": 18592330.0,
      "step": 8105
    },
    {
      "entropy": 5.366260242462158,
      "epoch": 0.7790585975024016,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947405907086905,
      "loss": 5.1969,
      "mean_token_accuracy": 0.21018484532833098,
      "num_tokens": 18604721.0,
      "step": 8110
    },
    {
      "entropy": 5.311359643936157,
      "epoch": 0.7795389048991355,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004947332188753419,
      "loss": 5.0674,
      "mean_token_accuracy": 0.21536423712968827,
      "num_tokens": 18616814.0,
      "step": 8115
    },
    {
      "entropy": 5.3130049228668215,
      "epoch": 0.7800192122958693,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947258419404405,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21512551456689835,
      "num_tokens": 18628224.0,
      "step": 8120
    },
    {
      "entropy": 5.322597169876099,
      "epoch": 0.7804995196926032,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947184599041576,
      "loss": 5.113,
      "mean_token_accuracy": 0.2130896508693695,
      "num_tokens": 18639777.0,
      "step": 8125
    },
    {
      "entropy": 5.247894668579102,
      "epoch": 0.7809798270893372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947110727666644,
      "loss": 5.0393,
      "mean_token_accuracy": 0.21201369166374207,
      "num_tokens": 18651044.0,
      "step": 8130
    },
    {
      "entropy": 5.340291166305542,
      "epoch": 0.7814601344860711,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947036805281325,
      "loss": 5.1051,
      "mean_token_accuracy": 0.21264140903949738,
      "num_tokens": 18663142.0,
      "step": 8135
    },
    {
      "entropy": 5.407776498794556,
      "epoch": 0.781940441882805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946962831887336,
      "loss": 5.1735,
      "mean_token_accuracy": 0.21024056524038315,
      "num_tokens": 18674079.0,
      "step": 8140
    },
    {
      "entropy": 5.303403282165528,
      "epoch": 0.7824207492795389,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946888807486393,
      "loss": 5.1552,
      "mean_token_accuracy": 0.205235031247139,
      "num_tokens": 18685744.0,
      "step": 8145
    },
    {
      "entropy": 5.248520612716675,
      "epoch": 0.7829010566762729,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946814732080214,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21602775305509567,
      "num_tokens": 18697049.0,
      "step": 8150
    },
    {
      "entropy": 5.258074617385864,
      "epoch": 0.7833813640730067,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946740605670523,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2155240595340729,
      "num_tokens": 18708765.0,
      "step": 8155
    },
    {
      "entropy": 5.2804708003997805,
      "epoch": 0.7838616714697406,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946666428259037,
      "loss": 5.0309,
      "mean_token_accuracy": 0.2173507660627365,
      "num_tokens": 18719819.0,
      "step": 8160
    },
    {
      "entropy": 5.360841226577759,
      "epoch": 0.7843419788664745,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946592199847478,
      "loss": 5.102,
      "mean_token_accuracy": 0.20799025595188142,
      "num_tokens": 18730668.0,
      "step": 8165
    },
    {
      "entropy": 5.232181358337402,
      "epoch": 0.7848222862632085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946517920437571,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21585632264614105,
      "num_tokens": 18741991.0,
      "step": 8170
    },
    {
      "entropy": 5.265512418746948,
      "epoch": 0.7853025936599424,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004946443590031041,
      "loss": 5.1293,
      "mean_token_accuracy": 0.21428398042917252,
      "num_tokens": 18753919.0,
      "step": 8175
    },
    {
      "entropy": 5.343008947372437,
      "epoch": 0.7857829010566763,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004946369208629613,
      "loss": 5.0371,
      "mean_token_accuracy": 0.21879064440727233,
      "num_tokens": 18764342.0,
      "step": 8180
    },
    {
      "entropy": 5.224815607070923,
      "epoch": 0.7862632084534101,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004946294776235013,
      "loss": 5.0573,
      "mean_token_accuracy": 0.21366302222013472,
      "num_tokens": 18776044.0,
      "step": 8185
    },
    {
      "entropy": 5.361653757095337,
      "epoch": 0.7867435158501441,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004946220292848971,
      "loss": 5.1236,
      "mean_token_accuracy": 0.21397276669740678,
      "num_tokens": 18787354.0,
      "step": 8190
    },
    {
      "entropy": 5.29656491279602,
      "epoch": 0.787223823246878,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004946145758473214,
      "loss": 5.021,
      "mean_token_accuracy": 0.2148450046777725,
      "num_tokens": 18797845.0,
      "step": 8195
    },
    {
      "entropy": 5.332850360870362,
      "epoch": 0.7877041306436119,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946071173109475,
      "loss": 5.1065,
      "mean_token_accuracy": 0.21267470121383666,
      "num_tokens": 18809253.0,
      "step": 8200
    },
    {
      "entropy": 5.281962251663208,
      "epoch": 0.7881844380403458,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004945996536759484,
      "loss": 4.9557,
      "mean_token_accuracy": 0.2228683426976204,
      "num_tokens": 18819768.0,
      "step": 8205
    },
    {
      "entropy": 5.376068496704102,
      "epoch": 0.7886647454370798,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945921849424974,
      "loss": 5.2085,
      "mean_token_accuracy": 0.20812487453222275,
      "num_tokens": 18831151.0,
      "step": 8210
    },
    {
      "entropy": 5.298772668838501,
      "epoch": 0.7891450528338136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945847111107679,
      "loss": 4.9938,
      "mean_token_accuracy": 0.2198323428630829,
      "num_tokens": 18842133.0,
      "step": 8215
    },
    {
      "entropy": 5.333953189849853,
      "epoch": 0.7896253602305475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945772321809334,
      "loss": 5.1538,
      "mean_token_accuracy": 0.20703756660223008,
      "num_tokens": 18853295.0,
      "step": 8220
    },
    {
      "entropy": 5.309771633148193,
      "epoch": 0.7901056676272814,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004945697481531677,
      "loss": 5.0931,
      "mean_token_accuracy": 0.21196287423372268,
      "num_tokens": 18865802.0,
      "step": 8225
    },
    {
      "entropy": 5.278874111175537,
      "epoch": 0.7905859750240154,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004945622590276443,
      "loss": 5.0205,
      "mean_token_accuracy": 0.2236466497182846,
      "num_tokens": 18877693.0,
      "step": 8230
    },
    {
      "entropy": 5.312116718292236,
      "epoch": 0.7910662824207493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945547648045373,
      "loss": 5.0824,
      "mean_token_accuracy": 0.2173798993229866,
      "num_tokens": 18888549.0,
      "step": 8235
    },
    {
      "entropy": 5.265875720977784,
      "epoch": 0.7915465898174832,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004945472654840206,
      "loss": 5.0175,
      "mean_token_accuracy": 0.2210939735174179,
      "num_tokens": 18899132.0,
      "step": 8240
    },
    {
      "entropy": 5.192903280258179,
      "epoch": 0.7920268972142172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004945397610662683,
      "loss": 5.0256,
      "mean_token_accuracy": 0.20927611887454986,
      "num_tokens": 18911774.0,
      "step": 8245
    },
    {
      "entropy": 5.341684484481812,
      "epoch": 0.792507204610951,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945322515514547,
      "loss": 5.0977,
      "mean_token_accuracy": 0.21303036212921142,
      "num_tokens": 18922806.0,
      "step": 8250
    },
    {
      "entropy": 5.304209280014038,
      "epoch": 0.7929875120076849,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000494524736939754,
      "loss": 5.0702,
      "mean_token_accuracy": 0.2213797003030777,
      "num_tokens": 18934861.0,
      "step": 8255
    },
    {
      "entropy": 5.3362915992736815,
      "epoch": 0.7934678194044188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945172172313408,
      "loss": 5.048,
      "mean_token_accuracy": 0.21906474530696868,
      "num_tokens": 18944269.0,
      "step": 8260
    },
    {
      "entropy": 5.304548740386963,
      "epoch": 0.7939481268011528,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945096924263896,
      "loss": 5.0814,
      "mean_token_accuracy": 0.2116611495614052,
      "num_tokens": 18957197.0,
      "step": 8265
    },
    {
      "entropy": 5.318814945220947,
      "epoch": 0.7944284341978867,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004945021625250753,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20900236517190934,
      "num_tokens": 18968254.0,
      "step": 8270
    },
    {
      "entropy": 5.243633127212524,
      "epoch": 0.7949087415946205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004944946275275724,
      "loss": 4.9881,
      "mean_token_accuracy": 0.22381478399038315,
      "num_tokens": 18979372.0,
      "step": 8275
    },
    {
      "entropy": 5.267259693145752,
      "epoch": 0.7953890489913544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944870874340561,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21762784272432328,
      "num_tokens": 18991075.0,
      "step": 8280
    },
    {
      "entropy": 5.216023206710815,
      "epoch": 0.7958693563880884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944795422447013,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21929397135972978,
      "num_tokens": 19002324.0,
      "step": 8285
    },
    {
      "entropy": 5.336474275588989,
      "epoch": 0.7963496637848223,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004944719919596835,
      "loss": 5.1218,
      "mean_token_accuracy": 0.21083132326602935,
      "num_tokens": 19014406.0,
      "step": 8290
    },
    {
      "entropy": 5.261001443862915,
      "epoch": 0.7968299711815562,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944644365791776,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22041784971952438,
      "num_tokens": 19025984.0,
      "step": 8295
    },
    {
      "entropy": 5.304883861541748,
      "epoch": 0.7973102785782901,
      "grad_norm": 1.0,
      "learning_rate": 0.000494456876103359,
      "loss": 5.113,
      "mean_token_accuracy": 0.2139579251408577,
      "num_tokens": 19036189.0,
      "step": 8300
    },
    {
      "entropy": 5.298628091812134,
      "epoch": 0.7977905859750241,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004944493105324035,
      "loss": 5.0976,
      "mean_token_accuracy": 0.21011899709701537,
      "num_tokens": 19047587.0,
      "step": 8305
    },
    {
      "entropy": 5.321364879608154,
      "epoch": 0.7982708933717579,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944417398664866,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21218617111444474,
      "num_tokens": 19058467.0,
      "step": 8310
    },
    {
      "entropy": 5.311764240264893,
      "epoch": 0.7987512007684918,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004944341641057843,
      "loss": 5.1437,
      "mean_token_accuracy": 0.21255367547273635,
      "num_tokens": 19070235.0,
      "step": 8315
    },
    {
      "entropy": 5.350745058059692,
      "epoch": 0.7992315081652257,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944265832504721,
      "loss": 5.1142,
      "mean_token_accuracy": 0.2127568766474724,
      "num_tokens": 19082005.0,
      "step": 8320
    },
    {
      "entropy": 5.33294358253479,
      "epoch": 0.7997118155619597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944189973007262,
      "loss": 5.1159,
      "mean_token_accuracy": 0.21368397921323776,
      "num_tokens": 19092922.0,
      "step": 8325
    },
    {
      "entropy": 5.3136883735656735,
      "epoch": 0.8001921229586936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004944114062567229,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20616220384836198,
      "num_tokens": 19104832.0,
      "step": 8330
    },
    {
      "entropy": 5.316489171981812,
      "epoch": 0.8006724303554275,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944038101186381,
      "loss": 5.1399,
      "mean_token_accuracy": 0.21133978068828582,
      "num_tokens": 19116261.0,
      "step": 8335
    },
    {
      "entropy": 5.345134830474853,
      "epoch": 0.8011527377521613,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943962088866483,
      "loss": 5.13,
      "mean_token_accuracy": 0.2103720039129257,
      "num_tokens": 19127195.0,
      "step": 8340
    },
    {
      "entropy": 5.354102516174317,
      "epoch": 0.8016330451488953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943886025609301,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20509569197893143,
      "num_tokens": 19138164.0,
      "step": 8345
    },
    {
      "entropy": 5.332017135620117,
      "epoch": 0.8021133525456292,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049438099114166,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20557694584131242,
      "num_tokens": 19149153.0,
      "step": 8350
    },
    {
      "entropy": 5.336039447784424,
      "epoch": 0.8025936599423631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004943733746290147,
      "loss": 5.1328,
      "mean_token_accuracy": 0.20478908866643905,
      "num_tokens": 19161023.0,
      "step": 8355
    },
    {
      "entropy": 5.268671226501465,
      "epoch": 0.803073967339097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494365753023171,
      "loss": 4.9874,
      "mean_token_accuracy": 0.22398886680603028,
      "num_tokens": 19172466.0,
      "step": 8360
    },
    {
      "entropy": 5.311010837554932,
      "epoch": 0.803554274735831,
      "grad_norm": 1.0,
      "learning_rate": 0.0004943581263243059,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21735257506370545,
      "num_tokens": 19183684.0,
      "step": 8365
    },
    {
      "entropy": 5.220240926742553,
      "epoch": 0.8040345821325648,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943504945325965,
      "loss": 4.9761,
      "mean_token_accuracy": 0.22056413292884827,
      "num_tokens": 19194479.0,
      "step": 8370
    },
    {
      "entropy": 5.32010817527771,
      "epoch": 0.8045148895292987,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004943428576482198,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20862782299518584,
      "num_tokens": 19206323.0,
      "step": 8375
    },
    {
      "entropy": 5.388244867324829,
      "epoch": 0.8049951969260326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004943352156713535,
      "loss": 5.1492,
      "mean_token_accuracy": 0.20637688040733337,
      "num_tokens": 19218849.0,
      "step": 8380
    },
    {
      "entropy": 5.249980735778808,
      "epoch": 0.8054755043227666,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943275686021747,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21490692347288132,
      "num_tokens": 19229603.0,
      "step": 8385
    },
    {
      "entropy": 5.330032634735107,
      "epoch": 0.8059558117195005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494319916440861,
      "loss": 5.1572,
      "mean_token_accuracy": 0.2045476630330086,
      "num_tokens": 19241318.0,
      "step": 8390
    },
    {
      "entropy": 5.319420433044433,
      "epoch": 0.8064361191162344,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943122591875901,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2140294134616852,
      "num_tokens": 19252640.0,
      "step": 8395
    },
    {
      "entropy": 5.231271123886108,
      "epoch": 0.8069164265129684,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004943045968425398,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21692116260528566,
      "num_tokens": 19262971.0,
      "step": 8400
    },
    {
      "entropy": 5.307226085662842,
      "epoch": 0.8073967339097022,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004942969294058878,
      "loss": 5.1256,
      "mean_token_accuracy": 0.21388705521821977,
      "num_tokens": 19274426.0,
      "step": 8405
    },
    {
      "entropy": 5.297821855545044,
      "epoch": 0.8078770413064361,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004942892568778125,
      "loss": 5.0831,
      "mean_token_accuracy": 0.216566564142704,
      "num_tokens": 19286806.0,
      "step": 8410
    },
    {
      "entropy": 5.2695924758911135,
      "epoch": 0.80835734870317,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942815792584917,
      "loss": 5.052,
      "mean_token_accuracy": 0.21475101560354232,
      "num_tokens": 19297997.0,
      "step": 8415
    },
    {
      "entropy": 5.280530595779419,
      "epoch": 0.808837656099904,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004942738965481038,
      "loss": 5.0278,
      "mean_token_accuracy": 0.21553569585084914,
      "num_tokens": 19309789.0,
      "step": 8420
    },
    {
      "entropy": 5.273676586151123,
      "epoch": 0.8093179634966379,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004942662087468272,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21843004673719407,
      "num_tokens": 19320688.0,
      "step": 8425
    },
    {
      "entropy": 5.272088527679443,
      "epoch": 0.8097982708933718,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942585158548402,
      "loss": 5.0827,
      "mean_token_accuracy": 0.21727826297283173,
      "num_tokens": 19331670.0,
      "step": 8430
    },
    {
      "entropy": 5.3112115383148195,
      "epoch": 0.8102785782901056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942508178723214,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21233619451522828,
      "num_tokens": 19343578.0,
      "step": 8435
    },
    {
      "entropy": 5.327071809768677,
      "epoch": 0.8107588856868396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942431147994499,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20958428382873534,
      "num_tokens": 19354875.0,
      "step": 8440
    },
    {
      "entropy": 5.278344964981079,
      "epoch": 0.8112391930835735,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942354066364042,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21839356869459153,
      "num_tokens": 19366353.0,
      "step": 8445
    },
    {
      "entropy": 5.199359893798828,
      "epoch": 0.8117195004803074,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004942276933833634,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22472579181194305,
      "num_tokens": 19377534.0,
      "step": 8450
    },
    {
      "entropy": 5.278019905090332,
      "epoch": 0.8121998078770413,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942199750405064,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20319292545318604,
      "num_tokens": 19388188.0,
      "step": 8455
    },
    {
      "entropy": 5.194910717010498,
      "epoch": 0.8126801152737753,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942122516080127,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22797464877367019,
      "num_tokens": 19399910.0,
      "step": 8460
    },
    {
      "entropy": 5.314953899383545,
      "epoch": 0.8131604226705091,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942045230860614,
      "loss": 5.0924,
      "mean_token_accuracy": 0.2186409667134285,
      "num_tokens": 19411715.0,
      "step": 8465
    },
    {
      "entropy": 5.35609073638916,
      "epoch": 0.813640730067243,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004941967894748319,
      "loss": 5.1859,
      "mean_token_accuracy": 0.20044416934251785,
      "num_tokens": 19423275.0,
      "step": 8470
    },
    {
      "entropy": 5.326424169540405,
      "epoch": 0.8141210374639769,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941890507745039,
      "loss": 5.097,
      "mean_token_accuracy": 0.21137083023786546,
      "num_tokens": 19436035.0,
      "step": 8475
    },
    {
      "entropy": 5.260590887069702,
      "epoch": 0.8146013448607109,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004941813069852569,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20944691002368926,
      "num_tokens": 19447755.0,
      "step": 8480
    },
    {
      "entropy": 5.320154428482056,
      "epoch": 0.8150816522574448,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004941735581072708,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21846593767404557,
      "num_tokens": 19459044.0,
      "step": 8485
    },
    {
      "entropy": 5.282533979415893,
      "epoch": 0.8155619596541787,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941658041407255,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21304513663053512,
      "num_tokens": 19471486.0,
      "step": 8490
    },
    {
      "entropy": 5.1884088039398195,
      "epoch": 0.8160422670509125,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494158045085801,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22496270686388015,
      "num_tokens": 19482428.0,
      "step": 8495
    },
    {
      "entropy": 5.263566970825195,
      "epoch": 0.8165225744476465,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004941502809426776,
      "loss": 5.1129,
      "mean_token_accuracy": 0.21201093643903732,
      "num_tokens": 19494616.0,
      "step": 8500
    },
    {
      "entropy": 5.311705684661865,
      "epoch": 0.8170028818443804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941425117115354,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2129536598920822,
      "num_tokens": 19504953.0,
      "step": 8505
    },
    {
      "entropy": 5.31019082069397,
      "epoch": 0.8174831892411143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941347373925547,
      "loss": 5.1061,
      "mean_token_accuracy": 0.21713680177927017,
      "num_tokens": 19517408.0,
      "step": 8510
    },
    {
      "entropy": 5.219766092300415,
      "epoch": 0.8179634966378482,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941269579859161,
      "loss": 5.0417,
      "mean_token_accuracy": 0.22046579122543336,
      "num_tokens": 19529190.0,
      "step": 8515
    },
    {
      "entropy": 5.269935750961304,
      "epoch": 0.8184438040345822,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004941191734918002,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20842382907867432,
      "num_tokens": 19540278.0,
      "step": 8520
    },
    {
      "entropy": 5.2772321701049805,
      "epoch": 0.818924111431316,
      "grad_norm": 1.0,
      "learning_rate": 0.000494111383910388,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21376252621412278,
      "num_tokens": 19551224.0,
      "step": 8525
    },
    {
      "entropy": 5.384726285934448,
      "epoch": 0.8194044188280499,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941035892418597,
      "loss": 5.079,
      "mean_token_accuracy": 0.2107518121600151,
      "num_tokens": 19564386.0,
      "step": 8530
    },
    {
      "entropy": 5.337044095993042,
      "epoch": 0.8198847262247838,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940957894863968,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21811206936836242,
      "num_tokens": 19576170.0,
      "step": 8535
    },
    {
      "entropy": 5.269604015350342,
      "epoch": 0.8203650336215178,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940879846441804,
      "loss": 5.054,
      "mean_token_accuracy": 0.2122248664498329,
      "num_tokens": 19587220.0,
      "step": 8540
    },
    {
      "entropy": 5.269298839569092,
      "epoch": 0.8208453410182517,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004940801747153914,
      "loss": 5.0404,
      "mean_token_accuracy": 0.2137961208820343,
      "num_tokens": 19598649.0,
      "step": 8545
    },
    {
      "entropy": 5.3189338684082035,
      "epoch": 0.8213256484149856,
      "grad_norm": 1.0,
      "learning_rate": 0.0004940723597002113,
      "loss": 5.094,
      "mean_token_accuracy": 0.21246107369661332,
      "num_tokens": 19610243.0,
      "step": 8550
    },
    {
      "entropy": 5.2762247085571286,
      "epoch": 0.8218059558117194,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940645395988216,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20773303806781768,
      "num_tokens": 19621467.0,
      "step": 8555
    },
    {
      "entropy": 5.373956537246704,
      "epoch": 0.8222862632084534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940567144114036,
      "loss": 5.1736,
      "mean_token_accuracy": 0.20707048177719117,
      "num_tokens": 19633367.0,
      "step": 8560
    },
    {
      "entropy": 5.264904308319092,
      "epoch": 0.8227665706051873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940488841381393,
      "loss": 5.042,
      "mean_token_accuracy": 0.21329474300146103,
      "num_tokens": 19643144.0,
      "step": 8565
    },
    {
      "entropy": 5.2883038997650145,
      "epoch": 0.8232468780019212,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940410487792103,
      "loss": 5.1158,
      "mean_token_accuracy": 0.21210776567459105,
      "num_tokens": 19654501.0,
      "step": 8570
    },
    {
      "entropy": 5.3344793796539305,
      "epoch": 0.8237271853986552,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940332083347986,
      "loss": 5.0726,
      "mean_token_accuracy": 0.21060514599084854,
      "num_tokens": 19665382.0,
      "step": 8575
    },
    {
      "entropy": 5.28497314453125,
      "epoch": 0.8242074927953891,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004940253628050861,
      "loss": 4.9169,
      "mean_token_accuracy": 0.2215944454073906,
      "num_tokens": 19677222.0,
      "step": 8580
    },
    {
      "entropy": 5.228455591201782,
      "epoch": 0.824687800192123,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940175121902552,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2186749666929245,
      "num_tokens": 19688550.0,
      "step": 8585
    },
    {
      "entropy": 5.3542252540588375,
      "epoch": 0.8251681075888568,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494009656490488,
      "loss": 5.1282,
      "mean_token_accuracy": 0.20551174283027648,
      "num_tokens": 19700125.0,
      "step": 8590
    },
    {
      "entropy": 5.346072673797607,
      "epoch": 0.8256484149855908,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004940017957059668,
      "loss": 5.1242,
      "mean_token_accuracy": 0.21430510133504868,
      "num_tokens": 19711969.0,
      "step": 8595
    },
    {
      "entropy": 5.289309549331665,
      "epoch": 0.8261287223823247,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939939298368742,
      "loss": 5.0862,
      "mean_token_accuracy": 0.21634469777345658,
      "num_tokens": 19723813.0,
      "step": 8600
    },
    {
      "entropy": 5.3289323329925535,
      "epoch": 0.8266090297790586,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004939860588833929,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20782829970121383,
      "num_tokens": 19735539.0,
      "step": 8605
    },
    {
      "entropy": 5.4116464138031,
      "epoch": 0.8270893371757925,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004939781828457055,
      "loss": 5.17,
      "mean_token_accuracy": 0.20933855473995208,
      "num_tokens": 19747106.0,
      "step": 8610
    },
    {
      "entropy": 5.297164821624756,
      "epoch": 0.8275696445725265,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493970301723995,
      "loss": 5.0903,
      "mean_token_accuracy": 0.22183773517608643,
      "num_tokens": 19760156.0,
      "step": 8615
    },
    {
      "entropy": 5.2421060562133786,
      "epoch": 0.8280499519692603,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939624155184443,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21803604662418366,
      "num_tokens": 19771256.0,
      "step": 8620
    },
    {
      "entropy": 5.2874242782592775,
      "epoch": 0.8285302593659942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939545242292365,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2207653284072876,
      "num_tokens": 19781332.0,
      "step": 8625
    },
    {
      "entropy": 5.342556381225586,
      "epoch": 0.8290105667627281,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939466278565547,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20954972207546235,
      "num_tokens": 19793573.0,
      "step": 8630
    },
    {
      "entropy": 5.270900583267212,
      "epoch": 0.8294908741594621,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939387264005825,
      "loss": 5.0522,
      "mean_token_accuracy": 0.21954465955495833,
      "num_tokens": 19803421.0,
      "step": 8635
    },
    {
      "entropy": 5.230138206481934,
      "epoch": 0.829971181556196,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939308198615031,
      "loss": 4.9319,
      "mean_token_accuracy": 0.22324181348085403,
      "num_tokens": 19814440.0,
      "step": 8640
    },
    {
      "entropy": 5.267609262466431,
      "epoch": 0.8304514889529299,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004939229082395001,
      "loss": 5.0623,
      "mean_token_accuracy": 0.212125663459301,
      "num_tokens": 19825721.0,
      "step": 8645
    },
    {
      "entropy": 5.2827794551849365,
      "epoch": 0.8309317963496637,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939149915347573,
      "loss": 4.988,
      "mean_token_accuracy": 0.22333361208438873,
      "num_tokens": 19837273.0,
      "step": 8650
    },
    {
      "entropy": 5.296159315109253,
      "epoch": 0.8314121037463977,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004939070697474585,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21818834990262986,
      "num_tokens": 19849436.0,
      "step": 8655
    },
    {
      "entropy": 5.228702163696289,
      "epoch": 0.8318924111431316,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938991428777875,
      "loss": 4.9171,
      "mean_token_accuracy": 0.22886181771755218,
      "num_tokens": 19860222.0,
      "step": 8660
    },
    {
      "entropy": 5.275073003768921,
      "epoch": 0.8323727185398655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004938912109259284,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22205011248588563,
      "num_tokens": 19870934.0,
      "step": 8665
    },
    {
      "entropy": 5.318188142776489,
      "epoch": 0.8328530259365994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938832738920654,
      "loss": 5.0825,
      "mean_token_accuracy": 0.2088473320007324,
      "num_tokens": 19882149.0,
      "step": 8670
    },
    {
      "entropy": 5.3246297359466555,
      "epoch": 0.8333333333333334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004938753317763826,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21744603663682938,
      "num_tokens": 19893276.0,
      "step": 8675
    },
    {
      "entropy": 5.317847013473511,
      "epoch": 0.8338136407300673,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938673845790646,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20508771240711213,
      "num_tokens": 19904723.0,
      "step": 8680
    },
    {
      "entropy": 5.305253314971924,
      "epoch": 0.8342939481268011,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004938594323002957,
      "loss": 5.0738,
      "mean_token_accuracy": 0.21140864938497544,
      "num_tokens": 19915226.0,
      "step": 8685
    },
    {
      "entropy": 5.2678173065185545,
      "epoch": 0.834774255523535,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004938514749402609,
      "loss": 5.078,
      "mean_token_accuracy": 0.21481076925992965,
      "num_tokens": 19927911.0,
      "step": 8690
    },
    {
      "entropy": 5.294334888458252,
      "epoch": 0.835254562920269,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004938435124991447,
      "loss": 5.1401,
      "mean_token_accuracy": 0.21382291465997696,
      "num_tokens": 19940356.0,
      "step": 8695
    },
    {
      "entropy": 5.338834476470947,
      "epoch": 0.8357348703170029,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938355449771318,
      "loss": 5.0982,
      "mean_token_accuracy": 0.209849514067173,
      "num_tokens": 19951108.0,
      "step": 8700
    },
    {
      "entropy": 5.2430558681488035,
      "epoch": 0.8362151777137368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938275723744075,
      "loss": 5.0327,
      "mean_token_accuracy": 0.22043029069900513,
      "num_tokens": 19962427.0,
      "step": 8705
    },
    {
      "entropy": 5.359533452987671,
      "epoch": 0.8366954851104706,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938195946911567,
      "loss": 5.1361,
      "mean_token_accuracy": 0.20995877534151078,
      "num_tokens": 19973476.0,
      "step": 8710
    },
    {
      "entropy": 5.330000019073486,
      "epoch": 0.8371757925072046,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938116119275645,
      "loss": 5.0722,
      "mean_token_accuracy": 0.21438438147306443,
      "num_tokens": 19984034.0,
      "step": 8715
    },
    {
      "entropy": 5.327744674682617,
      "epoch": 0.8376560999039385,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938036240838166,
      "loss": 5.1009,
      "mean_token_accuracy": 0.21417146921157837,
      "num_tokens": 19996035.0,
      "step": 8720
    },
    {
      "entropy": 5.346033334732056,
      "epoch": 0.8381364073006724,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937956311600983,
      "loss": 5.0285,
      "mean_token_accuracy": 0.2186152383685112,
      "num_tokens": 20007259.0,
      "step": 8725
    },
    {
      "entropy": 5.299717998504638,
      "epoch": 0.8386167146974063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004937876331565951,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21087097525596618,
      "num_tokens": 20018673.0,
      "step": 8730
    },
    {
      "entropy": 5.246153783798218,
      "epoch": 0.8390970220941403,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937796300734926,
      "loss": 5.0754,
      "mean_token_accuracy": 0.21319071799516678,
      "num_tokens": 20030767.0,
      "step": 8735
    },
    {
      "entropy": 5.349096250534058,
      "epoch": 0.8395773294908742,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937716219109769,
      "loss": 5.1071,
      "mean_token_accuracy": 0.2091129019856453,
      "num_tokens": 20041268.0,
      "step": 8740
    },
    {
      "entropy": 5.251312494277954,
      "epoch": 0.840057636887608,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004937636086692336,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2209983617067337,
      "num_tokens": 20052285.0,
      "step": 8745
    },
    {
      "entropy": 5.310549259185791,
      "epoch": 0.840537944284342,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493755590348449,
      "loss": 5.1681,
      "mean_token_accuracy": 0.21089137941598893,
      "num_tokens": 20065113.0,
      "step": 8750
    },
    {
      "entropy": 5.42260856628418,
      "epoch": 0.8410182516810759,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004937475669488091,
      "loss": 5.146,
      "mean_token_accuracy": 0.21910454481840133,
      "num_tokens": 20076151.0,
      "step": 8755
    },
    {
      "entropy": 5.260375308990478,
      "epoch": 0.8414985590778098,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937395384705004,
      "loss": 5.0748,
      "mean_token_accuracy": 0.21145062148571014,
      "num_tokens": 20088195.0,
      "step": 8760
    },
    {
      "entropy": 5.233353757858277,
      "epoch": 0.8419788664745437,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004937315049137089,
      "loss": 4.9429,
      "mean_token_accuracy": 0.22518647760152816,
      "num_tokens": 20098576.0,
      "step": 8765
    },
    {
      "entropy": 5.389544868469239,
      "epoch": 0.8424591738712777,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004937234662786216,
      "loss": 5.1577,
      "mean_token_accuracy": 0.20495210886001586,
      "num_tokens": 20110176.0,
      "step": 8770
    },
    {
      "entropy": 5.330946350097657,
      "epoch": 0.8429394812680115,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937154225654246,
      "loss": 5.0595,
      "mean_token_accuracy": 0.21761274039745332,
      "num_tokens": 20121713.0,
      "step": 8775
    },
    {
      "entropy": 5.38185715675354,
      "epoch": 0.8434197886647454,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937073737743051,
      "loss": 5.1662,
      "mean_token_accuracy": 0.2096487194299698,
      "num_tokens": 20134126.0,
      "step": 8780
    },
    {
      "entropy": 5.2658624172210695,
      "epoch": 0.8439000960614793,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936993199054499,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21947723925113677,
      "num_tokens": 20144791.0,
      "step": 8785
    },
    {
      "entropy": 5.370013427734375,
      "epoch": 0.8443804034582133,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004936912609590458,
      "loss": 5.1944,
      "mean_token_accuracy": 0.21194605380296708,
      "num_tokens": 20157214.0,
      "step": 8790
    },
    {
      "entropy": 5.4294819831848145,
      "epoch": 0.8448607108549472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049368319693528,
      "loss": 5.196,
      "mean_token_accuracy": 0.20246653705835344,
      "num_tokens": 20168952.0,
      "step": 8795
    },
    {
      "entropy": 5.3150876522064205,
      "epoch": 0.8453410182516811,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004936751278343397,
      "loss": 5.0794,
      "mean_token_accuracy": 0.2139264240860939,
      "num_tokens": 20181829.0,
      "step": 8800
    },
    {
      "entropy": 5.376411628723145,
      "epoch": 0.845821325648415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936670536564123,
      "loss": 5.2182,
      "mean_token_accuracy": 0.19499314427375794,
      "num_tokens": 20193362.0,
      "step": 8805
    },
    {
      "entropy": 5.334952402114868,
      "epoch": 0.8463016330451489,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936589744016853,
      "loss": 5.1317,
      "mean_token_accuracy": 0.21454751193523408,
      "num_tokens": 20205159.0,
      "step": 8810
    },
    {
      "entropy": 5.299839925765991,
      "epoch": 0.8467819404418828,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493650890070346,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2200593590736389,
      "num_tokens": 20217075.0,
      "step": 8815
    },
    {
      "entropy": 5.293173170089721,
      "epoch": 0.8472622478386167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004936428006625824,
      "loss": 5.0254,
      "mean_token_accuracy": 0.21470960974693298,
      "num_tokens": 20227901.0,
      "step": 8820
    },
    {
      "entropy": 5.319810628890991,
      "epoch": 0.8477425552353506,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004936347061785823,
      "loss": 5.0296,
      "mean_token_accuracy": 0.2155350401997566,
      "num_tokens": 20239945.0,
      "step": 8825
    },
    {
      "entropy": 5.354758548736572,
      "epoch": 0.8482228626320846,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004936266066185334,
      "loss": 5.1287,
      "mean_token_accuracy": 0.2104931354522705,
      "num_tokens": 20251503.0,
      "step": 8830
    },
    {
      "entropy": 5.282928800582885,
      "epoch": 0.8487031700288185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004936185019826239,
      "loss": 5.0391,
      "mean_token_accuracy": 0.21541307866573334,
      "num_tokens": 20263044.0,
      "step": 8835
    },
    {
      "entropy": 5.328686380386353,
      "epoch": 0.8491834774255523,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004936103922710419,
      "loss": 5.0284,
      "mean_token_accuracy": 0.21722408384084702,
      "num_tokens": 20273563.0,
      "step": 8840
    },
    {
      "entropy": 5.276285982131958,
      "epoch": 0.8496637848222862,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936022774839759,
      "loss": 5.0769,
      "mean_token_accuracy": 0.21417783200740814,
      "num_tokens": 20285482.0,
      "step": 8845
    },
    {
      "entropy": 5.399579095840454,
      "epoch": 0.8501440922190202,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935941576216141,
      "loss": 5.2558,
      "mean_token_accuracy": 0.20590045154094697,
      "num_tokens": 20296666.0,
      "step": 8850
    },
    {
      "entropy": 5.235482835769654,
      "epoch": 0.8506243996157541,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493586032684145,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22310832291841506,
      "num_tokens": 20308978.0,
      "step": 8855
    },
    {
      "entropy": 5.259793424606324,
      "epoch": 0.851104707012488,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004935779026717573,
      "loss": 5.0229,
      "mean_token_accuracy": 0.2138133704662323,
      "num_tokens": 20321488.0,
      "step": 8860
    },
    {
      "entropy": 5.216874647140503,
      "epoch": 0.8515850144092219,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004935697675846396,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2299219101667404,
      "num_tokens": 20332938.0,
      "step": 8865
    },
    {
      "entropy": 5.230583333969117,
      "epoch": 0.8520653218059558,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935616274229811,
      "loss": 4.9818,
      "mean_token_accuracy": 0.22525002360343932,
      "num_tokens": 20342986.0,
      "step": 8870
    },
    {
      "entropy": 5.284406518936157,
      "epoch": 0.8525456292026897,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004935534821869705,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2141647070646286,
      "num_tokens": 20355791.0,
      "step": 8875
    },
    {
      "entropy": 5.203439950942993,
      "epoch": 0.8530259365994236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004935453318767971,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2205710992217064,
      "num_tokens": 20367080.0,
      "step": 8880
    },
    {
      "entropy": 5.325229930877685,
      "epoch": 0.8535062439961575,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049353717649265,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2246568202972412,
      "num_tokens": 20376697.0,
      "step": 8885
    },
    {
      "entropy": 5.326195812225341,
      "epoch": 0.8539865513928915,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004935290160347185,
      "loss": 5.0512,
      "mean_token_accuracy": 0.2180757850408554,
      "num_tokens": 20387430.0,
      "step": 8890
    },
    {
      "entropy": 5.257937669754028,
      "epoch": 0.8544668587896254,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004935208505031922,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21423773616552352,
      "num_tokens": 20398488.0,
      "step": 8895
    },
    {
      "entropy": 5.200484323501587,
      "epoch": 0.8549471661863592,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935126798982606,
      "loss": 5.0538,
      "mean_token_accuracy": 0.21688165962696077,
      "num_tokens": 20410316.0,
      "step": 8900
    },
    {
      "entropy": 5.285229969024658,
      "epoch": 0.8554274735830932,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004935045042201135,
      "loss": 4.9688,
      "mean_token_accuracy": 0.22495245039463044,
      "num_tokens": 20421645.0,
      "step": 8905
    },
    {
      "entropy": 5.360074520111084,
      "epoch": 0.8559077809798271,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934963234689407,
      "loss": 5.0837,
      "mean_token_accuracy": 0.2081121250987053,
      "num_tokens": 20433397.0,
      "step": 8910
    },
    {
      "entropy": 5.260632276535034,
      "epoch": 0.856388088376561,
      "grad_norm": 0.96875,
      "learning_rate": 0.000493488137644932,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2203802779316902,
      "num_tokens": 20446218.0,
      "step": 8915
    },
    {
      "entropy": 5.333844804763794,
      "epoch": 0.8568683957732949,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004934799467482774,
      "loss": 5.0928,
      "mean_token_accuracy": 0.21161514967679979,
      "num_tokens": 20457265.0,
      "step": 8920
    },
    {
      "entropy": 5.287130737304688,
      "epoch": 0.8573487031700289,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934717507791673,
      "loss": 5.0819,
      "mean_token_accuracy": 0.21309166997671128,
      "num_tokens": 20468748.0,
      "step": 8925
    },
    {
      "entropy": 5.329586601257324,
      "epoch": 0.8578290105667628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934635497377919,
      "loss": 5.1014,
      "mean_token_accuracy": 0.2072741463780403,
      "num_tokens": 20481416.0,
      "step": 8930
    },
    {
      "entropy": 5.24630913734436,
      "epoch": 0.8583093179634966,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934553436243415,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21930938810110093,
      "num_tokens": 20493063.0,
      "step": 8935
    },
    {
      "entropy": 5.213432550430298,
      "epoch": 0.8587896253602305,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004934471324390067,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21791360080242156,
      "num_tokens": 20504000.0,
      "step": 8940
    },
    {
      "entropy": 5.317112588882447,
      "epoch": 0.8592699327569645,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004934389161819783,
      "loss": 5.0849,
      "mean_token_accuracy": 0.2180857315659523,
      "num_tokens": 20516419.0,
      "step": 8945
    },
    {
      "entropy": 5.269755411148071,
      "epoch": 0.8597502401536984,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004934306948534467,
      "loss": 5.035,
      "mean_token_accuracy": 0.2144977107644081,
      "num_tokens": 20527385.0,
      "step": 8950
    },
    {
      "entropy": 5.268268680572509,
      "epoch": 0.8602305475504323,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934224684536031,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21895914226770402,
      "num_tokens": 20538051.0,
      "step": 8955
    },
    {
      "entropy": 5.307522583007812,
      "epoch": 0.8607108549471661,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934142369826382,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21212247163057327,
      "num_tokens": 20550321.0,
      "step": 8960
    },
    {
      "entropy": 5.3252583026885985,
      "epoch": 0.8611911623439001,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934060004407434,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21834986060857772,
      "num_tokens": 20561229.0,
      "step": 8965
    },
    {
      "entropy": 5.257746934890747,
      "epoch": 0.861671469740634,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933977588281099,
      "loss": 4.9956,
      "mean_token_accuracy": 0.21840566098690034,
      "num_tokens": 20572040.0,
      "step": 8970
    },
    {
      "entropy": 5.220415210723877,
      "epoch": 0.8621517771373679,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933895121449288,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21771437376737596,
      "num_tokens": 20583110.0,
      "step": 8975
    },
    {
      "entropy": 5.246572065353393,
      "epoch": 0.8626320845341018,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004933812603913917,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21518482267856598,
      "num_tokens": 20593647.0,
      "step": 8980
    },
    {
      "entropy": 5.202943229675293,
      "epoch": 0.8631123919308358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933730035676903,
      "loss": 4.9012,
      "mean_token_accuracy": 0.23275296092033387,
      "num_tokens": 20604428.0,
      "step": 8985
    },
    {
      "entropy": 5.192376565933228,
      "epoch": 0.8635926993275697,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933647416740161,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2246390014886856,
      "num_tokens": 20615811.0,
      "step": 8990
    },
    {
      "entropy": 5.336459970474243,
      "epoch": 0.8640730067243035,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493356474710561,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21630200743675232,
      "num_tokens": 20627679.0,
      "step": 8995
    },
    {
      "entropy": 5.357700300216675,
      "epoch": 0.8645533141210374,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493348202677517,
      "loss": 5.0918,
      "mean_token_accuracy": 0.2130083382129669,
      "num_tokens": 20638879.0,
      "step": 9000
    },
    {
      "epoch": 0.8645533141210374,
      "eval_entropy": 5.09952763230088,
      "eval_loss": 5.063816070556641,
      "eval_mean_token_accuracy": 0.22313115184264815,
      "eval_num_tokens": 20638879.0,
      "eval_runtime": 20.6986,
      "eval_samples_per_second": 1585.372,
      "eval_steps_per_second": 198.178,
      "step": 9000
    },
    {
      "entropy": 5.275961780548096,
      "epoch": 0.8650336215177714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004933399255750761,
      "loss": 5.0842,
      "mean_token_accuracy": 0.219720821082592,
      "num_tokens": 20649729.0,
      "step": 9005
    },
    {
      "entropy": 5.318403005599976,
      "epoch": 0.8655139289145053,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004933316434034304,
      "loss": 5.1095,
      "mean_token_accuracy": 0.21900081485509873,
      "num_tokens": 20660473.0,
      "step": 9010
    },
    {
      "entropy": 5.354868984222412,
      "epoch": 0.8659942363112392,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933233561627723,
      "loss": 5.085,
      "mean_token_accuracy": 0.21625189036130904,
      "num_tokens": 20671776.0,
      "step": 9015
    },
    {
      "entropy": 5.28105936050415,
      "epoch": 0.866474543707973,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004933150638532942,
      "loss": 5.0614,
      "mean_token_accuracy": 0.21608456820249558,
      "num_tokens": 20684147.0,
      "step": 9020
    },
    {
      "entropy": 5.282857179641724,
      "epoch": 0.866954851104707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004933067664751885,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2147336944937706,
      "num_tokens": 20695248.0,
      "step": 9025
    },
    {
      "entropy": 5.283448982238769,
      "epoch": 0.8674351585014409,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000493298464028648,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21722276955842973,
      "num_tokens": 20707772.0,
      "step": 9030
    },
    {
      "entropy": 5.291160774230957,
      "epoch": 0.8679154658981748,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932901565138653,
      "loss": 5.0143,
      "mean_token_accuracy": 0.2133553460240364,
      "num_tokens": 20718813.0,
      "step": 9035
    },
    {
      "entropy": 5.228615617752075,
      "epoch": 0.8683957732949087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004932818439310334,
      "loss": 5.0032,
      "mean_token_accuracy": 0.22223004698753357,
      "num_tokens": 20730939.0,
      "step": 9040
    },
    {
      "entropy": 5.35224232673645,
      "epoch": 0.8688760806916427,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004932735262803452,
      "loss": 5.095,
      "mean_token_accuracy": 0.2111290842294693,
      "num_tokens": 20742990.0,
      "step": 9045
    },
    {
      "entropy": 5.23276653289795,
      "epoch": 0.8693563880883766,
      "grad_norm": 1.0,
      "learning_rate": 0.0004932652035619939,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2185356765985489,
      "num_tokens": 20754076.0,
      "step": 9050
    },
    {
      "entropy": 5.29207034111023,
      "epoch": 0.8698366954851104,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932568757761727,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22211582213640213,
      "num_tokens": 20765538.0,
      "step": 9055
    },
    {
      "entropy": 5.334879398345947,
      "epoch": 0.8703170028818443,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004932485429230748,
      "loss": 5.0852,
      "mean_token_accuracy": 0.211406147480011,
      "num_tokens": 20776359.0,
      "step": 9060
    },
    {
      "entropy": 5.2680652141571045,
      "epoch": 0.8707973102785783,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493240205002894,
      "loss": 5.0371,
      "mean_token_accuracy": 0.2201646402478218,
      "num_tokens": 20787581.0,
      "step": 9065
    },
    {
      "entropy": 5.334187412261963,
      "epoch": 0.8712776176753122,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004932318620158235,
      "loss": 5.0861,
      "mean_token_accuracy": 0.21241996735334395,
      "num_tokens": 20799904.0,
      "step": 9070
    },
    {
      "entropy": 5.325835514068603,
      "epoch": 0.8717579250720461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004932235139620574,
      "loss": 5.014,
      "mean_token_accuracy": 0.21864811480045318,
      "num_tokens": 20810238.0,
      "step": 9075
    },
    {
      "entropy": 5.258591938018799,
      "epoch": 0.8722382324687801,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932151608417892,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22118595540523528,
      "num_tokens": 20821349.0,
      "step": 9080
    },
    {
      "entropy": 5.214950275421143,
      "epoch": 0.872718539865514,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004932068026552127,
      "loss": 5.0212,
      "mean_token_accuracy": 0.21536519825458528,
      "num_tokens": 20834788.0,
      "step": 9085
    },
    {
      "entropy": 5.313766670227051,
      "epoch": 0.8731988472622478,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004931984394025224,
      "loss": 5.0866,
      "mean_token_accuracy": 0.21473322212696075,
      "num_tokens": 20845571.0,
      "step": 9090
    },
    {
      "entropy": 5.35056791305542,
      "epoch": 0.8736791546589817,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004931900710839123,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21386340260505676,
      "num_tokens": 20857209.0,
      "step": 9095
    },
    {
      "entropy": 5.363465023040772,
      "epoch": 0.8741594620557157,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004931816976995766,
      "loss": 5.1427,
      "mean_token_accuracy": 0.21253460496664048,
      "num_tokens": 20870624.0,
      "step": 9100
    },
    {
      "entropy": 5.2711201190948485,
      "epoch": 0.8746397694524496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931733192497097,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22064325511455535,
      "num_tokens": 20881769.0,
      "step": 9105
    },
    {
      "entropy": 5.22776312828064,
      "epoch": 0.8751200768491835,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004931649357345062,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2130696341395378,
      "num_tokens": 20892817.0,
      "step": 9110
    },
    {
      "entropy": 5.288740587234497,
      "epoch": 0.8756003842459174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004931565471541606,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22544212341308595,
      "num_tokens": 20903042.0,
      "step": 9115
    },
    {
      "entropy": 5.227994728088379,
      "epoch": 0.8760806916426513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004931481535088679,
      "loss": 4.9449,
      "mean_token_accuracy": 0.2286444276571274,
      "num_tokens": 20914684.0,
      "step": 9120
    },
    {
      "entropy": 5.225664138793945,
      "epoch": 0.8765609990393852,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931397547988229,
      "loss": 5.0003,
      "mean_token_accuracy": 0.22618058919906617,
      "num_tokens": 20926585.0,
      "step": 9125
    },
    {
      "entropy": 5.3330024719238285,
      "epoch": 0.8770413064361191,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931313510242204,
      "loss": 5.0646,
      "mean_token_accuracy": 0.21715231686830522,
      "num_tokens": 20939729.0,
      "step": 9130
    },
    {
      "entropy": 5.256030321121216,
      "epoch": 0.877521613832853,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004931229421852557,
      "loss": 4.9897,
      "mean_token_accuracy": 0.2193952441215515,
      "num_tokens": 20951697.0,
      "step": 9135
    },
    {
      "entropy": 5.296226263046265,
      "epoch": 0.878001921229587,
      "grad_norm": 1.0,
      "learning_rate": 0.000493114528282124,
      "loss": 5.113,
      "mean_token_accuracy": 0.21304744184017183,
      "num_tokens": 20962729.0,
      "step": 9140
    },
    {
      "entropy": 5.252359914779663,
      "epoch": 0.8784822286263209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004931061093150206,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22306902557611466,
      "num_tokens": 20973331.0,
      "step": 9145
    },
    {
      "entropy": 5.237087297439575,
      "epoch": 0.8789625360230547,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004930976852841409,
      "loss": 4.9777,
      "mean_token_accuracy": 0.2152834415435791,
      "num_tokens": 20985609.0,
      "step": 9150
    },
    {
      "entropy": 5.243382787704467,
      "epoch": 0.8794428434197886,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004930892561896806,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21254609972238542,
      "num_tokens": 20997231.0,
      "step": 9155
    },
    {
      "entropy": 5.308042573928833,
      "epoch": 0.8799231508165226,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930808220318354,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21969407051801682,
      "num_tokens": 21008511.0,
      "step": 9160
    },
    {
      "entropy": 5.257471323013306,
      "epoch": 0.8804034582132565,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004930723828108012,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2184320420026779,
      "num_tokens": 21019108.0,
      "step": 9165
    },
    {
      "entropy": 5.252701902389527,
      "epoch": 0.8808837656099904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930639385267736,
      "loss": 5.0158,
      "mean_token_accuracy": 0.2219120129942894,
      "num_tokens": 21030621.0,
      "step": 9170
    },
    {
      "entropy": 5.273743295669556,
      "epoch": 0.8813640730067243,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493055489179949,
      "loss": 4.983,
      "mean_token_accuracy": 0.2200742095708847,
      "num_tokens": 21041778.0,
      "step": 9175
    },
    {
      "entropy": 5.273216104507446,
      "epoch": 0.8818443804034583,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930470347705234,
      "loss": 5.009,
      "mean_token_accuracy": 0.22305000722408294,
      "num_tokens": 21054257.0,
      "step": 9180
    },
    {
      "entropy": 5.306494283676147,
      "epoch": 0.8823246878001921,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000493038575298693,
      "loss": 5.0568,
      "mean_token_accuracy": 0.21175634264945983,
      "num_tokens": 21066378.0,
      "step": 9185
    },
    {
      "entropy": 5.216196060180664,
      "epoch": 0.882804995196926,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004930301107646545,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21421182304620742,
      "num_tokens": 21078913.0,
      "step": 9190
    },
    {
      "entropy": 5.282477235794067,
      "epoch": 0.8832853025936599,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004930216411686042,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20952652841806413,
      "num_tokens": 21090500.0,
      "step": 9195
    },
    {
      "entropy": 5.275503540039063,
      "epoch": 0.8837656099903939,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004930131665107387,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21479604095220567,
      "num_tokens": 21102793.0,
      "step": 9200
    },
    {
      "entropy": 5.373411655426025,
      "epoch": 0.8842459173871278,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493004686791255,
      "loss": 5.1091,
      "mean_token_accuracy": 0.21017527282238008,
      "num_tokens": 21114504.0,
      "step": 9205
    },
    {
      "entropy": 5.269224405288696,
      "epoch": 0.8847262247838616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929962020103496,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21784576624631882,
      "num_tokens": 21126733.0,
      "step": 9210
    },
    {
      "entropy": 5.23415994644165,
      "epoch": 0.8852065321805955,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929877121682198,
      "loss": 4.9883,
      "mean_token_accuracy": 0.21848627924919128,
      "num_tokens": 21138045.0,
      "step": 9215
    },
    {
      "entropy": 5.258844518661499,
      "epoch": 0.8856868395773295,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004929792172650627,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21126711666584014,
      "num_tokens": 21151562.0,
      "step": 9220
    },
    {
      "entropy": 5.296943283081054,
      "epoch": 0.8861671469740634,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929707173010753,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21046533137559892,
      "num_tokens": 21162943.0,
      "step": 9225
    },
    {
      "entropy": 5.2792332649230955,
      "epoch": 0.8866474543707973,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004929622122764552,
      "loss": 5.0471,
      "mean_token_accuracy": 0.21485553234815596,
      "num_tokens": 21174392.0,
      "step": 9230
    },
    {
      "entropy": 5.227682876586914,
      "epoch": 0.8871277617675313,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004929537021913997,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22950513809919357,
      "num_tokens": 21185372.0,
      "step": 9235
    },
    {
      "entropy": 5.275816869735718,
      "epoch": 0.8876080691642652,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929451870461064,
      "loss": 5.0529,
      "mean_token_accuracy": 0.2113460347056389,
      "num_tokens": 21197044.0,
      "step": 9240
    },
    {
      "entropy": 5.239752006530762,
      "epoch": 0.888088376560999,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929366668407731,
      "loss": 4.99,
      "mean_token_accuracy": 0.22700745910406112,
      "num_tokens": 21207729.0,
      "step": 9245
    },
    {
      "entropy": 5.314344978332519,
      "epoch": 0.8885686839577329,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004929281415755974,
      "loss": 5.0269,
      "mean_token_accuracy": 0.21649389266967772,
      "num_tokens": 21218909.0,
      "step": 9250
    },
    {
      "entropy": 5.340578269958496,
      "epoch": 0.8890489913544669,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004929196112507775,
      "loss": 5.1189,
      "mean_token_accuracy": 0.21111754775047303,
      "num_tokens": 21230543.0,
      "step": 9255
    },
    {
      "entropy": 5.297629165649414,
      "epoch": 0.8895292987512008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929110758665112,
      "loss": 5.0658,
      "mean_token_accuracy": 0.21511633396148683,
      "num_tokens": 21242064.0,
      "step": 9260
    },
    {
      "entropy": 5.303161668777466,
      "epoch": 0.8900096061479347,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004929025354229969,
      "loss": 5.0943,
      "mean_token_accuracy": 0.21636799722909927,
      "num_tokens": 21254321.0,
      "step": 9265
    },
    {
      "entropy": 5.187288618087768,
      "epoch": 0.8904899135446686,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928939899204326,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22594969272613524,
      "num_tokens": 21264741.0,
      "step": 9270
    },
    {
      "entropy": 5.31791205406189,
      "epoch": 0.8909702209414025,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000492885439359017,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21543063521385192,
      "num_tokens": 21276834.0,
      "step": 9275
    },
    {
      "entropy": 5.233167314529419,
      "epoch": 0.8914505283381364,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004928768837389485,
      "loss": 4.9928,
      "mean_token_accuracy": 0.22092292010784148,
      "num_tokens": 21287108.0,
      "step": 9280
    },
    {
      "entropy": 5.2286313533782955,
      "epoch": 0.8919308357348703,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004928683230604257,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2186286047101021,
      "num_tokens": 21299942.0,
      "step": 9285
    },
    {
      "entropy": 5.405380344390869,
      "epoch": 0.8924111431316042,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928597573236474,
      "loss": 5.1528,
      "mean_token_accuracy": 0.2088737189769745,
      "num_tokens": 21311243.0,
      "step": 9290
    },
    {
      "entropy": 5.310605001449585,
      "epoch": 0.8928914505283382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004928511865288123,
      "loss": 4.9805,
      "mean_token_accuracy": 0.22331418246030807,
      "num_tokens": 21322291.0,
      "step": 9295
    },
    {
      "entropy": 5.283558702468872,
      "epoch": 0.8933717579250721,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004928426106761197,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21648231893777847,
      "num_tokens": 21333257.0,
      "step": 9300
    },
    {
      "entropy": 5.224516820907593,
      "epoch": 0.8938520653218059,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004928340297657685,
      "loss": 4.9831,
      "mean_token_accuracy": 0.21495548784732818,
      "num_tokens": 21345848.0,
      "step": 9305
    },
    {
      "entropy": 5.296548891067505,
      "epoch": 0.8943323727185398,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928254437979578,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2184055417776108,
      "num_tokens": 21357693.0,
      "step": 9310
    },
    {
      "entropy": 5.273598766326904,
      "epoch": 0.8948126801152738,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928168527728873,
      "loss": 5.1255,
      "mean_token_accuracy": 0.21267225593328476,
      "num_tokens": 21369653.0,
      "step": 9315
    },
    {
      "entropy": 5.358665990829468,
      "epoch": 0.8952929875120077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004928082566907562,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21815236955881118,
      "num_tokens": 21383924.0,
      "step": 9320
    },
    {
      "entropy": 5.314771413803101,
      "epoch": 0.8957732949087416,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004927996555517642,
      "loss": 5.0379,
      "mean_token_accuracy": 0.21375614851713182,
      "num_tokens": 21395963.0,
      "step": 9325
    },
    {
      "entropy": 5.281578779220581,
      "epoch": 0.8962536023054755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004927910493561109,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21775185763835908,
      "num_tokens": 21408200.0,
      "step": 9330
    },
    {
      "entropy": 5.351464796066284,
      "epoch": 0.8967339097022095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492782438103996,
      "loss": 5.1354,
      "mean_token_accuracy": 0.20740824490785598,
      "num_tokens": 21419963.0,
      "step": 9335
    },
    {
      "entropy": 5.336421394348145,
      "epoch": 0.8972142170989433,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927738217956197,
      "loss": 5.1068,
      "mean_token_accuracy": 0.214196614921093,
      "num_tokens": 21431824.0,
      "step": 9340
    },
    {
      "entropy": 5.192102336883545,
      "epoch": 0.8976945244956772,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927652004311819,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22789789885282516,
      "num_tokens": 21442354.0,
      "step": 9345
    },
    {
      "entropy": 5.302257823944092,
      "epoch": 0.8981748318924111,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004927565740108828,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21304932832717896,
      "num_tokens": 21453734.0,
      "step": 9350
    },
    {
      "entropy": 5.293565988540649,
      "epoch": 0.8986551392891451,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927479425349226,
      "loss": 5.0411,
      "mean_token_accuracy": 0.21963883489370345,
      "num_tokens": 21465471.0,
      "step": 9355
    },
    {
      "entropy": 5.346841669082641,
      "epoch": 0.899135446685879,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004927393060035018,
      "loss": 5.1696,
      "mean_token_accuracy": 0.2076309561729431,
      "num_tokens": 21477775.0,
      "step": 9360
    },
    {
      "entropy": 5.21389799118042,
      "epoch": 0.8996157540826129,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927306644168207,
      "loss": 4.9522,
      "mean_token_accuracy": 0.22712403088808059,
      "num_tokens": 21489319.0,
      "step": 9365
    },
    {
      "entropy": 5.385450458526611,
      "epoch": 0.9000960614793467,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004927220177750803,
      "loss": 5.1767,
      "mean_token_accuracy": 0.21272541135549544,
      "num_tokens": 21499742.0,
      "step": 9370
    },
    {
      "entropy": 5.272635746002197,
      "epoch": 0.9005763688760807,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004927133660784811,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22372090816497803,
      "num_tokens": 21511063.0,
      "step": 9375
    },
    {
      "entropy": 5.269239664077759,
      "epoch": 0.9010566762728146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004927047093272241,
      "loss": 4.9901,
      "mean_token_accuracy": 0.21752334386110306,
      "num_tokens": 21522500.0,
      "step": 9380
    },
    {
      "entropy": 5.312640619277954,
      "epoch": 0.9015369836695485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049269604752151,
      "loss": 5.0672,
      "mean_token_accuracy": 0.21184938102960588,
      "num_tokens": 21533578.0,
      "step": 9385
    },
    {
      "entropy": 5.277040815353393,
      "epoch": 0.9020172910662824,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926873806615403,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20969802588224412,
      "num_tokens": 21544296.0,
      "step": 9390
    },
    {
      "entropy": 5.300060558319092,
      "epoch": 0.9024975984630164,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926787087475158,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20543698966503143,
      "num_tokens": 21555386.0,
      "step": 9395
    },
    {
      "entropy": 5.256591033935547,
      "epoch": 0.9029779058597502,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004926700317796382,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2336789771914482,
      "num_tokens": 21566527.0,
      "step": 9400
    },
    {
      "entropy": 5.360135650634765,
      "epoch": 0.9034582132564841,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926613497581088,
      "loss": 5.039,
      "mean_token_accuracy": 0.2152058631181717,
      "num_tokens": 21576870.0,
      "step": 9405
    },
    {
      "entropy": 5.275247097015381,
      "epoch": 0.9039385206532181,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926526626831292,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21933796852827073,
      "num_tokens": 21588113.0,
      "step": 9410
    },
    {
      "entropy": 5.256653499603272,
      "epoch": 0.904418828049952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004926439705549011,
      "loss": 5.0317,
      "mean_token_accuracy": 0.2163578286767006,
      "num_tokens": 21599307.0,
      "step": 9415
    },
    {
      "entropy": 5.275369644165039,
      "epoch": 0.9048991354466859,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004926352733736262,
      "loss": 4.9861,
      "mean_token_accuracy": 0.22366381734609603,
      "num_tokens": 21609961.0,
      "step": 9420
    },
    {
      "entropy": 5.240253543853759,
      "epoch": 0.9053794428434198,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926265711395065,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21548174619674682,
      "num_tokens": 21622222.0,
      "step": 9425
    },
    {
      "entropy": 5.257960414886474,
      "epoch": 0.9058597502401537,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492617863852744,
      "loss": 5.0026,
      "mean_token_accuracy": 0.2248610481619835,
      "num_tokens": 21632843.0,
      "step": 9430
    },
    {
      "entropy": 5.2882537841796875,
      "epoch": 0.9063400576368876,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926091515135409,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21369160562753678,
      "num_tokens": 21645387.0,
      "step": 9435
    },
    {
      "entropy": 5.267239332199097,
      "epoch": 0.9068203650336215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926004341220995,
      "loss": 5.0148,
      "mean_token_accuracy": 0.22084719091653823,
      "num_tokens": 21656787.0,
      "step": 9440
    },
    {
      "entropy": 5.329501962661743,
      "epoch": 0.9073006724303554,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925917116786222,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21977902501821517,
      "num_tokens": 21667800.0,
      "step": 9445
    },
    {
      "entropy": 5.309007215499878,
      "epoch": 0.9077809798270894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925829841833114,
      "loss": 5.0708,
      "mean_token_accuracy": 0.21218023598194122,
      "num_tokens": 21679297.0,
      "step": 9450
    },
    {
      "entropy": 5.273968982696533,
      "epoch": 0.9082612872238233,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004925742516363699,
      "loss": 5.0957,
      "mean_token_accuracy": 0.20879530906677246,
      "num_tokens": 21692956.0,
      "step": 9455
    },
    {
      "entropy": 5.255225658416748,
      "epoch": 0.9087415946205571,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004925655140380002,
      "loss": 4.869,
      "mean_token_accuracy": 0.22657041251659393,
      "num_tokens": 21704852.0,
      "step": 9460
    },
    {
      "entropy": 5.309493637084961,
      "epoch": 0.909221902017291,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004925567713884054,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21910935789346694,
      "num_tokens": 21715886.0,
      "step": 9465
    },
    {
      "entropy": 5.236155891418457,
      "epoch": 0.909702209414025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925480236877884,
      "loss": 4.988,
      "mean_token_accuracy": 0.22265698313713073,
      "num_tokens": 21727157.0,
      "step": 9470
    },
    {
      "entropy": 5.252168321609497,
      "epoch": 0.9101825168107589,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925392709363522,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21956454664468766,
      "num_tokens": 21738232.0,
      "step": 9475
    },
    {
      "entropy": 5.249209022521972,
      "epoch": 0.9106628242074928,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004925305131343001,
      "loss": 5.0343,
      "mean_token_accuracy": 0.21398487985134124,
      "num_tokens": 21749029.0,
      "step": 9480
    },
    {
      "entropy": 5.329716444015503,
      "epoch": 0.9111431316042267,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004925217502818355,
      "loss": 5.0389,
      "mean_token_accuracy": 0.2130681172013283,
      "num_tokens": 21761243.0,
      "step": 9485
    },
    {
      "entropy": 5.252294445037842,
      "epoch": 0.9116234390009607,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004925129823791616,
      "loss": 5.015,
      "mean_token_accuracy": 0.21370399594306946,
      "num_tokens": 21772623.0,
      "step": 9490
    },
    {
      "entropy": 5.332067346572876,
      "epoch": 0.9121037463976945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925042094264822,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21759025752544403,
      "num_tokens": 21782611.0,
      "step": 9495
    },
    {
      "entropy": 5.203507804870606,
      "epoch": 0.9125840537944284,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492495431424001,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22751926332712175,
      "num_tokens": 21793946.0,
      "step": 9500
    },
    {
      "entropy": 5.3564716339111325,
      "epoch": 0.9130643611911623,
      "grad_norm": 1.0,
      "learning_rate": 0.0004924866483719216,
      "loss": 5.0971,
      "mean_token_accuracy": 0.21058327555656434,
      "num_tokens": 21803878.0,
      "step": 9505
    },
    {
      "entropy": 5.282685089111328,
      "epoch": 0.9135446685878963,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924778602704481,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21769251078367233,
      "num_tokens": 21815187.0,
      "step": 9510
    },
    {
      "entropy": 5.288547515869141,
      "epoch": 0.9140249759846302,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004924690671197845,
      "loss": 5.0221,
      "mean_token_accuracy": 0.2239673689007759,
      "num_tokens": 21825597.0,
      "step": 9515
    },
    {
      "entropy": 5.22894024848938,
      "epoch": 0.914505283381364,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924602689201348,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22132650017738342,
      "num_tokens": 21837110.0,
      "step": 9520
    },
    {
      "entropy": 5.465541505813599,
      "epoch": 0.9149855907780979,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924514656717034,
      "loss": 5.2398,
      "mean_token_accuracy": 0.2109512597322464,
      "num_tokens": 21847754.0,
      "step": 9525
    },
    {
      "entropy": 5.249364519119263,
      "epoch": 0.9154658981748319,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924426573746948,
      "loss": 5.0179,
      "mean_token_accuracy": 0.21390101462602615,
      "num_tokens": 21859162.0,
      "step": 9530
    },
    {
      "entropy": 5.286064147949219,
      "epoch": 0.9159462055715658,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004924338440293131,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21862683892250062,
      "num_tokens": 21870826.0,
      "step": 9535
    },
    {
      "entropy": 5.27528486251831,
      "epoch": 0.9164265129682997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004924250256357635,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21287347972393036,
      "num_tokens": 21882003.0,
      "step": 9540
    },
    {
      "entropy": 5.264393663406372,
      "epoch": 0.9169068203650336,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924162021942502,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21821916699409485,
      "num_tokens": 21894132.0,
      "step": 9545
    },
    {
      "entropy": 5.22781343460083,
      "epoch": 0.9173871277617676,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004924073737049784,
      "loss": 4.974,
      "mean_token_accuracy": 0.2241566598415375,
      "num_tokens": 21904951.0,
      "step": 9550
    },
    {
      "entropy": 5.239073276519775,
      "epoch": 0.9178674351585014,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923985401681528,
      "loss": 5.0213,
      "mean_token_accuracy": 0.22006384879350663,
      "num_tokens": 21917100.0,
      "step": 9555
    },
    {
      "entropy": 5.313463640213013,
      "epoch": 0.9183477425552353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923897015839788,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2168206363916397,
      "num_tokens": 21927541.0,
      "step": 9560
    },
    {
      "entropy": 5.293189191818238,
      "epoch": 0.9188280499519692,
      "grad_norm": 1.0,
      "learning_rate": 0.0004923808579526613,
      "loss": 4.983,
      "mean_token_accuracy": 0.21795238554477692,
      "num_tokens": 21938723.0,
      "step": 9565
    },
    {
      "entropy": 5.224343633651733,
      "epoch": 0.9193083573487032,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004923720092744059,
      "loss": 4.8948,
      "mean_token_accuracy": 0.2270112782716751,
      "num_tokens": 21950424.0,
      "step": 9570
    },
    {
      "entropy": 5.271740055084228,
      "epoch": 0.9197886647454371,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004923631555494179,
      "loss": 5.0692,
      "mean_token_accuracy": 0.22152447700500488,
      "num_tokens": 21961030.0,
      "step": 9575
    },
    {
      "entropy": 5.296582174301148,
      "epoch": 0.920268972142171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923542967779028,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22273160815238952,
      "num_tokens": 21971625.0,
      "step": 9580
    },
    {
      "entropy": 5.313235950469971,
      "epoch": 0.920749279538905,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923454329600664,
      "loss": 4.9936,
      "mean_token_accuracy": 0.22215207517147065,
      "num_tokens": 21983733.0,
      "step": 9585
    },
    {
      "entropy": 5.2137689113616945,
      "epoch": 0.9212295869356388,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004923365640961143,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22720693051815033,
      "num_tokens": 21995621.0,
      "step": 9590
    },
    {
      "entropy": 5.310101556777954,
      "epoch": 0.9217098943323727,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923276901862526,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21649552136659622,
      "num_tokens": 22007325.0,
      "step": 9595
    },
    {
      "entropy": 5.293409967422486,
      "epoch": 0.9221902017291066,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923188112306874,
      "loss": 4.9853,
      "mean_token_accuracy": 0.21534916013479233,
      "num_tokens": 22017733.0,
      "step": 9600
    },
    {
      "entropy": 5.26995439529419,
      "epoch": 0.9226705091258406,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923099272296246,
      "loss": 5.0661,
      "mean_token_accuracy": 0.2223100945353508,
      "num_tokens": 22030451.0,
      "step": 9605
    },
    {
      "entropy": 5.263484334945678,
      "epoch": 0.9231508165225745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923010381832706,
      "loss": 5.038,
      "mean_token_accuracy": 0.21055614501237868,
      "num_tokens": 22042626.0,
      "step": 9610
    },
    {
      "entropy": 5.2446061134338375,
      "epoch": 0.9236311239193083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004922921440918318,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21793048679828644,
      "num_tokens": 22053314.0,
      "step": 9615
    },
    {
      "entropy": 5.368663930892945,
      "epoch": 0.9241114313160422,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004922832449555144,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20827221274375915,
      "num_tokens": 22064395.0,
      "step": 9620
    },
    {
      "entropy": 5.2703649520874025,
      "epoch": 0.9245917387127762,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004922743407745255,
      "loss": 4.978,
      "mean_token_accuracy": 0.21472820341587068,
      "num_tokens": 22075960.0,
      "step": 9625
    },
    {
      "entropy": 5.31541805267334,
      "epoch": 0.9250720461095101,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922654315490714,
      "loss": 5.0761,
      "mean_token_accuracy": 0.21747578531503678,
      "num_tokens": 22086034.0,
      "step": 9630
    },
    {
      "entropy": 5.295312595367432,
      "epoch": 0.925552353506244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004922565172793593,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2151724874973297,
      "num_tokens": 22096184.0,
      "step": 9635
    },
    {
      "entropy": 5.302003860473633,
      "epoch": 0.9260326609029779,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004922475979655958,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21663730591535568,
      "num_tokens": 22108795.0,
      "step": 9640
    },
    {
      "entropy": 5.254413747787476,
      "epoch": 0.9265129682997119,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922386736079883,
      "loss": 5.0124,
      "mean_token_accuracy": 0.21520925909280778,
      "num_tokens": 22119608.0,
      "step": 9645
    },
    {
      "entropy": 5.278695297241211,
      "epoch": 0.9269932756964457,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922297442067438,
      "loss": 4.9888,
      "mean_token_accuracy": 0.22592408508062362,
      "num_tokens": 22131621.0,
      "step": 9650
    },
    {
      "entropy": 5.330060148239136,
      "epoch": 0.9274735830931796,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004922208097620697,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22365908324718475,
      "num_tokens": 22142745.0,
      "step": 9655
    },
    {
      "entropy": 5.249020385742187,
      "epoch": 0.9279538904899135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004922118702741735,
      "loss": 5.1544,
      "mean_token_accuracy": 0.21266472041606904,
      "num_tokens": 22155457.0,
      "step": 9660
    },
    {
      "entropy": 5.345229911804199,
      "epoch": 0.9284341978866475,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004922029257432625,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2287702515721321,
      "num_tokens": 22165955.0,
      "step": 9665
    },
    {
      "entropy": 5.28854513168335,
      "epoch": 0.9289145052833814,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921939761695446,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21616602540016175,
      "num_tokens": 22178142.0,
      "step": 9670
    },
    {
      "entropy": 5.190385627746582,
      "epoch": 0.9293948126801153,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921850215532275,
      "loss": 4.9331,
      "mean_token_accuracy": 0.22006936371326447,
      "num_tokens": 22190315.0,
      "step": 9675
    },
    {
      "entropy": 5.232425832748413,
      "epoch": 0.9298751200768491,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004921760618945192,
      "loss": 5.0003,
      "mean_token_accuracy": 0.22309097051620483,
      "num_tokens": 22201785.0,
      "step": 9680
    },
    {
      "entropy": 5.303860521316528,
      "epoch": 0.9303554274735831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004921670971936276,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22083207815885544,
      "num_tokens": 22212471.0,
      "step": 9685
    },
    {
      "entropy": 5.3013979434967045,
      "epoch": 0.930835734870317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921581274507607,
      "loss": 5.0367,
      "mean_token_accuracy": 0.22386121451854707,
      "num_tokens": 22223188.0,
      "step": 9690
    },
    {
      "entropy": 5.21935715675354,
      "epoch": 0.9313160422670509,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492149152666127,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21629127115011215,
      "num_tokens": 22233978.0,
      "step": 9695
    },
    {
      "entropy": 5.2537843704223635,
      "epoch": 0.9317963496637848,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004921401728399348,
      "loss": 4.9996,
      "mean_token_accuracy": 0.22612797021865844,
      "num_tokens": 22244713.0,
      "step": 9700
    },
    {
      "entropy": 5.294991779327392,
      "epoch": 0.9322766570605188,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004921311879723926,
      "loss": 5.0532,
      "mean_token_accuracy": 0.21627774238586425,
      "num_tokens": 22256192.0,
      "step": 9705
    },
    {
      "entropy": 5.3255462646484375,
      "epoch": 0.9327569644572526,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004921221980637088,
      "loss": 5.015,
      "mean_token_accuracy": 0.2136331245303154,
      "num_tokens": 22268294.0,
      "step": 9710
    },
    {
      "entropy": 5.313583946228027,
      "epoch": 0.9332372718539865,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921132031140925,
      "loss": 5.0461,
      "mean_token_accuracy": 0.22147811502218245,
      "num_tokens": 22278952.0,
      "step": 9715
    },
    {
      "entropy": 5.262536573410034,
      "epoch": 0.9337175792507204,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004921042031237521,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21746194660663604,
      "num_tokens": 22291057.0,
      "step": 9720
    },
    {
      "entropy": 5.350217485427857,
      "epoch": 0.9341978866474544,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004920951980928969,
      "loss": 5.0824,
      "mean_token_accuracy": 0.21769094318151475,
      "num_tokens": 22302479.0,
      "step": 9725
    },
    {
      "entropy": 5.382465982437134,
      "epoch": 0.9346781940441883,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004920861880217359,
      "loss": 5.143,
      "mean_token_accuracy": 0.2126326158642769,
      "num_tokens": 22315116.0,
      "step": 9730
    },
    {
      "entropy": 5.3078125,
      "epoch": 0.9351585014409222,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920771729104781,
      "loss": 5.0615,
      "mean_token_accuracy": 0.2142432764172554,
      "num_tokens": 22327548.0,
      "step": 9735
    },
    {
      "entropy": 5.2881035804748535,
      "epoch": 0.9356388088376562,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004920681527593329,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22308652251958846,
      "num_tokens": 22339154.0,
      "step": 9740
    },
    {
      "entropy": 5.2584796905517575,
      "epoch": 0.93611911623439,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920591275685098,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22078343629837036,
      "num_tokens": 22350781.0,
      "step": 9745
    },
    {
      "entropy": 5.298837900161743,
      "epoch": 0.9365994236311239,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920500973382184,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21960995495319366,
      "num_tokens": 22361990.0,
      "step": 9750
    },
    {
      "entropy": 5.314497470855713,
      "epoch": 0.9370797310278578,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004920410620686682,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2288723722100258,
      "num_tokens": 22372973.0,
      "step": 9755
    },
    {
      "entropy": 5.209956073760987,
      "epoch": 0.9375600384245918,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920320217600689,
      "loss": 4.9515,
      "mean_token_accuracy": 0.2249452903866768,
      "num_tokens": 22384369.0,
      "step": 9760
    },
    {
      "entropy": 5.259726858139038,
      "epoch": 0.9380403458213257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004920229764126306,
      "loss": 5.0559,
      "mean_token_accuracy": 0.22025755792856216,
      "num_tokens": 22395792.0,
      "step": 9765
    },
    {
      "entropy": 5.309612464904785,
      "epoch": 0.9385206532180596,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920139260265632,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2167820855975151,
      "num_tokens": 22408182.0,
      "step": 9770
    },
    {
      "entropy": 5.391679382324218,
      "epoch": 0.9390009606147934,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920048706020769,
      "loss": 5.1713,
      "mean_token_accuracy": 0.2068209633231163,
      "num_tokens": 22419774.0,
      "step": 9775
    },
    {
      "entropy": 5.190526628494263,
      "epoch": 0.9394812680115274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004919958101393817,
      "loss": 4.874,
      "mean_token_accuracy": 0.22302304208278656,
      "num_tokens": 22430210.0,
      "step": 9780
    },
    {
      "entropy": 5.292817544937134,
      "epoch": 0.9399615754082613,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004919867446386883,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2201724797487259,
      "num_tokens": 22442444.0,
      "step": 9785
    },
    {
      "entropy": 5.232239055633545,
      "epoch": 0.9404418828049952,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491977674100207,
      "loss": 4.9915,
      "mean_token_accuracy": 0.22096169888973236,
      "num_tokens": 22455521.0,
      "step": 9790
    },
    {
      "entropy": 5.309940671920776,
      "epoch": 0.9409221902017291,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919685985241483,
      "loss": 5.0488,
      "mean_token_accuracy": 0.21883592754602432,
      "num_tokens": 22466997.0,
      "step": 9795
    },
    {
      "entropy": 5.268366909027099,
      "epoch": 0.9414024975984631,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491959517910723,
      "loss": 4.961,
      "mean_token_accuracy": 0.22694261372089386,
      "num_tokens": 22477851.0,
      "step": 9800
    },
    {
      "entropy": 5.206273555755615,
      "epoch": 0.9418828049951969,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919504322601421,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2156046524643898,
      "num_tokens": 22489319.0,
      "step": 9805
    },
    {
      "entropy": 5.177804946899414,
      "epoch": 0.9423631123919308,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004919413415726162,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22635700404644013,
      "num_tokens": 22500847.0,
      "step": 9810
    },
    {
      "entropy": 5.3129349708557125,
      "epoch": 0.9428434197886647,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004919322458483566,
      "loss": 5.0026,
      "mean_token_accuracy": 0.22045049518346788,
      "num_tokens": 22512719.0,
      "step": 9815
    },
    {
      "entropy": 5.230278873443604,
      "epoch": 0.9433237271853987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004919231450875745,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22585716843605042,
      "num_tokens": 22522984.0,
      "step": 9820
    },
    {
      "entropy": 5.251847934722901,
      "epoch": 0.9438040345821326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004919140392904809,
      "loss": 5.0024,
      "mean_token_accuracy": 0.22210351824760438,
      "num_tokens": 22534816.0,
      "step": 9825
    },
    {
      "entropy": 5.2458123683929445,
      "epoch": 0.9442843419788665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004919049284572875,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21554143726825714,
      "num_tokens": 22545753.0,
      "step": 9830
    },
    {
      "entropy": 5.334679174423218,
      "epoch": 0.9447646493756003,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004918958125882058,
      "loss": 5.0644,
      "mean_token_accuracy": 0.2130831941962242,
      "num_tokens": 22557237.0,
      "step": 9835
    },
    {
      "entropy": 5.293295860290527,
      "epoch": 0.9452449567723343,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918866916834474,
      "loss": 4.9847,
      "mean_token_accuracy": 0.21890626102685928,
      "num_tokens": 22568909.0,
      "step": 9840
    },
    {
      "entropy": 5.365793514251709,
      "epoch": 0.9457252641690682,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918775657432239,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20735300928354264,
      "num_tokens": 22582162.0,
      "step": 9845
    },
    {
      "entropy": 5.237175178527832,
      "epoch": 0.9462055715658021,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004918684347677474,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21742402911186218,
      "num_tokens": 22592405.0,
      "step": 9850
    },
    {
      "entropy": 5.296424627304077,
      "epoch": 0.946685878962536,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004918592987572298,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2168272390961647,
      "num_tokens": 22603588.0,
      "step": 9855
    },
    {
      "entropy": 5.407385063171387,
      "epoch": 0.94716618635927,
      "grad_norm": 1.0,
      "learning_rate": 0.0004918501577118832,
      "loss": 5.0801,
      "mean_token_accuracy": 0.21735931783914567,
      "num_tokens": 22614995.0,
      "step": 9860
    },
    {
      "entropy": 5.285481071472168,
      "epoch": 0.9476464937560038,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049184101163192,
      "loss": 5.0377,
      "mean_token_accuracy": 0.21201010644435883,
      "num_tokens": 22627556.0,
      "step": 9865
    },
    {
      "entropy": 5.324817323684693,
      "epoch": 0.9481268011527377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918318605175522,
      "loss": 5.0208,
      "mean_token_accuracy": 0.22445139437913894,
      "num_tokens": 22638339.0,
      "step": 9870
    },
    {
      "entropy": 5.3253514766693115,
      "epoch": 0.9486071085494716,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918227043689924,
      "loss": 4.9768,
      "mean_token_accuracy": 0.22335925400257112,
      "num_tokens": 22648922.0,
      "step": 9875
    },
    {
      "entropy": 5.2410728454589846,
      "epoch": 0.9490874159462056,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004918135431864534,
      "loss": 5.0095,
      "mean_token_accuracy": 0.22066012471914292,
      "num_tokens": 22662006.0,
      "step": 9880
    },
    {
      "entropy": 5.216451406478882,
      "epoch": 0.9495677233429395,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004918043769701478,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22412959784269332,
      "num_tokens": 22674649.0,
      "step": 9885
    },
    {
      "entropy": 5.352089738845825,
      "epoch": 0.9500480307396734,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004917952057202882,
      "loss": 5.0572,
      "mean_token_accuracy": 0.2129005089402199,
      "num_tokens": 22685971.0,
      "step": 9890
    },
    {
      "entropy": 5.279277515411377,
      "epoch": 0.9505283381364072,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004917860294370877,
      "loss": 4.9337,
      "mean_token_accuracy": 0.22526732087135315,
      "num_tokens": 22696174.0,
      "step": 9895
    },
    {
      "entropy": 5.315207052230835,
      "epoch": 0.9510086455331412,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917768481207593,
      "loss": 5.046,
      "mean_token_accuracy": 0.2200968846678734,
      "num_tokens": 22706983.0,
      "step": 9900
    },
    {
      "entropy": 5.310713863372802,
      "epoch": 0.9514889529298751,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917676617715162,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2188078910112381,
      "num_tokens": 22718251.0,
      "step": 9905
    },
    {
      "entropy": 5.218424701690674,
      "epoch": 0.951969260326609,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917584703895717,
      "loss": 4.9715,
      "mean_token_accuracy": 0.227272130548954,
      "num_tokens": 22730071.0,
      "step": 9910
    },
    {
      "entropy": 5.250689363479614,
      "epoch": 0.952449567723343,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917492739751391,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21611586213111877,
      "num_tokens": 22742027.0,
      "step": 9915
    },
    {
      "entropy": 5.224556589126587,
      "epoch": 0.9529298751200769,
      "grad_norm": 0.875,
      "learning_rate": 0.000491740072528432,
      "loss": 4.906,
      "mean_token_accuracy": 0.2316231444478035,
      "num_tokens": 22752946.0,
      "step": 9920
    },
    {
      "entropy": 5.3047041416168215,
      "epoch": 0.9534101825168108,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000491730866049664,
      "loss": 4.99,
      "mean_token_accuracy": 0.22021090537309645,
      "num_tokens": 22763944.0,
      "step": 9925
    },
    {
      "entropy": 5.2356555461883545,
      "epoch": 0.9538904899135446,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004917216545390489,
      "loss": 4.957,
      "mean_token_accuracy": 0.22548570930957795,
      "num_tokens": 22774414.0,
      "step": 9930
    },
    {
      "entropy": 5.229720401763916,
      "epoch": 0.9543707973102786,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917124379968004,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23052207231521607,
      "num_tokens": 22785533.0,
      "step": 9935
    },
    {
      "entropy": 5.214996576309204,
      "epoch": 0.9548511047070125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917032164231327,
      "loss": 4.851,
      "mean_token_accuracy": 0.23044895827770234,
      "num_tokens": 22795809.0,
      "step": 9940
    },
    {
      "entropy": 5.268262195587158,
      "epoch": 0.9553314121037464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004916939898182598,
      "loss": 5.0942,
      "mean_token_accuracy": 0.21662921607494354,
      "num_tokens": 22807705.0,
      "step": 9945
    },
    {
      "entropy": 5.322963428497315,
      "epoch": 0.9558117195004803,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916847581823958,
      "loss": 4.9956,
      "mean_token_accuracy": 0.2204671248793602,
      "num_tokens": 22818852.0,
      "step": 9950
    },
    {
      "entropy": 5.252047204971314,
      "epoch": 0.9562920268972143,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004916755215157552,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22474018335342408,
      "num_tokens": 22829146.0,
      "step": 9955
    },
    {
      "entropy": 5.237692785263062,
      "epoch": 0.9567723342939481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004916662798185524,
      "loss": 4.9904,
      "mean_token_accuracy": 0.22866752445697786,
      "num_tokens": 22840088.0,
      "step": 9960
    },
    {
      "entropy": 5.269695711135864,
      "epoch": 0.957252641690682,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916570330910019,
      "loss": 5.0164,
      "mean_token_accuracy": 0.2274025246500969,
      "num_tokens": 22852470.0,
      "step": 9965
    },
    {
      "entropy": 5.327831745147705,
      "epoch": 0.9577329490874159,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004916477813333185,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21092079132795333,
      "num_tokens": 22863673.0,
      "step": 9970
    },
    {
      "entropy": 5.253059434890747,
      "epoch": 0.9582132564841499,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004916385245457168,
      "loss": 5.0332,
      "mean_token_accuracy": 0.21443782299757003,
      "num_tokens": 22874888.0,
      "step": 9975
    },
    {
      "entropy": 5.270570945739746,
      "epoch": 0.9586935638808838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000491629262728412,
      "loss": 5.0397,
      "mean_token_accuracy": 0.21091512590646744,
      "num_tokens": 22886811.0,
      "step": 9980
    },
    {
      "entropy": 5.333252477645874,
      "epoch": 0.9591738712776177,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916199958816188,
      "loss": 4.9971,
      "mean_token_accuracy": 0.21687838435173035,
      "num_tokens": 22898777.0,
      "step": 9985
    },
    {
      "entropy": 5.302281188964844,
      "epoch": 0.9596541786743515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004916107240055527,
      "loss": 4.9597,
      "mean_token_accuracy": 0.22339343279600143,
      "num_tokens": 22910804.0,
      "step": 9990
    },
    {
      "entropy": 5.317459630966186,
      "epoch": 0.9601344860710855,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916014471004287,
      "loss": 5.1104,
      "mean_token_accuracy": 0.2199084520339966,
      "num_tokens": 22922002.0,
      "step": 9995
    },
    {
      "entropy": 5.299905061721802,
      "epoch": 0.9606147934678194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004915921651664622,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21688541918992996,
      "num_tokens": 22933471.0,
      "step": 10000
    },
    {
      "entropy": 5.200297498703003,
      "epoch": 0.9610951008645533,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000491582878203869,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22504227608442307,
      "num_tokens": 22945303.0,
      "step": 10005
    },
    {
      "entropy": 5.238185453414917,
      "epoch": 0.9615754082612872,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915735862128643,
      "loss": 5.0522,
      "mean_token_accuracy": 0.2127025917172432,
      "num_tokens": 22956620.0,
      "step": 10010
    },
    {
      "entropy": 5.314497995376587,
      "epoch": 0.9620557156580212,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915642891936641,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21933964937925338,
      "num_tokens": 22968941.0,
      "step": 10015
    },
    {
      "entropy": 5.32964186668396,
      "epoch": 0.962536023054755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915549871464841,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2148846358060837,
      "num_tokens": 22980222.0,
      "step": 10020
    },
    {
      "entropy": 5.329556703567505,
      "epoch": 0.9630163304514889,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915456800715403,
      "loss": 5.0394,
      "mean_token_accuracy": 0.21293448209762572,
      "num_tokens": 22991156.0,
      "step": 10025
    },
    {
      "entropy": 5.339330101013184,
      "epoch": 0.9634966378482228,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491536367969049,
      "loss": 5.176,
      "mean_token_accuracy": 0.20958519876003265,
      "num_tokens": 23002939.0,
      "step": 10030
    },
    {
      "entropy": 5.383946752548217,
      "epoch": 0.9639769452449568,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004915270508392261,
      "loss": 5.0258,
      "mean_token_accuracy": 0.21836361289024353,
      "num_tokens": 23015590.0,
      "step": 10035
    },
    {
      "entropy": 5.2001707553863525,
      "epoch": 0.9644572526416907,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000491517728682288,
      "loss": 4.908,
      "mean_token_accuracy": 0.2255134701728821,
      "num_tokens": 23026387.0,
      "step": 10040
    },
    {
      "entropy": 5.171950578689575,
      "epoch": 0.9649375600384246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915084014984512,
      "loss": 4.9816,
      "mean_token_accuracy": 0.22336099445819854,
      "num_tokens": 23037475.0,
      "step": 10045
    },
    {
      "entropy": 5.248710346221924,
      "epoch": 0.9654178674351584,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004914990692879322,
      "loss": 4.9539,
      "mean_token_accuracy": 0.22828294187784196,
      "num_tokens": 23049305.0,
      "step": 10050
    },
    {
      "entropy": 5.296443319320678,
      "epoch": 0.9658981748318924,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914897320509478,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22198242992162703,
      "num_tokens": 23061765.0,
      "step": 10055
    },
    {
      "entropy": 5.221637773513794,
      "epoch": 0.9663784822286263,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914803897877146,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21211671084165573,
      "num_tokens": 23072355.0,
      "step": 10060
    },
    {
      "entropy": 5.309389686584472,
      "epoch": 0.9668587896253602,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004914710424984495,
      "loss": 4.997,
      "mean_token_accuracy": 0.21252110302448274,
      "num_tokens": 23085583.0,
      "step": 10065
    },
    {
      "entropy": 5.340275287628174,
      "epoch": 0.9673390970220941,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914616901833696,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2156083181500435,
      "num_tokens": 23095942.0,
      "step": 10070
    },
    {
      "entropy": 5.243502807617188,
      "epoch": 0.9678194044188281,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000491452332842692,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22689363807439805,
      "num_tokens": 23106540.0,
      "step": 10075
    },
    {
      "entropy": 5.269853830337524,
      "epoch": 0.968299711815562,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491442970476634,
      "loss": 5.0274,
      "mean_token_accuracy": 0.22504184693098067,
      "num_tokens": 23118006.0,
      "step": 10080
    },
    {
      "entropy": 5.247406387329102,
      "epoch": 0.9687800192122958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004914336030854129,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22886505872011184,
      "num_tokens": 23129103.0,
      "step": 10085
    },
    {
      "entropy": 5.205692768096924,
      "epoch": 0.9692603266090298,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004914242306692461,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22711916714906694,
      "num_tokens": 23140009.0,
      "step": 10090
    },
    {
      "entropy": 5.260742330551148,
      "epoch": 0.9697406340057637,
      "grad_norm": 0.875,
      "learning_rate": 0.0004914148532283516,
      "loss": 4.9949,
      "mean_token_accuracy": 0.22203567028045654,
      "num_tokens": 23150982.0,
      "step": 10095
    },
    {
      "entropy": 5.300123643875122,
      "epoch": 0.9702209414024976,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914054707629466,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2162649765610695,
      "num_tokens": 23161834.0,
      "step": 10100
    },
    {
      "entropy": 5.306209993362427,
      "epoch": 0.9707012487992315,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913960832732493,
      "loss": 5.1296,
      "mean_token_accuracy": 0.2124180629849434,
      "num_tokens": 23173355.0,
      "step": 10105
    },
    {
      "entropy": 5.306587505340576,
      "epoch": 0.9711815561959655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913866907594774,
      "loss": 5.0572,
      "mean_token_accuracy": 0.21493452936410903,
      "num_tokens": 23185075.0,
      "step": 10110
    },
    {
      "entropy": 5.278501987457275,
      "epoch": 0.9716618635926993,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004913772932218491,
      "loss": 5.02,
      "mean_token_accuracy": 0.22556180357933045,
      "num_tokens": 23195590.0,
      "step": 10115
    },
    {
      "entropy": 5.310996961593628,
      "epoch": 0.9721421709894332,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913678906605825,
      "loss": 5.0042,
      "mean_token_accuracy": 0.21865737140178682,
      "num_tokens": 23207668.0,
      "step": 10120
    },
    {
      "entropy": 5.226266670227051,
      "epoch": 0.9726224783861671,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004913584830758961,
      "loss": 4.9752,
      "mean_token_accuracy": 0.23079055100679396,
      "num_tokens": 23218497.0,
      "step": 10125
    },
    {
      "entropy": 5.356466436386109,
      "epoch": 0.9731027857829011,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004913490704680081,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21443286836147307,
      "num_tokens": 23230575.0,
      "step": 10130
    },
    {
      "entropy": 5.304587697982788,
      "epoch": 0.973583093179635,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004913396528371371,
      "loss": 5.0577,
      "mean_token_accuracy": 0.21829638183116912,
      "num_tokens": 23242348.0,
      "step": 10135
    },
    {
      "entropy": 5.294364261627197,
      "epoch": 0.9740634005763689,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913302301835018,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21942730247974396,
      "num_tokens": 23253297.0,
      "step": 10140
    },
    {
      "entropy": 5.244136905670166,
      "epoch": 0.9745437079731027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491320802507321,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22819057404994963,
      "num_tokens": 23265830.0,
      "step": 10145
    },
    {
      "entropy": 5.309697675704956,
      "epoch": 0.9750240153698367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004913113698088133,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21807650923728944,
      "num_tokens": 23275591.0,
      "step": 10150
    },
    {
      "entropy": 5.262397241592407,
      "epoch": 0.9755043227665706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491301932088198,
      "loss": 4.941,
      "mean_token_accuracy": 0.22249893248081207,
      "num_tokens": 23286685.0,
      "step": 10155
    },
    {
      "entropy": 5.216834354400635,
      "epoch": 0.9759846301633045,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912924893456942,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22529801577329636,
      "num_tokens": 23298776.0,
      "step": 10160
    },
    {
      "entropy": 5.239245653152466,
      "epoch": 0.9764649375600384,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000491283041581521,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22041333615779876,
      "num_tokens": 23310498.0,
      "step": 10165
    },
    {
      "entropy": 5.278992366790772,
      "epoch": 0.9769452449567724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004912735887958978,
      "loss": 5.001,
      "mean_token_accuracy": 0.22294555008411407,
      "num_tokens": 23321089.0,
      "step": 10170
    },
    {
      "entropy": 5.278607797622681,
      "epoch": 0.9774255523535063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912641309890441,
      "loss": 4.9773,
      "mean_token_accuracy": 0.22036618441343309,
      "num_tokens": 23332142.0,
      "step": 10175
    },
    {
      "entropy": 5.205700206756592,
      "epoch": 0.9779058597502401,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004912546681611794,
      "loss": 4.9428,
      "mean_token_accuracy": 0.2297223225235939,
      "num_tokens": 23343014.0,
      "step": 10180
    },
    {
      "entropy": 5.277859115600586,
      "epoch": 0.978386167146974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912452003125234,
      "loss": 4.9384,
      "mean_token_accuracy": 0.2263363406062126,
      "num_tokens": 23354611.0,
      "step": 10185
    },
    {
      "entropy": 5.273922157287598,
      "epoch": 0.978866474543708,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491235727443296,
      "loss": 4.987,
      "mean_token_accuracy": 0.2161739006638527,
      "num_tokens": 23365608.0,
      "step": 10190
    },
    {
      "entropy": 5.288276243209839,
      "epoch": 0.9793467819404419,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912262495537171,
      "loss": 5.0096,
      "mean_token_accuracy": 0.2223093569278717,
      "num_tokens": 23377884.0,
      "step": 10195
    },
    {
      "entropy": 5.184326696395874,
      "epoch": 0.9798270893371758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004912167666440068,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22272637337446213,
      "num_tokens": 23389553.0,
      "step": 10200
    },
    {
      "entropy": 5.211463260650635,
      "epoch": 0.9803073967339097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912072787143852,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22205393761396408,
      "num_tokens": 23401079.0,
      "step": 10205
    },
    {
      "entropy": 5.23992509841919,
      "epoch": 0.9807877041306436,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004911977857650725,
      "loss": 4.9777,
      "mean_token_accuracy": 0.221847040951252,
      "num_tokens": 23412886.0,
      "step": 10210
    },
    {
      "entropy": 5.308433580398559,
      "epoch": 0.9812680115273775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004911882877962893,
      "loss": 5.033,
      "mean_token_accuracy": 0.21274041682481765,
      "num_tokens": 23424758.0,
      "step": 10215
    },
    {
      "entropy": 5.220535373687744,
      "epoch": 0.9817483189241114,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911787848082559,
      "loss": 4.9,
      "mean_token_accuracy": 0.23109544515609742,
      "num_tokens": 23435552.0,
      "step": 10220
    },
    {
      "entropy": 5.188296937942505,
      "epoch": 0.9822286263208453,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911692768011931,
      "loss": 4.909,
      "mean_token_accuracy": 0.22794807702302933,
      "num_tokens": 23446584.0,
      "step": 10225
    },
    {
      "entropy": 5.351886177062989,
      "epoch": 0.9827089337175793,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004911597637753217,
      "loss": 5.1375,
      "mean_token_accuracy": 0.20629780739545822,
      "num_tokens": 23458452.0,
      "step": 10230
    },
    {
      "entropy": 5.298908948898315,
      "epoch": 0.9831892411143132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911502457308623,
      "loss": 5.005,
      "mean_token_accuracy": 0.2151845797896385,
      "num_tokens": 23470310.0,
      "step": 10235
    },
    {
      "entropy": 5.344496870040894,
      "epoch": 0.983669548511047,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491140722668036,
      "loss": 5.0387,
      "mean_token_accuracy": 0.22020720392465593,
      "num_tokens": 23481166.0,
      "step": 10240
    },
    {
      "entropy": 5.268197011947632,
      "epoch": 0.984149855907781,
      "grad_norm": 0.953125,
      "learning_rate": 0.000491131194587064,
      "loss": 5.0228,
      "mean_token_accuracy": 0.22076378464698793,
      "num_tokens": 23493134.0,
      "step": 10245
    },
    {
      "entropy": 5.2058045864105225,
      "epoch": 0.9846301633045149,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004911216614881675,
      "loss": 4.9065,
      "mean_token_accuracy": 0.2243533819913864,
      "num_tokens": 23504983.0,
      "step": 10250
    },
    {
      "entropy": 5.317254447937012,
      "epoch": 0.9851104707012488,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004911121233715677,
      "loss": 5.0004,
      "mean_token_accuracy": 0.22312158197164536,
      "num_tokens": 23516119.0,
      "step": 10255
    },
    {
      "entropy": 5.348520135879516,
      "epoch": 0.9855907780979827,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004911025802374861,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21397202908992768,
      "num_tokens": 23528242.0,
      "step": 10260
    },
    {
      "entropy": 5.210879993438721,
      "epoch": 0.9860710854947167,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910930320861442,
      "loss": 4.916,
      "mean_token_accuracy": 0.23077184706926346,
      "num_tokens": 23539738.0,
      "step": 10265
    },
    {
      "entropy": 5.127011775970459,
      "epoch": 0.9865513928914506,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004910834789177639,
      "loss": 4.9322,
      "mean_token_accuracy": 0.224748694896698,
      "num_tokens": 23551228.0,
      "step": 10270
    },
    {
      "entropy": 5.303711557388306,
      "epoch": 0.9870317002881844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004910739207325668,
      "loss": 4.9876,
      "mean_token_accuracy": 0.22397853881120683,
      "num_tokens": 23563084.0,
      "step": 10275
    },
    {
      "entropy": 5.270771455764771,
      "epoch": 0.9875120076849183,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004910643575307749,
      "loss": 4.9513,
      "mean_token_accuracy": 0.2266560062766075,
      "num_tokens": 23574328.0,
      "step": 10280
    },
    {
      "entropy": 5.1867015838623045,
      "epoch": 0.9879923150816523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910547893126102,
      "loss": 4.9598,
      "mean_token_accuracy": 0.225901797413826,
      "num_tokens": 23585230.0,
      "step": 10285
    },
    {
      "entropy": 5.22535891532898,
      "epoch": 0.9884726224783862,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004910452160782948,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21333660036325455,
      "num_tokens": 23596951.0,
      "step": 10290
    },
    {
      "entropy": 5.326337051391602,
      "epoch": 0.9889529298751201,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491035637828051,
      "loss": 5.0075,
      "mean_token_accuracy": 0.2236792489886284,
      "num_tokens": 23607759.0,
      "step": 10295
    },
    {
      "entropy": 5.289549160003662,
      "epoch": 0.989433237271854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910260545621012,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21728943884372712,
      "num_tokens": 23619631.0,
      "step": 10300
    },
    {
      "entropy": 5.241495561599732,
      "epoch": 0.9899135446685879,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004910164662806679,
      "loss": 5.0305,
      "mean_token_accuracy": 0.22212643921375275,
      "num_tokens": 23630601.0,
      "step": 10305
    },
    {
      "entropy": 5.301242208480835,
      "epoch": 0.9903938520653218,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004910068729839736,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22495593279600143,
      "num_tokens": 23641330.0,
      "step": 10310
    },
    {
      "entropy": 5.301807260513305,
      "epoch": 0.9908741594620557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004909972746722413,
      "loss": 5.0198,
      "mean_token_accuracy": 0.21776878833770752,
      "num_tokens": 23651492.0,
      "step": 10315
    },
    {
      "entropy": 5.211068248748779,
      "epoch": 0.9913544668587896,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004909876713456935,
      "loss": 4.907,
      "mean_token_accuracy": 0.22746839225292206,
      "num_tokens": 23661773.0,
      "step": 10320
    },
    {
      "entropy": 5.259599733352661,
      "epoch": 0.9918347742555236,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004909780630045534,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2244741290807724,
      "num_tokens": 23673534.0,
      "step": 10325
    },
    {
      "entropy": 5.300588703155517,
      "epoch": 0.9923150816522575,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000490968449649044,
      "loss": 5.0549,
      "mean_token_accuracy": 0.22067267298698426,
      "num_tokens": 23684892.0,
      "step": 10330
    },
    {
      "entropy": 5.330635261535645,
      "epoch": 0.9927953890489913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004909588312793884,
      "loss": 5.0443,
      "mean_token_accuracy": 0.21706904023885726,
      "num_tokens": 23696076.0,
      "step": 10335
    },
    {
      "entropy": 5.325946092605591,
      "epoch": 0.9932756964457252,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004909492078958101,
      "loss": 5.0261,
      "mean_token_accuracy": 0.21531579941511153,
      "num_tokens": 23707795.0,
      "step": 10340
    },
    {
      "entropy": 5.249337148666382,
      "epoch": 0.9937560038424592,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004909395794985324,
      "loss": 4.9383,
      "mean_token_accuracy": 0.2242613449692726,
      "num_tokens": 23720802.0,
      "step": 10345
    },
    {
      "entropy": 5.248753881454467,
      "epoch": 0.9942363112391931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909299460877788,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21588589549064635,
      "num_tokens": 23732854.0,
      "step": 10350
    },
    {
      "entropy": 5.310943174362182,
      "epoch": 0.994716618635927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909203076637732,
      "loss": 5.0629,
      "mean_token_accuracy": 0.21553382277488708,
      "num_tokens": 23743593.0,
      "step": 10355
    },
    {
      "entropy": 5.303987073898315,
      "epoch": 0.9951969260326609,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909106642267392,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21632853001356125,
      "num_tokens": 23755447.0,
      "step": 10360
    },
    {
      "entropy": 5.241162872314453,
      "epoch": 0.9956772334293948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004909010157769006,
      "loss": 5.0171,
      "mean_token_accuracy": 0.21970560401678085,
      "num_tokens": 23767181.0,
      "step": 10365
    },
    {
      "entropy": 5.399766731262207,
      "epoch": 0.9961575408261287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004908913623144814,
      "loss": 5.0938,
      "mean_token_accuracy": 0.2096856191754341,
      "num_tokens": 23776356.0,
      "step": 10370
    },
    {
      "entropy": 5.27275276184082,
      "epoch": 0.9966378482228626,
      "grad_norm": 1.015625,
      "learning_rate": 0.000490881703839706,
      "loss": 5.0546,
      "mean_token_accuracy": 0.2227531149983406,
      "num_tokens": 23787965.0,
      "step": 10375
    },
    {
      "entropy": 5.205696630477905,
      "epoch": 0.9971181556195965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908720403527984,
      "loss": 4.9103,
      "mean_token_accuracy": 0.2274794787168503,
      "num_tokens": 23800327.0,
      "step": 10380
    },
    {
      "entropy": 5.243367147445679,
      "epoch": 0.9975984630163305,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000490862371853983,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22548123598098754,
      "num_tokens": 23812845.0,
      "step": 10385
    },
    {
      "entropy": 5.353694248199463,
      "epoch": 0.9980787704130644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004908526983434844,
      "loss": 5.0803,
      "mean_token_accuracy": 0.2143431007862091,
      "num_tokens": 23824831.0,
      "step": 10390
    },
    {
      "entropy": 5.27176308631897,
      "epoch": 0.9985590778097982,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000490843019821527,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22548753172159194,
      "num_tokens": 23836697.0,
      "step": 10395
    },
    {
      "entropy": 5.225878715515137,
      "epoch": 0.9990393852065321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004908333362883358,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2244037002325058,
      "num_tokens": 23847112.0,
      "step": 10400
    },
    {
      "entropy": 5.3030657291412355,
      "epoch": 0.9995196926032661,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004908236477441353,
      "loss": 5.0518,
      "mean_token_accuracy": 0.22107020318508147,
      "num_tokens": 23858185.0,
      "step": 10405
    },
    {
      "entropy": 5.167287015914917,
      "epoch": 1.0,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004908139541891505,
      "loss": 4.8275,
      "mean_token_accuracy": 0.23650120943784714,
      "num_tokens": 23868536.0,
      "step": 10410
    },
    {
      "entropy": 5.274187660217285,
      "epoch": 1.0004803073967339,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004908042556236066,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23181226551532746,
      "num_tokens": 23880283.0,
      "step": 10415
    },
    {
      "entropy": 5.309299087524414,
      "epoch": 1.0009606147934678,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004907945520477286,
      "loss": 4.9602,
      "mean_token_accuracy": 0.22216319739818574,
      "num_tokens": 23892413.0,
      "step": 10420
    },
    {
      "entropy": 5.365721607208252,
      "epoch": 1.0014409221902016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004907848434617419,
      "loss": 5.0658,
      "mean_token_accuracy": 0.21287601441144943,
      "num_tokens": 23903977.0,
      "step": 10425
    },
    {
      "entropy": 5.333603477478027,
      "epoch": 1.0019212295869357,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000490775129865872,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2165704056620598,
      "num_tokens": 23915153.0,
      "step": 10430
    },
    {
      "entropy": 5.2360066890716555,
      "epoch": 1.0024015369836696,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907654112603442,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2211843192577362,
      "num_tokens": 23926043.0,
      "step": 10435
    },
    {
      "entropy": 5.193376445770264,
      "epoch": 1.0028818443804035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004907556876453843,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23309382051229477,
      "num_tokens": 23936658.0,
      "step": 10440
    },
    {
      "entropy": 5.263293027877808,
      "epoch": 1.0033621517771374,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000490745959021218,
      "loss": 4.9178,
      "mean_token_accuracy": 0.21794203966856002,
      "num_tokens": 23947676.0,
      "step": 10445
    },
    {
      "entropy": 5.253292465209961,
      "epoch": 1.0038424591738713,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004907362253880711,
      "loss": 4.901,
      "mean_token_accuracy": 0.22251911461353302,
      "num_tokens": 23959130.0,
      "step": 10450
    },
    {
      "entropy": 5.35165114402771,
      "epoch": 1.0043227665706052,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907264867461697,
      "loss": 5.0073,
      "mean_token_accuracy": 0.21690160930156707,
      "num_tokens": 23969905.0,
      "step": 10455
    },
    {
      "entropy": 5.225652074813842,
      "epoch": 1.004803073967339,
      "grad_norm": 0.875,
      "learning_rate": 0.0004907167430957399,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22616911083459854,
      "num_tokens": 23982016.0,
      "step": 10460
    },
    {
      "entropy": 5.268946123123169,
      "epoch": 1.005283381364073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907069944370077,
      "loss": 5.0237,
      "mean_token_accuracy": 0.22152279764413835,
      "num_tokens": 23994200.0,
      "step": 10465
    },
    {
      "entropy": 5.368046569824219,
      "epoch": 1.005763688760807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004906972407701998,
      "loss": 4.9216,
      "mean_token_accuracy": 0.23149920254945755,
      "num_tokens": 24004695.0,
      "step": 10470
    },
    {
      "entropy": 5.236816310882569,
      "epoch": 1.006243996157541,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004906874820955423,
      "loss": 4.869,
      "mean_token_accuracy": 0.2293284609913826,
      "num_tokens": 24015922.0,
      "step": 10475
    },
    {
      "entropy": 5.195282602310181,
      "epoch": 1.0067243035542748,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004906777184132621,
      "loss": 4.8784,
      "mean_token_accuracy": 0.22464821636676788,
      "num_tokens": 24026759.0,
      "step": 10480
    },
    {
      "entropy": 5.286918687820434,
      "epoch": 1.0072046109510087,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004906679497235856,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21885820776224135,
      "num_tokens": 24037988.0,
      "step": 10485
    },
    {
      "entropy": 5.320085763931274,
      "epoch": 1.0076849183477425,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906581760267397,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22248762249946594,
      "num_tokens": 24050837.0,
      "step": 10490
    },
    {
      "entropy": 5.268432140350342,
      "epoch": 1.0081652257444764,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906483973229513,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2319096013903618,
      "num_tokens": 24063085.0,
      "step": 10495
    },
    {
      "entropy": 5.209267330169678,
      "epoch": 1.0086455331412103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906386136124476,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22396856397390366,
      "num_tokens": 24074082.0,
      "step": 10500
    },
    {
      "entropy": 5.32150444984436,
      "epoch": 1.0091258405379442,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004906288248954554,
      "loss": 4.9697,
      "mean_token_accuracy": 0.22197814285755157,
      "num_tokens": 24085050.0,
      "step": 10505
    },
    {
      "entropy": 5.264161014556885,
      "epoch": 1.0096061479346783,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004906190311722023,
      "loss": 4.938,
      "mean_token_accuracy": 0.23015548437833785,
      "num_tokens": 24095523.0,
      "step": 10510
    },
    {
      "entropy": 5.207053565979004,
      "epoch": 1.0100864553314122,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906092324429155,
      "loss": 4.9036,
      "mean_token_accuracy": 0.21892459243535994,
      "num_tokens": 24107157.0,
      "step": 10515
    },
    {
      "entropy": 5.304631280899048,
      "epoch": 1.010566762728146,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905994287078227,
      "loss": 4.9049,
      "mean_token_accuracy": 0.22785793095827103,
      "num_tokens": 24118668.0,
      "step": 10520
    },
    {
      "entropy": 5.28090009689331,
      "epoch": 1.01104707012488,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004905896199671512,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22385577261447906,
      "num_tokens": 24129563.0,
      "step": 10525
    },
    {
      "entropy": 5.362709951400757,
      "epoch": 1.0115273775216138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490579806221129,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21373135298490525,
      "num_tokens": 24139965.0,
      "step": 10530
    },
    {
      "entropy": 5.3374659538269045,
      "epoch": 1.0120076849183477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004905699874699838,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22102486491203308,
      "num_tokens": 24152109.0,
      "step": 10535
    },
    {
      "entropy": 5.257688045501709,
      "epoch": 1.0124879923150816,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004905601637139436,
      "loss": 4.9285,
      "mean_token_accuracy": 0.2160330221056938,
      "num_tokens": 24162558.0,
      "step": 10540
    },
    {
      "entropy": 5.183422565460205,
      "epoch": 1.0129682997118155,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004905503349532365,
      "loss": 4.8315,
      "mean_token_accuracy": 0.2374105155467987,
      "num_tokens": 24173963.0,
      "step": 10545
    },
    {
      "entropy": 5.246429920196533,
      "epoch": 1.0134486071085496,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905405011880906,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23231059908866883,
      "num_tokens": 24184369.0,
      "step": 10550
    },
    {
      "entropy": 5.336789512634278,
      "epoch": 1.0139289145052834,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905306624187343,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21425141245126725,
      "num_tokens": 24194692.0,
      "step": 10555
    },
    {
      "entropy": 5.306349563598633,
      "epoch": 1.0144092219020173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004905208186453961,
      "loss": 4.9549,
      "mean_token_accuracy": 0.2222945660352707,
      "num_tokens": 24205363.0,
      "step": 10560
    },
    {
      "entropy": 5.330207252502442,
      "epoch": 1.0148895292987512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905109698683044,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21622021049261092,
      "num_tokens": 24216678.0,
      "step": 10565
    },
    {
      "entropy": 5.200223541259765,
      "epoch": 1.015369836695485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905011160876878,
      "loss": 4.8535,
      "mean_token_accuracy": 0.2206648886203766,
      "num_tokens": 24227541.0,
      "step": 10570
    },
    {
      "entropy": 5.2698938846588135,
      "epoch": 1.015850144092219,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004904912573037753,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22845771312713622,
      "num_tokens": 24238118.0,
      "step": 10575
    },
    {
      "entropy": 5.286724519729614,
      "epoch": 1.0163304514889528,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004904813935167957,
      "loss": 4.99,
      "mean_token_accuracy": 0.21181708872318267,
      "num_tokens": 24250044.0,
      "step": 10580
    },
    {
      "entropy": 5.321779775619507,
      "epoch": 1.0168107588856867,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904715247269779,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21287375539541245,
      "num_tokens": 24262805.0,
      "step": 10585
    },
    {
      "entropy": 5.270455360412598,
      "epoch": 1.0172910662824208,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004904616509345514,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21490428298711778,
      "num_tokens": 24274287.0,
      "step": 10590
    },
    {
      "entropy": 5.169078636169433,
      "epoch": 1.0177713736791547,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004904517721397449,
      "loss": 4.8153,
      "mean_token_accuracy": 0.22800626754760742,
      "num_tokens": 24284839.0,
      "step": 10595
    },
    {
      "entropy": 5.241527700424195,
      "epoch": 1.0182516810758886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004904418883427881,
      "loss": 4.9239,
      "mean_token_accuracy": 0.2220812037587166,
      "num_tokens": 24295812.0,
      "step": 10600
    },
    {
      "entropy": 5.2704771041870115,
      "epoch": 1.0187319884726225,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004904319995439104,
      "loss": 4.899,
      "mean_token_accuracy": 0.2299889162182808,
      "num_tokens": 24306365.0,
      "step": 10605
    },
    {
      "entropy": 5.291464614868164,
      "epoch": 1.0192122958693564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904221057433412,
      "loss": 4.8934,
      "mean_token_accuracy": 0.2234057828783989,
      "num_tokens": 24317557.0,
      "step": 10610
    },
    {
      "entropy": 5.23344349861145,
      "epoch": 1.0196926032660902,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904122069413105,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23403415977954864,
      "num_tokens": 24328874.0,
      "step": 10615
    },
    {
      "entropy": 5.362671613693237,
      "epoch": 1.0201729106628241,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000490402303138048,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21525761783123015,
      "num_tokens": 24340918.0,
      "step": 10620
    },
    {
      "entropy": 5.290171432495117,
      "epoch": 1.0206532180595582,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004903923943337836,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2204394206404686,
      "num_tokens": 24352642.0,
      "step": 10625
    },
    {
      "entropy": 5.225270366668701,
      "epoch": 1.021133525456292,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004903824805287475,
      "loss": 4.8615,
      "mean_token_accuracy": 0.22953014224767684,
      "num_tokens": 24364874.0,
      "step": 10630
    },
    {
      "entropy": 5.22464771270752,
      "epoch": 1.021613832853026,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903725617231696,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22557385563850402,
      "num_tokens": 24376566.0,
      "step": 10635
    },
    {
      "entropy": 5.261400747299194,
      "epoch": 1.0220941402497599,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004903626379172805,
      "loss": 4.7999,
      "mean_token_accuracy": 0.23345146626234053,
      "num_tokens": 24387283.0,
      "step": 10640
    },
    {
      "entropy": 5.336586952209473,
      "epoch": 1.0225744476464937,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903527091113102,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21613558977842331,
      "num_tokens": 24397970.0,
      "step": 10645
    },
    {
      "entropy": 5.252213621139527,
      "epoch": 1.0230547550432276,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903427753054897,
      "loss": 4.9388,
      "mean_token_accuracy": 0.2184563547372818,
      "num_tokens": 24411223.0,
      "step": 10650
    },
    {
      "entropy": 5.291300868988037,
      "epoch": 1.0235350624399615,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903328365000492,
      "loss": 4.9626,
      "mean_token_accuracy": 0.22010653764009475,
      "num_tokens": 24422117.0,
      "step": 10655
    },
    {
      "entropy": 5.272748899459839,
      "epoch": 1.0240153698366954,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004903228926952199,
      "loss": 4.9162,
      "mean_token_accuracy": 0.22435684502124786,
      "num_tokens": 24434142.0,
      "step": 10660
    },
    {
      "entropy": 5.26851863861084,
      "epoch": 1.0244956772334295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004903129438912322,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2322851151227951,
      "num_tokens": 24445295.0,
      "step": 10665
    },
    {
      "entropy": 5.244488048553467,
      "epoch": 1.0249759846301634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004903029900883174,
      "loss": 4.8947,
      "mean_token_accuracy": 0.23116625994443893,
      "num_tokens": 24456092.0,
      "step": 10670
    },
    {
      "entropy": 5.240031433105469,
      "epoch": 1.0254562920268973,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902930312867063,
      "loss": 5.007,
      "mean_token_accuracy": 0.21438848972320557,
      "num_tokens": 24467653.0,
      "step": 10675
    },
    {
      "entropy": 5.302840900421143,
      "epoch": 1.0259365994236311,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004902830674866306,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2225495994091034,
      "num_tokens": 24479164.0,
      "step": 10680
    },
    {
      "entropy": 5.234032011032104,
      "epoch": 1.026416906820365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004902730986883211,
      "loss": 4.8289,
      "mean_token_accuracy": 0.2315092533826828,
      "num_tokens": 24489785.0,
      "step": 10685
    },
    {
      "entropy": 5.274800729751587,
      "epoch": 1.026897214217099,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004902631248920096,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2291500523686409,
      "num_tokens": 24500158.0,
      "step": 10690
    },
    {
      "entropy": 5.3104212284088135,
      "epoch": 1.0273775216138328,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004902531460979274,
      "loss": 5.0214,
      "mean_token_accuracy": 0.2151598334312439,
      "num_tokens": 24512851.0,
      "step": 10695
    },
    {
      "entropy": 5.27370138168335,
      "epoch": 1.0278578290105667,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004902431623063065,
      "loss": 4.9658,
      "mean_token_accuracy": 0.2228020891547203,
      "num_tokens": 24524016.0,
      "step": 10700
    },
    {
      "entropy": 5.190254259109497,
      "epoch": 1.0283381364073008,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004902331735173785,
      "loss": 4.8372,
      "mean_token_accuracy": 0.22253914773464203,
      "num_tokens": 24536348.0,
      "step": 10705
    },
    {
      "entropy": 5.226603221893311,
      "epoch": 1.0288184438040346,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004902231797313752,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22179499566555022,
      "num_tokens": 24548718.0,
      "step": 10710
    },
    {
      "entropy": 5.2236779689788815,
      "epoch": 1.0292987512007685,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902131809485288,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23917854726314544,
      "num_tokens": 24560567.0,
      "step": 10715
    },
    {
      "entropy": 5.2486608028411865,
      "epoch": 1.0297790585975024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902031771690713,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22624286711215974,
      "num_tokens": 24572610.0,
      "step": 10720
    },
    {
      "entropy": 5.261171913146972,
      "epoch": 1.0302593659942363,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004901931683932352,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22562562972307204,
      "num_tokens": 24584738.0,
      "step": 10725
    },
    {
      "entropy": 5.307566785812378,
      "epoch": 1.0307396733909702,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004901831546212526,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22423602491617203,
      "num_tokens": 24596603.0,
      "step": 10730
    },
    {
      "entropy": 5.238448238372802,
      "epoch": 1.031219980787704,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901731358533562,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22502353191375732,
      "num_tokens": 24607061.0,
      "step": 10735
    },
    {
      "entropy": 5.194182538986206,
      "epoch": 1.031700288184438,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004901631120897785,
      "loss": 4.914,
      "mean_token_accuracy": 0.22342673391103746,
      "num_tokens": 24619177.0,
      "step": 10740
    },
    {
      "entropy": 5.192566633224487,
      "epoch": 1.032180595581172,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004901530833307522,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2271680235862732,
      "num_tokens": 24631336.0,
      "step": 10745
    },
    {
      "entropy": 5.222535276412964,
      "epoch": 1.032660902977906,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901430495765103,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23529910743236543,
      "num_tokens": 24641743.0,
      "step": 10750
    },
    {
      "entropy": 5.248483800888062,
      "epoch": 1.0331412103746398,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901330108272855,
      "loss": 4.8505,
      "mean_token_accuracy": 0.22924480587244034,
      "num_tokens": 24652318.0,
      "step": 10755
    },
    {
      "entropy": 5.2008123874664305,
      "epoch": 1.0336215177713737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004901229670833111,
      "loss": 4.8571,
      "mean_token_accuracy": 0.22193673402070999,
      "num_tokens": 24664129.0,
      "step": 10760
    },
    {
      "entropy": 5.303077459335327,
      "epoch": 1.0341018251681076,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901129183448201,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21791775077581405,
      "num_tokens": 24674921.0,
      "step": 10765
    },
    {
      "entropy": 5.22287335395813,
      "epoch": 1.0345821325648414,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004901028646120459,
      "loss": 4.914,
      "mean_token_accuracy": 0.22063460499048232,
      "num_tokens": 24686052.0,
      "step": 10770
    },
    {
      "entropy": 5.230008888244629,
      "epoch": 1.0350624399615753,
      "grad_norm": 0.859375,
      "learning_rate": 0.000490092805885222,
      "loss": 4.8845,
      "mean_token_accuracy": 0.22437179386615752,
      "num_tokens": 24698625.0,
      "step": 10775
    },
    {
      "entropy": 5.222806644439697,
      "epoch": 1.0355427473583094,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004900827421645816,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2338532879948616,
      "num_tokens": 24709322.0,
      "step": 10780
    },
    {
      "entropy": 5.190197324752807,
      "epoch": 1.0360230547550433,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004900726734503589,
      "loss": 4.829,
      "mean_token_accuracy": 0.22988853305578233,
      "num_tokens": 24719512.0,
      "step": 10785
    },
    {
      "entropy": 5.211903429031372,
      "epoch": 1.0365033621517772,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004900625997427872,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22865433245897293,
      "num_tokens": 24729947.0,
      "step": 10790
    },
    {
      "entropy": 5.167942047119141,
      "epoch": 1.036983669548511,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004900525210421006,
      "loss": 4.8633,
      "mean_token_accuracy": 0.2347295880317688,
      "num_tokens": 24741423.0,
      "step": 10795
    },
    {
      "entropy": 5.264333057403564,
      "epoch": 1.037463976945245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900424373485329,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22298837453126907,
      "num_tokens": 24753403.0,
      "step": 10800
    },
    {
      "entropy": 5.3256793975830075,
      "epoch": 1.0379442843419788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004900323486623185,
      "loss": 5.0007,
      "mean_token_accuracy": 0.22211142927408217,
      "num_tokens": 24763660.0,
      "step": 10805
    },
    {
      "entropy": 5.284540367126465,
      "epoch": 1.0384245917387127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004900222549836914,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22663730531930923,
      "num_tokens": 24775061.0,
      "step": 10810
    },
    {
      "entropy": 5.263784599304199,
      "epoch": 1.0389048991354466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490012156312886,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21866078823804855,
      "num_tokens": 24785248.0,
      "step": 10815
    },
    {
      "entropy": 5.20162935256958,
      "epoch": 1.0393852065321807,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900020526501369,
      "loss": 4.8024,
      "mean_token_accuracy": 0.2338803619146347,
      "num_tokens": 24797024.0,
      "step": 10820
    },
    {
      "entropy": 5.363827180862427,
      "epoch": 1.0398655139289146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899919439956785,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21360716521739959,
      "num_tokens": 24808500.0,
      "step": 10825
    },
    {
      "entropy": 5.3585193157196045,
      "epoch": 1.0403458213256485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899818303497455,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21760460287332534,
      "num_tokens": 24818805.0,
      "step": 10830
    },
    {
      "entropy": 5.300438928604126,
      "epoch": 1.0408261287223823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004899717117125728,
      "loss": 4.9254,
      "mean_token_accuracy": 0.2188895270228386,
      "num_tokens": 24829247.0,
      "step": 10835
    },
    {
      "entropy": 5.183730602264404,
      "epoch": 1.0413064361191162,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004899615880843953,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2305304154753685,
      "num_tokens": 24840139.0,
      "step": 10840
    },
    {
      "entropy": 5.299179220199585,
      "epoch": 1.04178674351585,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004899514594654481,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21626512110233306,
      "num_tokens": 24851734.0,
      "step": 10845
    },
    {
      "entropy": 5.278581333160401,
      "epoch": 1.042267050912584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004899413258559662,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2247483789920807,
      "num_tokens": 24863424.0,
      "step": 10850
    },
    {
      "entropy": 5.249710273742676,
      "epoch": 1.0427473583093179,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899311872561849,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22249127179384232,
      "num_tokens": 24875086.0,
      "step": 10855
    },
    {
      "entropy": 5.219966316223145,
      "epoch": 1.043227665706052,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004899210436663398,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22666616588830948,
      "num_tokens": 24888408.0,
      "step": 10860
    },
    {
      "entropy": 5.19952621459961,
      "epoch": 1.0437079731027858,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004899108950866661,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23532017916440964,
      "num_tokens": 24900357.0,
      "step": 10865
    },
    {
      "entropy": 5.25680947303772,
      "epoch": 1.0441882804995197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899007415173997,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22540972828865052,
      "num_tokens": 24910790.0,
      "step": 10870
    },
    {
      "entropy": 5.246063089370727,
      "epoch": 1.0446685878962536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004898905829587762,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22390599101781844,
      "num_tokens": 24922124.0,
      "step": 10875
    },
    {
      "entropy": 5.181507158279419,
      "epoch": 1.0451488952929875,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004898804194110313,
      "loss": 4.8044,
      "mean_token_accuracy": 0.2386482760310173,
      "num_tokens": 24933591.0,
      "step": 10880
    },
    {
      "entropy": 5.237972736358643,
      "epoch": 1.0456292026897214,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004898702508744012,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2282120779156685,
      "num_tokens": 24944708.0,
      "step": 10885
    },
    {
      "entropy": 5.1982612133026125,
      "epoch": 1.0461095100864553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004898600773491221,
      "loss": 4.8049,
      "mean_token_accuracy": 0.23107136338949202,
      "num_tokens": 24955966.0,
      "step": 10890
    },
    {
      "entropy": 5.274428415298462,
      "epoch": 1.0465898174831891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004898498988354297,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22845429033041,
      "num_tokens": 24967292.0,
      "step": 10895
    },
    {
      "entropy": 5.278221225738525,
      "epoch": 1.0470701248799232,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004898397153335608,
      "loss": 4.9649,
      "mean_token_accuracy": 0.22311765402555467,
      "num_tokens": 24977407.0,
      "step": 10900
    },
    {
      "entropy": 5.357841444015503,
      "epoch": 1.0475504322766571,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004898295268437517,
      "loss": 5.0209,
      "mean_token_accuracy": 0.22102802693843843,
      "num_tokens": 24988804.0,
      "step": 10905
    },
    {
      "entropy": 5.270384407043457,
      "epoch": 1.048030739673391,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004898193333662388,
      "loss": 4.9613,
      "mean_token_accuracy": 0.2215827852487564,
      "num_tokens": 25000297.0,
      "step": 10910
    },
    {
      "entropy": 5.260179901123047,
      "epoch": 1.0485110470701249,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898091349012588,
      "loss": 4.8837,
      "mean_token_accuracy": 0.228547041118145,
      "num_tokens": 25012135.0,
      "step": 10915
    },
    {
      "entropy": 5.176730966567993,
      "epoch": 1.0489913544668588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004897989314490486,
      "loss": 4.869,
      "mean_token_accuracy": 0.2326307326555252,
      "num_tokens": 25023572.0,
      "step": 10920
    },
    {
      "entropy": 5.258727645874023,
      "epoch": 1.0494716618635926,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897887230098451,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22730273008346558,
      "num_tokens": 25035015.0,
      "step": 10925
    },
    {
      "entropy": 5.2376354217529295,
      "epoch": 1.0499519692603265,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897785095838852,
      "loss": 4.8785,
      "mean_token_accuracy": 0.2296762079000473,
      "num_tokens": 25045931.0,
      "step": 10930
    },
    {
      "entropy": 5.290501642227173,
      "epoch": 1.0504322766570606,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897682911714061,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22674432545900344,
      "num_tokens": 25056767.0,
      "step": 10935
    },
    {
      "entropy": 5.188278007507324,
      "epoch": 1.0509125840537945,
      "grad_norm": 0.90625,
      "learning_rate": 0.000489758067772645,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2343797877430916,
      "num_tokens": 25068731.0,
      "step": 10940
    },
    {
      "entropy": 5.278168725967407,
      "epoch": 1.0513928914505284,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004897478393878392,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22133503407239913,
      "num_tokens": 25081268.0,
      "step": 10945
    },
    {
      "entropy": 5.189077520370484,
      "epoch": 1.0518731988472623,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004897376060172264,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23504106402397157,
      "num_tokens": 25093105.0,
      "step": 10950
    },
    {
      "entropy": 5.204558372497559,
      "epoch": 1.0523535062439962,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897273676610438,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2320477247238159,
      "num_tokens": 25103766.0,
      "step": 10955
    },
    {
      "entropy": 5.262063694000244,
      "epoch": 1.05283381364073,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004897171243195295,
      "loss": 4.9921,
      "mean_token_accuracy": 0.22442621737718582,
      "num_tokens": 25115675.0,
      "step": 10960
    },
    {
      "entropy": 5.246082544326782,
      "epoch": 1.053314121037464,
      "grad_norm": 0.890625,
      "learning_rate": 0.000489706875992921,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22830186188220977,
      "num_tokens": 25127907.0,
      "step": 10965
    },
    {
      "entropy": 5.251580810546875,
      "epoch": 1.0537944284341978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896966226814565,
      "loss": 4.8897,
      "mean_token_accuracy": 0.2297168716788292,
      "num_tokens": 25139675.0,
      "step": 10970
    },
    {
      "entropy": 5.2940576553344725,
      "epoch": 1.054274735830932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896863643853739,
      "loss": 4.995,
      "mean_token_accuracy": 0.22323969900608062,
      "num_tokens": 25150960.0,
      "step": 10975
    },
    {
      "entropy": 5.31819224357605,
      "epoch": 1.0547550432276658,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004896761011049114,
      "loss": 4.9801,
      "mean_token_accuracy": 0.22037239670753478,
      "num_tokens": 25163676.0,
      "step": 10980
    },
    {
      "entropy": 5.2066856861114506,
      "epoch": 1.0552353506243997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896658328403074,
      "loss": 4.906,
      "mean_token_accuracy": 0.22275177836418153,
      "num_tokens": 25174317.0,
      "step": 10985
    },
    {
      "entropy": 5.205246210098267,
      "epoch": 1.0557156580211335,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004896555595918001,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22735002189874648,
      "num_tokens": 25186585.0,
      "step": 10990
    },
    {
      "entropy": 5.289089775085449,
      "epoch": 1.0561959654178674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004896452813596281,
      "loss": 4.9326,
      "mean_token_accuracy": 0.2227770507335663,
      "num_tokens": 25199014.0,
      "step": 10995
    },
    {
      "entropy": 5.314134168624878,
      "epoch": 1.0566762728146013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004896349981440301,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21744184345006942,
      "num_tokens": 25210544.0,
      "step": 11000
    },
    {
      "entropy": 5.251679992675781,
      "epoch": 1.0571565802113352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004896247099452447,
      "loss": 4.9118,
      "mean_token_accuracy": 0.23253205567598342,
      "num_tokens": 25221583.0,
      "step": 11005
    },
    {
      "entropy": 5.29381537437439,
      "epoch": 1.057636887608069,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004896144167635108,
      "loss": 4.8912,
      "mean_token_accuracy": 0.23377034217119216,
      "num_tokens": 25231724.0,
      "step": 11010
    },
    {
      "entropy": 5.290861701965332,
      "epoch": 1.0581171950048032,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896041185990675,
      "loss": 4.9927,
      "mean_token_accuracy": 0.22005500942468642,
      "num_tokens": 25243021.0,
      "step": 11015
    },
    {
      "entropy": 5.2868507385253904,
      "epoch": 1.058597502401537,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895938154521538,
      "loss": 4.9698,
      "mean_token_accuracy": 0.22316930890083314,
      "num_tokens": 25254189.0,
      "step": 11020
    },
    {
      "entropy": 5.305389165878296,
      "epoch": 1.059077809798271,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895835073230089,
      "loss": 4.9939,
      "mean_token_accuracy": 0.226844684779644,
      "num_tokens": 25265556.0,
      "step": 11025
    },
    {
      "entropy": 5.1989898681640625,
      "epoch": 1.0595581171950048,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004895731942118722,
      "loss": 4.8287,
      "mean_token_accuracy": 0.22370388060808183,
      "num_tokens": 25276789.0,
      "step": 11030
    },
    {
      "entropy": 5.278147459030151,
      "epoch": 1.0600384245917387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004895628761189829,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21584384590387345,
      "num_tokens": 25288505.0,
      "step": 11035
    },
    {
      "entropy": 5.210616397857666,
      "epoch": 1.0605187319884726,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004895525530445809,
      "loss": 4.8557,
      "mean_token_accuracy": 0.23931586146354675,
      "num_tokens": 25301490.0,
      "step": 11040
    },
    {
      "entropy": 5.234191465377807,
      "epoch": 1.0609990393852065,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895422249889057,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22233884483575822,
      "num_tokens": 25313303.0,
      "step": 11045
    },
    {
      "entropy": 5.342378234863281,
      "epoch": 1.0614793467819403,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895318919521971,
      "loss": 5.0013,
      "mean_token_accuracy": 0.21327349841594695,
      "num_tokens": 25324379.0,
      "step": 11050
    },
    {
      "entropy": 5.2233179092407225,
      "epoch": 1.0619596541786744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004895215539346949,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22978077530860902,
      "num_tokens": 25335834.0,
      "step": 11055
    },
    {
      "entropy": 5.226367568969726,
      "epoch": 1.0624399615754083,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004895112109366393,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22513044476509095,
      "num_tokens": 25347591.0,
      "step": 11060
    },
    {
      "entropy": 5.234711599349976,
      "epoch": 1.0629202689721422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895008629582703,
      "loss": 4.8848,
      "mean_token_accuracy": 0.224392306804657,
      "num_tokens": 25358483.0,
      "step": 11065
    },
    {
      "entropy": 5.222918605804443,
      "epoch": 1.063400576368876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894905099998283,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22331410497426987,
      "num_tokens": 25369434.0,
      "step": 11070
    },
    {
      "entropy": 5.255740737915039,
      "epoch": 1.06388088376561,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004894801520615535,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2277396261692047,
      "num_tokens": 25381007.0,
      "step": 11075
    },
    {
      "entropy": 5.2619866847991945,
      "epoch": 1.0643611911623438,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894697891436863,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22221987098455429,
      "num_tokens": 25393809.0,
      "step": 11080
    },
    {
      "entropy": 5.287129640579224,
      "epoch": 1.0648414985590777,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004894594212464676,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22959400564432145,
      "num_tokens": 25404967.0,
      "step": 11085
    },
    {
      "entropy": 5.2617950439453125,
      "epoch": 1.0653218059558118,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894490483701381,
      "loss": 4.8735,
      "mean_token_accuracy": 0.2364131838083267,
      "num_tokens": 25417092.0,
      "step": 11090
    },
    {
      "entropy": 5.294598245620728,
      "epoch": 1.0658021133525457,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894386705149382,
      "loss": 4.9763,
      "mean_token_accuracy": 0.21499801874160768,
      "num_tokens": 25428425.0,
      "step": 11095
    },
    {
      "entropy": 5.234216403961182,
      "epoch": 1.0662824207492796,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894282876811093,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2386219844222069,
      "num_tokens": 25440134.0,
      "step": 11100
    },
    {
      "entropy": 5.215627288818359,
      "epoch": 1.0667627281460135,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004894178998688921,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22861060202121736,
      "num_tokens": 25452222.0,
      "step": 11105
    },
    {
      "entropy": 5.251913261413574,
      "epoch": 1.0672430355427474,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004894075070785281,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2259517028927803,
      "num_tokens": 25464541.0,
      "step": 11110
    },
    {
      "entropy": 5.331733274459839,
      "epoch": 1.0677233429394812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893971093102585,
      "loss": 4.9187,
      "mean_token_accuracy": 0.21849103271961212,
      "num_tokens": 25476537.0,
      "step": 11115
    },
    {
      "entropy": 5.355682849884033,
      "epoch": 1.0682036503362151,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004893867065643245,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21579207926988603,
      "num_tokens": 25486737.0,
      "step": 11120
    },
    {
      "entropy": 5.243563938140869,
      "epoch": 1.068683957732949,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893762988409678,
      "loss": 4.9235,
      "mean_token_accuracy": 0.22094615548849106,
      "num_tokens": 25497278.0,
      "step": 11125
    },
    {
      "entropy": 5.267231082916259,
      "epoch": 1.069164265129683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004893658861404301,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22320848256349562,
      "num_tokens": 25508716.0,
      "step": 11130
    },
    {
      "entropy": 5.318556690216065,
      "epoch": 1.069644572526417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893554684629529,
      "loss": 4.9867,
      "mean_token_accuracy": 0.21895406544208526,
      "num_tokens": 25519439.0,
      "step": 11135
    },
    {
      "entropy": 5.219156265258789,
      "epoch": 1.0701248799231509,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893450458087784,
      "loss": 4.7853,
      "mean_token_accuracy": 0.2339973643422127,
      "num_tokens": 25530911.0,
      "step": 11140
    },
    {
      "entropy": 5.298265838623047,
      "epoch": 1.0706051873198847,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004893346181781483,
      "loss": 5.0125,
      "mean_token_accuracy": 0.2111084207892418,
      "num_tokens": 25542452.0,
      "step": 11145
    },
    {
      "entropy": 5.267034435272217,
      "epoch": 1.0710854947166186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004893241855713048,
      "loss": 4.9304,
      "mean_token_accuracy": 0.21265296787023544,
      "num_tokens": 25554105.0,
      "step": 11150
    },
    {
      "entropy": 5.371748208999634,
      "epoch": 1.0715658021133525,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004893137479884903,
      "loss": 4.9956,
      "mean_token_accuracy": 0.22250245362520218,
      "num_tokens": 25564806.0,
      "step": 11155
    },
    {
      "entropy": 5.215398979187012,
      "epoch": 1.0720461095100864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893033054299468,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22798677235841752,
      "num_tokens": 25575664.0,
      "step": 11160
    },
    {
      "entropy": 5.232505559921265,
      "epoch": 1.0725264169068203,
      "grad_norm": 1.09375,
      "learning_rate": 0.000489292857895917,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23367771804332732,
      "num_tokens": 25586227.0,
      "step": 11165
    },
    {
      "entropy": 5.2712836265563965,
      "epoch": 1.0730067243035544,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004892824053866432,
      "loss": 4.9869,
      "mean_token_accuracy": 0.2226012423634529,
      "num_tokens": 25597475.0,
      "step": 11170
    },
    {
      "entropy": 5.205050897598267,
      "epoch": 1.0734870317002883,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004892719479023683,
      "loss": 4.8628,
      "mean_token_accuracy": 0.23120887726545333,
      "num_tokens": 25608098.0,
      "step": 11175
    },
    {
      "entropy": 5.3060407638549805,
      "epoch": 1.0739673390970221,
      "grad_norm": 1.0,
      "learning_rate": 0.000489261485443335,
      "loss": 4.983,
      "mean_token_accuracy": 0.21792962700128554,
      "num_tokens": 25620053.0,
      "step": 11180
    },
    {
      "entropy": 5.287705707550049,
      "epoch": 1.074447646493756,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892510180097863,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22067270129919053,
      "num_tokens": 25630534.0,
      "step": 11185
    },
    {
      "entropy": 5.293157434463501,
      "epoch": 1.07492795389049,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004892405456019651,
      "loss": 4.9457,
      "mean_token_accuracy": 0.2270430624485016,
      "num_tokens": 25641413.0,
      "step": 11190
    },
    {
      "entropy": 5.20750880241394,
      "epoch": 1.0754082612872238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004892300682201147,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23234288692474364,
      "num_tokens": 25652081.0,
      "step": 11195
    },
    {
      "entropy": 5.2989599227905275,
      "epoch": 1.0758885686839577,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004892195858644782,
      "loss": 4.96,
      "mean_token_accuracy": 0.22862013876438142,
      "num_tokens": 25664282.0,
      "step": 11200
    },
    {
      "entropy": 5.295254325866699,
      "epoch": 1.0763688760806915,
      "grad_norm": 0.984375,
      "learning_rate": 0.000489209098535299,
      "loss": 5.0294,
      "mean_token_accuracy": 0.21955449730157853,
      "num_tokens": 25675310.0,
      "step": 11205
    },
    {
      "entropy": 5.364900398254394,
      "epoch": 1.0768491834774256,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891986062328205,
      "loss": 5.0151,
      "mean_token_accuracy": 0.22300100177526475,
      "num_tokens": 25686895.0,
      "step": 11210
    },
    {
      "entropy": 5.261716699600219,
      "epoch": 1.0773294908741595,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004891881089572865,
      "loss": 4.8142,
      "mean_token_accuracy": 0.22860780954360962,
      "num_tokens": 25697778.0,
      "step": 11215
    },
    {
      "entropy": 5.217953538894653,
      "epoch": 1.0778097982708934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891776067089406,
      "loss": 4.9539,
      "mean_token_accuracy": 0.2138178676366806,
      "num_tokens": 25708602.0,
      "step": 11220
    },
    {
      "entropy": 5.242786598205567,
      "epoch": 1.0782901056676273,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004891670994880266,
      "loss": 4.855,
      "mean_token_accuracy": 0.22719742059707643,
      "num_tokens": 25719671.0,
      "step": 11225
    },
    {
      "entropy": 5.341251707077026,
      "epoch": 1.0787704130643612,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004891565872947888,
      "loss": 5.018,
      "mean_token_accuracy": 0.2188740625977516,
      "num_tokens": 25731797.0,
      "step": 11230
    },
    {
      "entropy": 5.228974342346191,
      "epoch": 1.079250720461095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004891460701294706,
      "loss": 4.8879,
      "mean_token_accuracy": 0.23104191869497298,
      "num_tokens": 25743984.0,
      "step": 11235
    },
    {
      "entropy": 5.178681135177612,
      "epoch": 1.079731027857829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004891355479923167,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2300044924020767,
      "num_tokens": 25755252.0,
      "step": 11240
    },
    {
      "entropy": 5.184259462356567,
      "epoch": 1.080211335254563,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004891250208835712,
      "loss": 4.8056,
      "mean_token_accuracy": 0.2283560663461685,
      "num_tokens": 25765715.0,
      "step": 11245
    },
    {
      "entropy": 5.317605352401733,
      "epoch": 1.080691642651297,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891144888034784,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22156814038753508,
      "num_tokens": 25777866.0,
      "step": 11250
    },
    {
      "entropy": 5.257466983795166,
      "epoch": 1.0811719500480308,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004891039517522832,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22886780947446822,
      "num_tokens": 25791199.0,
      "step": 11255
    },
    {
      "entropy": 5.160863161087036,
      "epoch": 1.0816522574447647,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004890934097302299,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2412295773625374,
      "num_tokens": 25802979.0,
      "step": 11260
    },
    {
      "entropy": 5.1385103225708,
      "epoch": 1.0821325648414986,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004890828627375632,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22546306103467942,
      "num_tokens": 25814696.0,
      "step": 11265
    },
    {
      "entropy": 5.28546257019043,
      "epoch": 1.0826128722382324,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004890723107745283,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21339473724365235,
      "num_tokens": 25825376.0,
      "step": 11270
    },
    {
      "entropy": 5.291554737091064,
      "epoch": 1.0830931796349663,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004890617538413699,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22395792454481125,
      "num_tokens": 25835491.0,
      "step": 11275
    },
    {
      "entropy": 5.279152345657349,
      "epoch": 1.0835734870317002,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004890511919383333,
      "loss": 4.9562,
      "mean_token_accuracy": 0.2216746136546135,
      "num_tokens": 25848154.0,
      "step": 11280
    },
    {
      "entropy": 5.224534606933593,
      "epoch": 1.084053794428434,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004890406250656636,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22763658910989762,
      "num_tokens": 25859471.0,
      "step": 11285
    },
    {
      "entropy": 5.227497577667236,
      "epoch": 1.0845341018251682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004890300532236062,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22896967828273773,
      "num_tokens": 25869460.0,
      "step": 11290
    },
    {
      "entropy": 5.255751848220825,
      "epoch": 1.085014409221902,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004890194764124064,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23003373444080352,
      "num_tokens": 25881892.0,
      "step": 11295
    },
    {
      "entropy": 5.284740686416626,
      "epoch": 1.085494716618636,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004890088946323099,
      "loss": 4.9876,
      "mean_token_accuracy": 0.21472073644399642,
      "num_tokens": 25893774.0,
      "step": 11300
    },
    {
      "entropy": 5.202382373809814,
      "epoch": 1.0859750240153698,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004889983078835623,
      "loss": 4.805,
      "mean_token_accuracy": 0.23075173795223236,
      "num_tokens": 25904758.0,
      "step": 11305
    },
    {
      "entropy": 5.222729444503784,
      "epoch": 1.0864553314121037,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004889877161664096,
      "loss": 4.9119,
      "mean_token_accuracy": 0.2221290424466133,
      "num_tokens": 25917013.0,
      "step": 11310
    },
    {
      "entropy": 5.358620548248291,
      "epoch": 1.0869356388088376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889771194810974,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22622882276773454,
      "num_tokens": 25927780.0,
      "step": 11315
    },
    {
      "entropy": 5.213243627548218,
      "epoch": 1.0874159462055715,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889665178278719,
      "loss": 4.8567,
      "mean_token_accuracy": 0.232503642141819,
      "num_tokens": 25939339.0,
      "step": 11320
    },
    {
      "entropy": 5.239845132827758,
      "epoch": 1.0878962536023056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004889559112069792,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22841319888830186,
      "num_tokens": 25950440.0,
      "step": 11325
    },
    {
      "entropy": 5.264793491363525,
      "epoch": 1.0883765609990395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004889452996186657,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2217724710702896,
      "num_tokens": 25962849.0,
      "step": 11330
    },
    {
      "entropy": 5.277474021911621,
      "epoch": 1.0888568683957733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004889346830631774,
      "loss": 4.9357,
      "mean_token_accuracy": 0.2179882138967514,
      "num_tokens": 25973616.0,
      "step": 11335
    },
    {
      "entropy": 5.24298791885376,
      "epoch": 1.0893371757925072,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000488924061540761,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23404917269945144,
      "num_tokens": 25984727.0,
      "step": 11340
    },
    {
      "entropy": 5.262585687637329,
      "epoch": 1.089817483189241,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889134350516633,
      "loss": 4.9344,
      "mean_token_accuracy": 0.232413649559021,
      "num_tokens": 25996431.0,
      "step": 11345
    },
    {
      "entropy": 5.283836746215821,
      "epoch": 1.090297790585975,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889028035961308,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22559195756912231,
      "num_tokens": 26008936.0,
      "step": 11350
    },
    {
      "entropy": 5.338031148910522,
      "epoch": 1.0907780979827089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888921671744103,
      "loss": 4.9659,
      "mean_token_accuracy": 0.21589563488960267,
      "num_tokens": 26019308.0,
      "step": 11355
    },
    {
      "entropy": 5.268791151046753,
      "epoch": 1.0912584053794427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004888815257867488,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23384385108947753,
      "num_tokens": 26030705.0,
      "step": 11360
    },
    {
      "entropy": 5.261903476715088,
      "epoch": 1.0917387127761768,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004888708794333934,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2257483884692192,
      "num_tokens": 26042759.0,
      "step": 11365
    },
    {
      "entropy": 5.264266586303711,
      "epoch": 1.0922190201729107,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004888602281145913,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2276270568370819,
      "num_tokens": 26054719.0,
      "step": 11370
    },
    {
      "entropy": 5.243171834945679,
      "epoch": 1.0926993275696446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004888495718305897,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2174038767814636,
      "num_tokens": 26065765.0,
      "step": 11375
    },
    {
      "entropy": 5.152737092971802,
      "epoch": 1.0931796349663785,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488838910581636,
      "loss": 4.8333,
      "mean_token_accuracy": 0.228248430788517,
      "num_tokens": 26077719.0,
      "step": 11380
    },
    {
      "entropy": 5.312730216979981,
      "epoch": 1.0936599423631124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004888282443679777,
      "loss": 4.9773,
      "mean_token_accuracy": 0.21503383815288543,
      "num_tokens": 26089924.0,
      "step": 11385
    },
    {
      "entropy": 5.389603757858277,
      "epoch": 1.0941402497598463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004888175731898627,
      "loss": 5.0433,
      "mean_token_accuracy": 0.21609580963850022,
      "num_tokens": 26100312.0,
      "step": 11390
    },
    {
      "entropy": 5.243638277053833,
      "epoch": 1.0946205571565801,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004888068970475384,
      "loss": 4.9379,
      "mean_token_accuracy": 0.22016473710536957,
      "num_tokens": 26111932.0,
      "step": 11395
    },
    {
      "entropy": 5.270091867446899,
      "epoch": 1.0951008645533142,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004887962159412529,
      "loss": 4.848,
      "mean_token_accuracy": 0.2285095825791359,
      "num_tokens": 26123989.0,
      "step": 11400
    },
    {
      "entropy": 5.328083324432373,
      "epoch": 1.0955811719500481,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887855298712541,
      "loss": 4.9588,
      "mean_token_accuracy": 0.2235036998987198,
      "num_tokens": 26135589.0,
      "step": 11405
    },
    {
      "entropy": 5.1671537399292,
      "epoch": 1.096061479346782,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048877483883779,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22971481829881668,
      "num_tokens": 26147069.0,
      "step": 11410
    },
    {
      "entropy": 5.232102155685425,
      "epoch": 1.0965417867435159,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004887641428411091,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22444375306367875,
      "num_tokens": 26159331.0,
      "step": 11415
    },
    {
      "entropy": 5.292403268814087,
      "epoch": 1.0970220941402498,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004887534418814595,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22335042506456376,
      "num_tokens": 26169863.0,
      "step": 11420
    },
    {
      "entropy": 5.2318507671356205,
      "epoch": 1.0975024015369836,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004887427359590897,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22035748213529588,
      "num_tokens": 26182888.0,
      "step": 11425
    },
    {
      "entropy": 5.2402245044708256,
      "epoch": 1.0979827089337175,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004887320250742482,
      "loss": 4.8945,
      "mean_token_accuracy": 0.23108669072389604,
      "num_tokens": 26194979.0,
      "step": 11430
    },
    {
      "entropy": 5.206297588348389,
      "epoch": 1.0984630163304514,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004887213092271838,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23503292351961136,
      "num_tokens": 26207309.0,
      "step": 11435
    },
    {
      "entropy": 5.232964611053466,
      "epoch": 1.0989433237271853,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004887105884181451,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22167981863021852,
      "num_tokens": 26219231.0,
      "step": 11440
    },
    {
      "entropy": 5.325558423995972,
      "epoch": 1.0994236311239194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886998626473813,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22034439891576768,
      "num_tokens": 26229355.0,
      "step": 11445
    },
    {
      "entropy": 5.289998388290405,
      "epoch": 1.0999039385206533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886891319151411,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23249689042568206,
      "num_tokens": 26239069.0,
      "step": 11450
    },
    {
      "entropy": 5.289576053619385,
      "epoch": 1.1003842459173871,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886783962216738,
      "loss": 4.991,
      "mean_token_accuracy": 0.22010688334703446,
      "num_tokens": 26250403.0,
      "step": 11455
    },
    {
      "entropy": 5.30884370803833,
      "epoch": 1.100864553314121,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004886676555672287,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2255308985710144,
      "num_tokens": 26262926.0,
      "step": 11460
    },
    {
      "entropy": 5.352726078033447,
      "epoch": 1.101344860710855,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886569099520551,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21332792341709136,
      "num_tokens": 26274030.0,
      "step": 11465
    },
    {
      "entropy": 5.215978574752808,
      "epoch": 1.1018251681075888,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004886461593764024,
      "loss": 4.9111,
      "mean_token_accuracy": 0.23133158385753633,
      "num_tokens": 26284799.0,
      "step": 11470
    },
    {
      "entropy": 5.126020431518555,
      "epoch": 1.1023054755043227,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004886354038405204,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2370050370693207,
      "num_tokens": 26295154.0,
      "step": 11475
    },
    {
      "entropy": 5.291110277175903,
      "epoch": 1.1027857829010568,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886246433446586,
      "loss": 4.9298,
      "mean_token_accuracy": 0.21937124729156493,
      "num_tokens": 26306181.0,
      "step": 11480
    },
    {
      "entropy": 5.3846197605133055,
      "epoch": 1.1032660902977907,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886138778890669,
      "loss": 5.0081,
      "mean_token_accuracy": 0.21463223397731782,
      "num_tokens": 26318674.0,
      "step": 11485
    },
    {
      "entropy": 5.335513257980347,
      "epoch": 1.1037463976945245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004886031074739953,
      "loss": 4.9417,
      "mean_token_accuracy": 0.2109510198235512,
      "num_tokens": 26330257.0,
      "step": 11490
    },
    {
      "entropy": 5.264688491821289,
      "epoch": 1.1042267050912584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004885923320996938,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2251171126961708,
      "num_tokens": 26342035.0,
      "step": 11495
    },
    {
      "entropy": 5.261469888687134,
      "epoch": 1.1047070124879923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885815517664127,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2246149256825447,
      "num_tokens": 26352703.0,
      "step": 11500
    },
    {
      "entropy": 5.302320957183838,
      "epoch": 1.1051873198847262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488570766474402,
      "loss": 4.9367,
      "mean_token_accuracy": 0.22746830582618713,
      "num_tokens": 26362324.0,
      "step": 11505
    },
    {
      "entropy": 5.297534322738647,
      "epoch": 1.10566762728146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004885599762239124,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21410372108221054,
      "num_tokens": 26373540.0,
      "step": 11510
    },
    {
      "entropy": 5.244863557815552,
      "epoch": 1.106147934678194,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004885491810151943,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22253721952438354,
      "num_tokens": 26384259.0,
      "step": 11515
    },
    {
      "entropy": 5.199947309494019,
      "epoch": 1.106628242074928,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004885383808484982,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22784973531961442,
      "num_tokens": 26395193.0,
      "step": 11520
    },
    {
      "entropy": 5.213193607330322,
      "epoch": 1.107108549471662,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004885275757240751,
      "loss": 4.84,
      "mean_token_accuracy": 0.22606273591518403,
      "num_tokens": 26408556.0,
      "step": 11525
    },
    {
      "entropy": 5.235076475143432,
      "epoch": 1.1075888568683958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004885167656421757,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2248816281557083,
      "num_tokens": 26420066.0,
      "step": 11530
    },
    {
      "entropy": 5.230745935440064,
      "epoch": 1.1080691642651297,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488505950603051,
      "loss": 4.8903,
      "mean_token_accuracy": 0.2255407989025116,
      "num_tokens": 26432533.0,
      "step": 11535
    },
    {
      "entropy": 5.232820320129394,
      "epoch": 1.1085494716618636,
      "grad_norm": 1.109375,
      "learning_rate": 0.000488495130606952,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22794847041368485,
      "num_tokens": 26443878.0,
      "step": 11540
    },
    {
      "entropy": 5.286282634735107,
      "epoch": 1.1090297790585975,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004884843056541302,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2211151286959648,
      "num_tokens": 26455111.0,
      "step": 11545
    },
    {
      "entropy": 5.258453893661499,
      "epoch": 1.1095100864553313,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004884734757448367,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22411396354436874,
      "num_tokens": 26466577.0,
      "step": 11550
    },
    {
      "entropy": 5.239039278030395,
      "epoch": 1.1099903938520654,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000488462640879323,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22907019406557083,
      "num_tokens": 26479637.0,
      "step": 11555
    },
    {
      "entropy": 5.212899684906006,
      "epoch": 1.1104707012487993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004884518010578405,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2268904745578766,
      "num_tokens": 26489923.0,
      "step": 11560
    },
    {
      "entropy": 5.283234930038452,
      "epoch": 1.1109510086455332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884409562806411,
      "loss": 4.9852,
      "mean_token_accuracy": 0.2160591796040535,
      "num_tokens": 26501236.0,
      "step": 11565
    },
    {
      "entropy": 5.244903326034546,
      "epoch": 1.111431316042267,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004884301065479765,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22698431164026261,
      "num_tokens": 26511509.0,
      "step": 11570
    },
    {
      "entropy": 5.16862301826477,
      "epoch": 1.111911623439001,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004884192518600986,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23667658120393753,
      "num_tokens": 26520931.0,
      "step": 11575
    },
    {
      "entropy": 5.181065940856934,
      "epoch": 1.1123919308357348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004884083922172593,
      "loss": 4.8313,
      "mean_token_accuracy": 0.23801038414239883,
      "num_tokens": 26531333.0,
      "step": 11580
    },
    {
      "entropy": 5.203880310058594,
      "epoch": 1.1128722382324687,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004883975276197108,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2308025598526001,
      "num_tokens": 26543696.0,
      "step": 11585
    },
    {
      "entropy": 5.258868551254272,
      "epoch": 1.1133525456292026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004883866580677055,
      "loss": 4.8515,
      "mean_token_accuracy": 0.23199667185544967,
      "num_tokens": 26556292.0,
      "step": 11590
    },
    {
      "entropy": 5.273058319091797,
      "epoch": 1.1138328530259365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004883757835614956,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2188606083393097,
      "num_tokens": 26568023.0,
      "step": 11595
    },
    {
      "entropy": 5.1909692764282225,
      "epoch": 1.1143131604226706,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004883649041013335,
      "loss": 4.881,
      "mean_token_accuracy": 0.22893589735031128,
      "num_tokens": 26579469.0,
      "step": 11600
    },
    {
      "entropy": 5.339939546585083,
      "epoch": 1.1147934678194045,
      "grad_norm": 1.015625,
      "learning_rate": 0.000488354019687472,
      "loss": 4.9226,
      "mean_token_accuracy": 0.2237112268805504,
      "num_tokens": 26591592.0,
      "step": 11605
    },
    {
      "entropy": 5.281352519989014,
      "epoch": 1.1152737752161384,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004883431303201636,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2264537200331688,
      "num_tokens": 26604075.0,
      "step": 11610
    },
    {
      "entropy": 5.163708019256592,
      "epoch": 1.1157540826128722,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883322359996613,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23240430355072023,
      "num_tokens": 26616448.0,
      "step": 11615
    },
    {
      "entropy": 5.236646556854248,
      "epoch": 1.1162343900096061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004883213367262179,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23349083364009857,
      "num_tokens": 26628749.0,
      "step": 11620
    },
    {
      "entropy": 5.25415997505188,
      "epoch": 1.11671469740634,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883104325000866,
      "loss": 4.853,
      "mean_token_accuracy": 0.23319512754678726,
      "num_tokens": 26640163.0,
      "step": 11625
    },
    {
      "entropy": 5.266146850585938,
      "epoch": 1.1171950048030739,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004882995233215203,
      "loss": 4.9055,
      "mean_token_accuracy": 0.2249453768134117,
      "num_tokens": 26650729.0,
      "step": 11630
    },
    {
      "entropy": 5.220187854766846,
      "epoch": 1.117675312199808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004882886091907726,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22713975459337235,
      "num_tokens": 26661995.0,
      "step": 11635
    },
    {
      "entropy": 5.230469131469727,
      "epoch": 1.1181556195965419,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048827769010809666,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23094468265771867,
      "num_tokens": 26673224.0,
      "step": 11640
    },
    {
      "entropy": 5.265944719314575,
      "epoch": 1.1186359269932757,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048826676607374606,
      "loss": 4.8795,
      "mean_token_accuracy": 0.21974054276943206,
      "num_tokens": 26686331.0,
      "step": 11645
    },
    {
      "entropy": 5.346266746520996,
      "epoch": 1.1191162343900096,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048825583708797434,
      "loss": 4.9245,
      "mean_token_accuracy": 0.22869265377521514,
      "num_tokens": 26696816.0,
      "step": 11650
    },
    {
      "entropy": 5.233740711212159,
      "epoch": 1.1195965417867435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004882449031510354,
      "loss": 4.884,
      "mean_token_accuracy": 0.2303701013326645,
      "num_tokens": 26708126.0,
      "step": 11655
    },
    {
      "entropy": 5.319062805175781,
      "epoch": 1.1200768491834774,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000488233964263183,
      "loss": 4.9947,
      "mean_token_accuracy": 0.2256132334470749,
      "num_tokens": 26718951.0,
      "step": 11660
    },
    {
      "entropy": 5.2943941116333,
      "epoch": 1.1205571565802113,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048822302042467115,
      "loss": 4.9612,
      "mean_token_accuracy": 0.21830258518457413,
      "num_tokens": 26730550.0,
      "step": 11665
    },
    {
      "entropy": 5.1769775390625,
      "epoch": 1.1210374639769451,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004882120716357539,
      "loss": 4.784,
      "mean_token_accuracy": 0.23301633447408676,
      "num_tokens": 26741485.0,
      "step": 11670
    },
    {
      "entropy": 5.151652574539185,
      "epoch": 1.1215177713736793,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004882011178966854,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23126445859670638,
      "num_tokens": 26753947.0,
      "step": 11675
    },
    {
      "entropy": 5.2239611625671385,
      "epoch": 1.1219980787704131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004881901592077201,
      "loss": 4.8507,
      "mean_token_accuracy": 0.2377290904521942,
      "num_tokens": 26764921.0,
      "step": 11680
    },
    {
      "entropy": 5.252098941802979,
      "epoch": 1.122478386167147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004881791955691123,
      "loss": 4.8273,
      "mean_token_accuracy": 0.2292026996612549,
      "num_tokens": 26777384.0,
      "step": 11685
    },
    {
      "entropy": 5.255926132202148,
      "epoch": 1.122958693563881,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048816822698111655,
      "loss": 4.9438,
      "mean_token_accuracy": 0.22267890870571136,
      "num_tokens": 26789916.0,
      "step": 11690
    },
    {
      "entropy": 5.233464622497559,
      "epoch": 1.1234390009606148,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048815725344398766,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22951712608337402,
      "num_tokens": 26801167.0,
      "step": 11695
    },
    {
      "entropy": 5.3065758228302,
      "epoch": 1.1239193083573487,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048814627495798017,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22306582778692247,
      "num_tokens": 26813235.0,
      "step": 11700
    },
    {
      "entropy": 5.308640480041504,
      "epoch": 1.1243996157540825,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004881352915233492,
      "loss": 4.9403,
      "mean_token_accuracy": 0.2272568464279175,
      "num_tokens": 26824758.0,
      "step": 11705
    },
    {
      "entropy": 5.2614795684814455,
      "epoch": 1.1248799231508164,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048812430314034956,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22273175567388534,
      "num_tokens": 26836877.0,
      "step": 11710
    },
    {
      "entropy": 5.163290023803711,
      "epoch": 1.1253602305475505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004881133098092365,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23946761190891266,
      "num_tokens": 26848394.0,
      "step": 11715
    },
    {
      "entropy": 5.1863288402557375,
      "epoch": 1.1258405379442844,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004881023115302652,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22795673161745073,
      "num_tokens": 26859064.0,
      "step": 11720
    },
    {
      "entropy": 5.326487922668457,
      "epoch": 1.1263208453410183,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488091308303691,
      "loss": 5.0177,
      "mean_token_accuracy": 0.20961594432592393,
      "num_tokens": 26870401.0,
      "step": 11725
    },
    {
      "entropy": 5.293287754058838,
      "epoch": 1.1268011527377522,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004880803001297694,
      "loss": 4.8975,
      "mean_token_accuracy": 0.22611468732357026,
      "num_tokens": 26881767.0,
      "step": 11730
    },
    {
      "entropy": 5.207429265975952,
      "epoch": 1.127281460134486,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004880692870087559,
      "loss": 4.8075,
      "mean_token_accuracy": 0.22658294290304185,
      "num_tokens": 26892854.0,
      "step": 11735
    },
    {
      "entropy": 5.326664876937866,
      "epoch": 1.12776176753122,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048805826894090626,
      "loss": 4.9742,
      "mean_token_accuracy": 0.2178993284702301,
      "num_tokens": 26905079.0,
      "step": 11740
    },
    {
      "entropy": 5.351021432876587,
      "epoch": 1.1282420749279538,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048804724592647626,
      "loss": 5.006,
      "mean_token_accuracy": 0.21630996465682983,
      "num_tokens": 26915640.0,
      "step": 11745
    },
    {
      "entropy": 5.237456655502319,
      "epoch": 1.1287223823246877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004880362179657218,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2279524624347687,
      "num_tokens": 26927039.0,
      "step": 11750
    },
    {
      "entropy": 5.234236383438111,
      "epoch": 1.1292026897214218,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048802518505889904,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2271624818444252,
      "num_tokens": 26939368.0,
      "step": 11755
    },
    {
      "entropy": 5.237351179122925,
      "epoch": 1.1296829971181557,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048801414720626404,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20983987897634507,
      "num_tokens": 26952051.0,
      "step": 11760
    },
    {
      "entropy": 5.223706579208374,
      "epoch": 1.1301633045148896,
      "grad_norm": 1.0,
      "learning_rate": 0.00048800310440807294,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22461564540863038,
      "num_tokens": 26963680.0,
      "step": 11765
    },
    {
      "entropy": 5.188236951828003,
      "epoch": 1.1306436119116234,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004879920566645823,
      "loss": 4.7675,
      "mean_token_accuracy": 0.23430795967578888,
      "num_tokens": 26973899.0,
      "step": 11770
    },
    {
      "entropy": 5.226254081726074,
      "epoch": 1.1311239193083573,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004879810039760486,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2183161422610283,
      "num_tokens": 26985692.0,
      "step": 11775
    },
    {
      "entropy": 5.268548631668091,
      "epoch": 1.1316042267050912,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004879699463427284,
      "loss": 4.8469,
      "mean_token_accuracy": 0.23019198179244996,
      "num_tokens": 26996354.0,
      "step": 11780
    },
    {
      "entropy": 5.292991399765015,
      "epoch": 1.132084534101825,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004879588837648785,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22216602563858032,
      "num_tokens": 27007925.0,
      "step": 11785
    },
    {
      "entropy": 5.18213119506836,
      "epoch": 1.1325648414985592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048794781624275554,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23503227680921554,
      "num_tokens": 27019222.0,
      "step": 11790
    },
    {
      "entropy": 5.159797048568725,
      "epoch": 1.133045148895293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048793674377661664,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23339744210243224,
      "num_tokens": 27029973.0,
      "step": 11795
    },
    {
      "entropy": 5.3194386005401615,
      "epoch": 1.133525456292027,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048792566636671886,
      "loss": 4.9972,
      "mean_token_accuracy": 0.22451780438423158,
      "num_tokens": 27042002.0,
      "step": 11800
    },
    {
      "entropy": 5.263101720809937,
      "epoch": 1.1340057636887608,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004879145840133194,
      "loss": 4.8993,
      "mean_token_accuracy": 0.23220417648553848,
      "num_tokens": 27052770.0,
      "step": 11805
    },
    {
      "entropy": 5.26320161819458,
      "epoch": 1.1344860710854947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004879034967166755,
      "loss": 4.9325,
      "mean_token_accuracy": 0.22670873254537582,
      "num_tokens": 27063375.0,
      "step": 11810
    },
    {
      "entropy": 5.202679204940796,
      "epoch": 1.1349663784822286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004878924044770446,
      "loss": 4.8754,
      "mean_token_accuracy": 0.2284083142876625,
      "num_tokens": 27075908.0,
      "step": 11815
    },
    {
      "entropy": 5.247083282470703,
      "epoch": 1.1354466858789625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004878813072946843,
      "loss": 4.9198,
      "mean_token_accuracy": 0.22700789868831633,
      "num_tokens": 27087590.0,
      "step": 11820
    },
    {
      "entropy": 5.254173612594604,
      "epoch": 1.1359269932756964,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048787020516985203,
      "loss": 4.8397,
      "mean_token_accuracy": 0.23286038488149643,
      "num_tokens": 27098076.0,
      "step": 11825
    },
    {
      "entropy": 5.217121601104736,
      "epoch": 1.1364073006724302,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048785909810280576,
      "loss": 4.8469,
      "mean_token_accuracy": 0.2364824041724205,
      "num_tokens": 27109851.0,
      "step": 11830
    },
    {
      "entropy": 5.219685983657837,
      "epoch": 1.1368876080691643,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004878479860938033,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23431069999933243,
      "num_tokens": 27121288.0,
      "step": 11835
    },
    {
      "entropy": 5.230327129364014,
      "epoch": 1.1373679154658982,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048783686914310266,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2252708852291107,
      "num_tokens": 27133877.0,
      "step": 11840
    },
    {
      "entropy": 5.105851125717163,
      "epoch": 1.137848222862632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004878257472509619,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24098218977451324,
      "num_tokens": 27145616.0,
      "step": 11845
    },
    {
      "entropy": 5.155728530883789,
      "epoch": 1.138328530259366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004878146204176392,
      "loss": 4.851,
      "mean_token_accuracy": 0.22704108208417892,
      "num_tokens": 27157182.0,
      "step": 11850
    },
    {
      "entropy": 5.282539892196655,
      "epoch": 1.1388088376560999,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000487803488643393,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2355527713894844,
      "num_tokens": 27167609.0,
      "step": 11855
    },
    {
      "entropy": 5.316936397552491,
      "epoch": 1.1392891450528337,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048779235192848166,
      "loss": 4.9688,
      "mean_token_accuracy": 0.2221071556210518,
      "num_tokens": 27180043.0,
      "step": 11860
    },
    {
      "entropy": 5.167005109786987,
      "epoch": 1.1397694524495678,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004877812102731638,
      "loss": 4.7787,
      "mean_token_accuracy": 0.23531264215707778,
      "num_tokens": 27191870.0,
      "step": 11865
    },
    {
      "entropy": 5.255460786819458,
      "epoch": 1.1402497598463017,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048777006367769804,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22846313267946244,
      "num_tokens": 27203291.0,
      "step": 11870
    },
    {
      "entropy": 5.2547942161560055,
      "epoch": 1.1407300672430356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004877589121423432,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22856287509202958,
      "num_tokens": 27214607.0,
      "step": 11875
    },
    {
      "entropy": 5.221647453308106,
      "epoch": 1.1412103746397695,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004877477556673582,
      "loss": 4.7942,
      "mean_token_accuracy": 0.23272537738084792,
      "num_tokens": 27224658.0,
      "step": 11880
    },
    {
      "entropy": 5.19312801361084,
      "epoch": 1.1416906820365034,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000487736594253002,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22440452128648758,
      "num_tokens": 27235306.0,
      "step": 11885
    },
    {
      "entropy": 5.280982494354248,
      "epoch": 1.1421709894332372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048772542789953384,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22780535817146302,
      "num_tokens": 27246679.0,
      "step": 11890
    },
    {
      "entropy": 5.3368322372436525,
      "epoch": 1.1426512968299711,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048771425660721284,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21749826222658158,
      "num_tokens": 27257799.0,
      "step": 11895
    },
    {
      "entropy": 5.220373678207397,
      "epoch": 1.143131604226705,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048770308037629853,
      "loss": 4.9208,
      "mean_token_accuracy": 0.22285577952861785,
      "num_tokens": 27268651.0,
      "step": 11900
    },
    {
      "entropy": 5.212119436264038,
      "epoch": 1.143611911623439,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004876918992070502,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2291291117668152,
      "num_tokens": 27280655.0,
      "step": 11905
    },
    {
      "entropy": 5.309642267227173,
      "epoch": 1.144092219020173,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004876807130997276,
      "loss": 4.9609,
      "mean_token_accuracy": 0.22083690464496614,
      "num_tokens": 27292664.0,
      "step": 11910
    },
    {
      "entropy": 5.260783433914185,
      "epoch": 1.1445725264169069,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004876695220545903,
      "loss": 4.9254,
      "mean_token_accuracy": 0.22563991397619249,
      "num_tokens": 27306143.0,
      "step": 11915
    },
    {
      "entropy": 5.235976982116699,
      "epoch": 1.1450528338136408,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048765832607189824,
      "loss": 4.818,
      "mean_token_accuracy": 0.23104555159807205,
      "num_tokens": 27316926.0,
      "step": 11920
    },
    {
      "entropy": 5.339642763137817,
      "epoch": 1.1455331412103746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048764712515191136,
      "loss": 4.94,
      "mean_token_accuracy": 0.2277303606271744,
      "num_tokens": 27327472.0,
      "step": 11925
    },
    {
      "entropy": 5.31283745765686,
      "epoch": 1.1460134486071085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048763591929488966,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2293119341135025,
      "num_tokens": 27338388.0,
      "step": 11930
    },
    {
      "entropy": 5.228753757476807,
      "epoch": 1.1464937560038424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004876247085010933,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22910197377204894,
      "num_tokens": 27349879.0,
      "step": 11935
    },
    {
      "entropy": 5.22907018661499,
      "epoch": 1.1469740634005763,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048761349277078253,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23338637948036195,
      "num_tokens": 27361226.0,
      "step": 11940
    },
    {
      "entropy": 5.237310695648193,
      "epoch": 1.1474543707973104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048760227210421775,
      "loss": 4.82,
      "mean_token_accuracy": 0.23125654309988022,
      "num_tokens": 27373512.0,
      "step": 11945
    },
    {
      "entropy": 5.171060943603516,
      "epoch": 1.1479346781940443,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004875910465016596,
      "loss": 4.8741,
      "mean_token_accuracy": 0.23239589631557464,
      "num_tokens": 27384950.0,
      "step": 11950
    },
    {
      "entropy": 5.350432443618774,
      "epoch": 1.1484149855907781,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004875798159633686,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22146188467741013,
      "num_tokens": 27395877.0,
      "step": 11955
    },
    {
      "entropy": 5.278548526763916,
      "epoch": 1.148895292987512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004875685804896055,
      "loss": 4.9718,
      "mean_token_accuracy": 0.22865064293146134,
      "num_tokens": 27407783.0,
      "step": 11960
    },
    {
      "entropy": 5.2086889266967775,
      "epoch": 1.149375600384246,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004875573400806312,
      "loss": 4.862,
      "mean_token_accuracy": 0.22700949013233185,
      "num_tokens": 27420598.0,
      "step": 11965
    },
    {
      "entropy": 5.262592887878418,
      "epoch": 1.1498559077809798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048754609473670654,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22339701354503633,
      "num_tokens": 27431961.0,
      "step": 11970
    },
    {
      "entropy": 5.240012741088867,
      "epoch": 1.1503362151777137,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004875348444580927,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23154822289943694,
      "num_tokens": 27444010.0,
      "step": 11975
    },
    {
      "entropy": 5.26583571434021,
      "epoch": 1.1508165225744476,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000487523589245051,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2296842008829117,
      "num_tokens": 27455576.0,
      "step": 11980
    },
    {
      "entropy": 5.228506135940552,
      "epoch": 1.1512968299711814,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004875123290978425,
      "loss": 4.8921,
      "mean_token_accuracy": 0.22496830374002458,
      "num_tokens": 27468957.0,
      "step": 11985
    },
    {
      "entropy": 5.23967170715332,
      "epoch": 1.1517771373679155,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048750106401672876,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2303827092051506,
      "num_tokens": 27479992.0,
      "step": 11990
    },
    {
      "entropy": 5.316565418243409,
      "epoch": 1.1522574447646494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048748979400197134,
      "loss": 4.9501,
      "mean_token_accuracy": 0.22238598614931107,
      "num_tokens": 27490878.0,
      "step": 11995
    },
    {
      "entropy": 5.193116998672485,
      "epoch": 1.1527377521613833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048747851905383183,
      "loss": 4.7884,
      "mean_token_accuracy": 0.2362989977002144,
      "num_tokens": 27502009.0,
      "step": 12000
    },
    {
      "epoch": 1.1527377521613833,
      "eval_entropy": 5.038512420910036,
      "eval_loss": 4.977294921875,
      "eval_mean_token_accuracy": 0.23085267144799784,
      "eval_num_tokens": 27502009.0,
      "eval_runtime": 20.4494,
      "eval_samples_per_second": 1604.696,
      "eval_steps_per_second": 200.593,
      "step": 12000
    },
    {
      "entropy": 5.24969973564148,
      "epoch": 1.1532180595581172,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004874672391725721,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21951764523983003,
      "num_tokens": 27513376.0,
      "step": 12005
    },
    {
      "entropy": 5.200477838516235,
      "epoch": 1.153698366954851,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874559543584539,
      "loss": 4.8354,
      "mean_token_accuracy": 0.2266063928604126,
      "num_tokens": 27525166.0,
      "step": 12010
    },
    {
      "entropy": 5.175713348388672,
      "epoch": 1.154178674351585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004874446646117394,
      "loss": 4.7592,
      "mean_token_accuracy": 0.2451384887099266,
      "num_tokens": 27535994.0,
      "step": 12015
    },
    {
      "entropy": 5.20946102142334,
      "epoch": 1.154658981748319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004874333699326906,
      "loss": 4.8459,
      "mean_token_accuracy": 0.2289209097623825,
      "num_tokens": 27546883.0,
      "step": 12020
    },
    {
      "entropy": 5.308850383758545,
      "epoch": 1.155139289145053,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004874220703215697,
      "loss": 4.9771,
      "mean_token_accuracy": 0.22082567512989043,
      "num_tokens": 27558840.0,
      "step": 12025
    },
    {
      "entropy": 5.247687911987304,
      "epoch": 1.1556195965417868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004874107657786391,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22178890258073808,
      "num_tokens": 27569776.0,
      "step": 12030
    },
    {
      "entropy": 5.2705670356750485,
      "epoch": 1.1560999039385207,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048739945630416124,
      "loss": 4.8286,
      "mean_token_accuracy": 0.23441084027290343,
      "num_tokens": 27580097.0,
      "step": 12035
    },
    {
      "entropy": 5.299001884460449,
      "epoch": 1.1565802113352546,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004873881418983987,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22369632869958878,
      "num_tokens": 27592035.0,
      "step": 12040
    },
    {
      "entropy": 5.267014789581299,
      "epoch": 1.1570605187319885,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004873768225616141,
      "loss": 4.9215,
      "mean_token_accuracy": 0.22802343368530273,
      "num_tokens": 27604409.0,
      "step": 12045
    },
    {
      "entropy": 5.259671926498413,
      "epoch": 1.1575408261287223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048736549829407047,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22798727601766586,
      "num_tokens": 27615019.0,
      "step": 12050
    },
    {
      "entropy": 5.322322082519531,
      "epoch": 1.1580211335254562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004873541690960305,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22090902030467988,
      "num_tokens": 27627924.0,
      "step": 12055
    },
    {
      "entropy": 5.220972156524658,
      "epoch": 1.15850144092219,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004873428349677573,
      "loss": 4.8003,
      "mean_token_accuracy": 0.22881386131048204,
      "num_tokens": 27639188.0,
      "step": 12060
    },
    {
      "entropy": 5.318940687179565,
      "epoch": 1.1589817483189242,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000487331495909514,
      "loss": 5.0121,
      "mean_token_accuracy": 0.2153981953859329,
      "num_tokens": 27652621.0,
      "step": 12065
    },
    {
      "entropy": 5.296951389312744,
      "epoch": 1.159462055715658,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048732015192156383,
      "loss": 4.9466,
      "mean_token_accuracy": 0.21933044940233232,
      "num_tokens": 27665726.0,
      "step": 12070
    },
    {
      "entropy": 5.261237335205078,
      "epoch": 1.159942363112392,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048730880300417015,
      "loss": 4.912,
      "mean_token_accuracy": 0.2217643141746521,
      "num_tokens": 27676984.0,
      "step": 12075
    },
    {
      "entropy": 5.285104894638062,
      "epoch": 1.1604226705091258,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048729744915759657,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22514020651578903,
      "num_tokens": 27688812.0,
      "step": 12080
    },
    {
      "entropy": 5.166171073913574,
      "epoch": 1.1609029779058597,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048728609038210655,
      "loss": 4.7674,
      "mean_token_accuracy": 0.23518061339855195,
      "num_tokens": 27699152.0,
      "step": 12085
    },
    {
      "entropy": 5.257433938980102,
      "epoch": 1.1613832853025936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048727472667796395,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22507894933223724,
      "num_tokens": 27710347.0,
      "step": 12090
    },
    {
      "entropy": 5.1937000274658205,
      "epoch": 1.1618635926993275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004872633580454325,
      "loss": 4.837,
      "mean_token_accuracy": 0.23211393803358077,
      "num_tokens": 27721112.0,
      "step": 12095
    },
    {
      "entropy": 5.268783330917358,
      "epoch": 1.1623439000960616,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048725198448477616,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22265601605176927,
      "num_tokens": 27731766.0,
      "step": 12100
    },
    {
      "entropy": 5.206387138366699,
      "epoch": 1.1628242074927955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048724060599625893,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23101575970649718,
      "num_tokens": 27743718.0,
      "step": 12105
    },
    {
      "entropy": 5.281051063537598,
      "epoch": 1.1633045148895294,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048722922258014506,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22674826234579087,
      "num_tokens": 27754999.0,
      "step": 12110
    },
    {
      "entropy": 5.311943292617798,
      "epoch": 1.1637848222862632,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004872178342366989,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21628783643245697,
      "num_tokens": 27767684.0,
      "step": 12115
    },
    {
      "entropy": 5.224699974060059,
      "epoch": 1.1642651296829971,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048720644096618475,
      "loss": 4.8881,
      "mean_token_accuracy": 0.22770496755838393,
      "num_tokens": 27779103.0,
      "step": 12120
    },
    {
      "entropy": 5.219120645523072,
      "epoch": 1.164745437079731,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004871950427688672,
      "loss": 4.9162,
      "mean_token_accuracy": 0.23188477009534836,
      "num_tokens": 27791223.0,
      "step": 12125
    },
    {
      "entropy": 5.2410088062286375,
      "epoch": 1.1652257444764649,
      "grad_norm": 1.0,
      "learning_rate": 0.00048718363964501087,
      "loss": 4.8364,
      "mean_token_accuracy": 0.23128343671560286,
      "num_tokens": 27802660.0,
      "step": 12130
    },
    {
      "entropy": 5.309169864654541,
      "epoch": 1.1657060518731988,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004871722315948805,
      "loss": 4.9197,
      "mean_token_accuracy": 0.2205914944410324,
      "num_tokens": 27814370.0,
      "step": 12135
    },
    {
      "entropy": 5.230017852783203,
      "epoch": 1.1661863592699326,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004871608186187408,
      "loss": 4.8277,
      "mean_token_accuracy": 0.234578412771225,
      "num_tokens": 27825035.0,
      "step": 12140
    },
    {
      "entropy": 5.149127340316772,
      "epoch": 1.1666666666666667,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048714940071685703,
      "loss": 4.7782,
      "mean_token_accuracy": 0.2336522564291954,
      "num_tokens": 27837276.0,
      "step": 12145
    },
    {
      "entropy": 5.326910400390625,
      "epoch": 1.1671469740634006,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048713797788949405,
      "loss": 4.961,
      "mean_token_accuracy": 0.22117829769849778,
      "num_tokens": 27848618.0,
      "step": 12150
    },
    {
      "entropy": 5.182396173477173,
      "epoch": 1.1676272814601345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048712655013691714,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22152728140354155,
      "num_tokens": 27861556.0,
      "step": 12155
    },
    {
      "entropy": 5.267713451385498,
      "epoch": 1.1681075888568684,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048711511745939165,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22337129563093186,
      "num_tokens": 27873211.0,
      "step": 12160
    },
    {
      "entropy": 5.276123046875,
      "epoch": 1.1685878962536023,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000487103679857183,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2314464658498764,
      "num_tokens": 27883365.0,
      "step": 12165
    },
    {
      "entropy": 5.125260639190674,
      "epoch": 1.1690682036503361,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004870922373305567,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23140246272087098,
      "num_tokens": 27894669.0,
      "step": 12170
    },
    {
      "entropy": 5.217029714584351,
      "epoch": 1.1695485110470702,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048708078987977837,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22900984138250352,
      "num_tokens": 27906480.0,
      "step": 12175
    },
    {
      "entropy": 5.230155420303345,
      "epoch": 1.1700288184438041,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048706933750511394,
      "loss": 4.8549,
      "mean_token_accuracy": 0.23662770837545394,
      "num_tokens": 27918383.0,
      "step": 12180
    },
    {
      "entropy": 5.312373876571655,
      "epoch": 1.170509125840538,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004870578802068292,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22901555448770522,
      "num_tokens": 27928944.0,
      "step": 12185
    },
    {
      "entropy": 5.1691266059875485,
      "epoch": 1.170989433237272,
      "grad_norm": 0.875,
      "learning_rate": 0.00048704641798519006,
      "loss": 4.8214,
      "mean_token_accuracy": 0.22792013883590698,
      "num_tokens": 27941105.0,
      "step": 12190
    },
    {
      "entropy": 5.228753900527954,
      "epoch": 1.1714697406340058,
      "grad_norm": 1.0,
      "learning_rate": 0.00048703495084046286,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2313213974237442,
      "num_tokens": 27952925.0,
      "step": 12195
    },
    {
      "entropy": 5.199661540985107,
      "epoch": 1.1719500480307397,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004870234787729137,
      "loss": 4.8543,
      "mean_token_accuracy": 0.2347971275448799,
      "num_tokens": 27965504.0,
      "step": 12200
    },
    {
      "entropy": 5.195229530334473,
      "epoch": 1.1724303554274735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004870120017828089,
      "loss": 4.8363,
      "mean_token_accuracy": 0.23301661014556885,
      "num_tokens": 27975985.0,
      "step": 12205
    },
    {
      "entropy": 5.35689754486084,
      "epoch": 1.1729106628242074,
      "grad_norm": 0.984375,
      "learning_rate": 0.000487000519870415,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2112668976187706,
      "num_tokens": 27987784.0,
      "step": 12210
    },
    {
      "entropy": 5.2515922546386715,
      "epoch": 1.1733909702209413,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004869890330359986,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22466825395822526,
      "num_tokens": 27999623.0,
      "step": 12215
    },
    {
      "entropy": 5.250215244293213,
      "epoch": 1.1738712776176754,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004869775412798262,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22351137846708297,
      "num_tokens": 28011561.0,
      "step": 12220
    },
    {
      "entropy": 5.272357606887818,
      "epoch": 1.1743515850144093,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048696604460216476,
      "loss": 4.9163,
      "mean_token_accuracy": 0.23218585401773453,
      "num_tokens": 28022108.0,
      "step": 12225
    },
    {
      "entropy": 5.273452281951904,
      "epoch": 1.1748318924111432,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048695454300328123,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21990286707878112,
      "num_tokens": 28033497.0,
      "step": 12230
    },
    {
      "entropy": 5.310796928405762,
      "epoch": 1.175312199807877,
      "grad_norm": 1.0,
      "learning_rate": 0.00048694303648344256,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22657402902841567,
      "num_tokens": 28044790.0,
      "step": 12235
    },
    {
      "entropy": 5.219969987869263,
      "epoch": 1.175792507204611,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048693152504291595,
      "loss": 4.8675,
      "mean_token_accuracy": 0.2293829694390297,
      "num_tokens": 28056000.0,
      "step": 12240
    },
    {
      "entropy": 5.2074648380279545,
      "epoch": 1.1762728146013448,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004869200086819686,
      "loss": 4.8865,
      "mean_token_accuracy": 0.22641449570655822,
      "num_tokens": 28068264.0,
      "step": 12245
    },
    {
      "entropy": 5.274576568603516,
      "epoch": 1.1767531219980787,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048690848740086796,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23296197056770324,
      "num_tokens": 28080220.0,
      "step": 12250
    },
    {
      "entropy": 5.285139513015747,
      "epoch": 1.1772334293948128,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004868969611998814,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23126309365034103,
      "num_tokens": 28091531.0,
      "step": 12255
    },
    {
      "entropy": 5.221493482589722,
      "epoch": 1.1777137367915467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004868854300792767,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22601612508296967,
      "num_tokens": 28102982.0,
      "step": 12260
    },
    {
      "entropy": 5.245751619338989,
      "epoch": 1.1781940441882806,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048687389403932144,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2264834851026535,
      "num_tokens": 28114131.0,
      "step": 12265
    },
    {
      "entropy": 5.238775682449341,
      "epoch": 1.1786743515850144,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004868623530802835,
      "loss": 4.843,
      "mean_token_accuracy": 0.2272892639040947,
      "num_tokens": 28125637.0,
      "step": 12270
    },
    {
      "entropy": 5.32103796005249,
      "epoch": 1.1791546589817483,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048685080720243086,
      "loss": 4.9001,
      "mean_token_accuracy": 0.23009310513734818,
      "num_tokens": 28137284.0,
      "step": 12275
    },
    {
      "entropy": 5.269804000854492,
      "epoch": 1.1796349663784822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004868392564060315,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23387300223112106,
      "num_tokens": 28148909.0,
      "step": 12280
    },
    {
      "entropy": 5.275269460678101,
      "epoch": 1.180115273775216,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004868277006913537,
      "loss": 4.901,
      "mean_token_accuracy": 0.23067858219146728,
      "num_tokens": 28160392.0,
      "step": 12285
    },
    {
      "entropy": 5.272730541229248,
      "epoch": 1.18059558117195,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004868161400586656,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21420984864234924,
      "num_tokens": 28172957.0,
      "step": 12290
    },
    {
      "entropy": 5.378000974655151,
      "epoch": 1.1810758885686838,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004868045745082357,
      "loss": 4.9534,
      "mean_token_accuracy": 0.22122086435556412,
      "num_tokens": 28183239.0,
      "step": 12295
    },
    {
      "entropy": 5.227760982513428,
      "epoch": 1.181556195965418,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004867930040403326,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2293841779232025,
      "num_tokens": 28195382.0,
      "step": 12300
    },
    {
      "entropy": 5.21633505821228,
      "epoch": 1.1820365033621518,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048678142865522475,
      "loss": 4.9329,
      "mean_token_accuracy": 0.22046357691287993,
      "num_tokens": 28206645.0,
      "step": 12305
    },
    {
      "entropy": 5.289435291290284,
      "epoch": 1.1825168107588857,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000486769848353181,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2272959381341934,
      "num_tokens": 28218420.0,
      "step": 12310
    },
    {
      "entropy": 5.329785966873169,
      "epoch": 1.1829971181556196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048675826313447027,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22620476335287093,
      "num_tokens": 28229458.0,
      "step": 12315
    },
    {
      "entropy": 5.277183628082275,
      "epoch": 1.1834774255523535,
      "grad_norm": 1.0,
      "learning_rate": 0.00048674667299936135,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22461260259151458,
      "num_tokens": 28242240.0,
      "step": 12320
    },
    {
      "entropy": 5.380187225341797,
      "epoch": 1.1839577329490873,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048673507794812356,
      "loss": 5.0065,
      "mean_token_accuracy": 0.216791932284832,
      "num_tokens": 28254597.0,
      "step": 12325
    },
    {
      "entropy": 5.28867359161377,
      "epoch": 1.1844380403458212,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004867234779810259,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22054771780967714,
      "num_tokens": 28266674.0,
      "step": 12330
    },
    {
      "entropy": 5.230553150177002,
      "epoch": 1.1849183477425553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004867118730983378,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22803467363119126,
      "num_tokens": 28278671.0,
      "step": 12335
    },
    {
      "entropy": 5.29219388961792,
      "epoch": 1.1853986551392892,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004867002633003286,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23344966173171997,
      "num_tokens": 28291085.0,
      "step": 12340
    },
    {
      "entropy": 5.235214948654175,
      "epoch": 1.185878962536023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000486688648587268,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23029283285140992,
      "num_tokens": 28303457.0,
      "step": 12345
    },
    {
      "entropy": 5.256423044204712,
      "epoch": 1.186359269932757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866770289594256,
      "loss": 4.9258,
      "mean_token_accuracy": 0.2240041419863701,
      "num_tokens": 28314599.0,
      "step": 12350
    },
    {
      "entropy": 5.370691013336182,
      "epoch": 1.1868395773294909,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048666540441707107,
      "loss": 5.0002,
      "mean_token_accuracy": 0.22366588264703752,
      "num_tokens": 28326266.0,
      "step": 12355
    },
    {
      "entropy": 5.270044374465942,
      "epoch": 1.1873198847262247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004866537749604744,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22437173426151275,
      "num_tokens": 28337238.0,
      "step": 12360
    },
    {
      "entropy": 5.186470794677734,
      "epoch": 1.1878001921229586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048664214058990546,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2318959966301918,
      "num_tokens": 28348060.0,
      "step": 12365
    },
    {
      "entropy": 5.239836740493774,
      "epoch": 1.1882804995196925,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004866305013056346,
      "loss": 4.9071,
      "mean_token_accuracy": 0.22301804423332214,
      "num_tokens": 28359442.0,
      "step": 12370
    },
    {
      "entropy": 5.311597871780395,
      "epoch": 1.1887608069164266,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004866188571079318,
      "loss": 4.9383,
      "mean_token_accuracy": 0.22664508372545242,
      "num_tokens": 28370827.0,
      "step": 12375
    },
    {
      "entropy": 5.327049589157104,
      "epoch": 1.1892411143131605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866072079970676,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21761683672666549,
      "num_tokens": 28383115.0,
      "step": 12380
    },
    {
      "entropy": 5.2556510925292965,
      "epoch": 1.1897214217098944,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048659555397331236,
      "loss": 4.8647,
      "mean_token_accuracy": 0.23380745202302933,
      "num_tokens": 28394904.0,
      "step": 12385
    },
    {
      "entropy": 5.240615320205689,
      "epoch": 1.1902017291066282,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004865838950369366,
      "loss": 4.904,
      "mean_token_accuracy": 0.2215904861688614,
      "num_tokens": 28407357.0,
      "step": 12390
    },
    {
      "entropy": 5.239888763427734,
      "epoch": 1.1906820365033621,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048657223118821116,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23132071495056153,
      "num_tokens": 28418088.0,
      "step": 12395
    },
    {
      "entropy": 5.1296289443969725,
      "epoch": 1.191162343900096,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048656056242740665,
      "loss": 4.7661,
      "mean_token_accuracy": 0.23616935312747955,
      "num_tokens": 28430022.0,
      "step": 12400
    },
    {
      "entropy": 5.254455804824829,
      "epoch": 1.19164265129683,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004865488887547942,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22510038167238236,
      "num_tokens": 28440530.0,
      "step": 12405
    },
    {
      "entropy": 5.301360368728638,
      "epoch": 1.192122958693564,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004865372101706446,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2224045291543007,
      "num_tokens": 28452707.0,
      "step": 12410
    },
    {
      "entropy": 5.159613800048828,
      "epoch": 1.1926032660902979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004865255266752292,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23615646213293076,
      "num_tokens": 28465131.0,
      "step": 12415
    },
    {
      "entropy": 5.184052276611328,
      "epoch": 1.1930835734870318,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004865138382688191,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23025415688753129,
      "num_tokens": 28476136.0,
      "step": 12420
    },
    {
      "entropy": 5.256198930740356,
      "epoch": 1.1935638808837656,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004865021449516859,
      "loss": 4.8157,
      "mean_token_accuracy": 0.22713397145271302,
      "num_tokens": 28488374.0,
      "step": 12425
    },
    {
      "entropy": 5.234571123123169,
      "epoch": 1.1940441882804995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004864904467241008,
      "loss": 4.858,
      "mean_token_accuracy": 0.2224708989262581,
      "num_tokens": 28499585.0,
      "step": 12430
    },
    {
      "entropy": 5.164283657073975,
      "epoch": 1.1945244956772334,
      "grad_norm": 1.0,
      "learning_rate": 0.00048647874358633556,
      "loss": 4.7999,
      "mean_token_accuracy": 0.22950491458177566,
      "num_tokens": 28510707.0,
      "step": 12435
    },
    {
      "entropy": 5.323966693878174,
      "epoch": 1.1950048030739673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048646703553866183,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22271956503391266,
      "num_tokens": 28522398.0,
      "step": 12440
    },
    {
      "entropy": 5.257711744308471,
      "epoch": 1.1954851104707012,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004864553225813515,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22629288882017135,
      "num_tokens": 28532949.0,
      "step": 12445
    },
    {
      "entropy": 5.160993432998657,
      "epoch": 1.195965417867435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864436047146765,
      "loss": 4.8039,
      "mean_token_accuracy": 0.23674183189868928,
      "num_tokens": 28544292.0,
      "step": 12450
    },
    {
      "entropy": 5.265891647338867,
      "epoch": 1.1964457252641691,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048643188193890874,
      "loss": 4.9575,
      "mean_token_accuracy": 0.22365434169769288,
      "num_tokens": 28556430.0,
      "step": 12455
    },
    {
      "entropy": 5.288662719726562,
      "epoch": 1.196926032660903,
      "grad_norm": 1.0,
      "learning_rate": 0.0004864201542543206,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21955835372209548,
      "num_tokens": 28568475.0,
      "step": 12460
    },
    {
      "entropy": 5.285154962539673,
      "epoch": 1.197406340057637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864084216611843,
      "loss": 4.8831,
      "mean_token_accuracy": 0.22876946330070497,
      "num_tokens": 28579653.0,
      "step": 12465
    },
    {
      "entropy": 5.268446063995361,
      "epoch": 1.1978866474543708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048639668415977207,
      "loss": 4.8334,
      "mean_token_accuracy": 0.2224608138203621,
      "num_tokens": 28590108.0,
      "step": 12470
    },
    {
      "entropy": 5.250721502304077,
      "epoch": 1.1983669548511047,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048638494175035665,
      "loss": 4.85,
      "mean_token_accuracy": 0.23002714514732361,
      "num_tokens": 28602008.0,
      "step": 12475
    },
    {
      "entropy": 5.205025768280029,
      "epoch": 1.1988472622478386,
      "grad_norm": 1.125,
      "learning_rate": 0.0004863731944332105,
      "loss": 4.766,
      "mean_token_accuracy": 0.23381862342357634,
      "num_tokens": 28613286.0,
      "step": 12480
    },
    {
      "entropy": 5.215325832366943,
      "epoch": 1.1993275696445724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004863614422086065,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23262013047933577,
      "num_tokens": 28625497.0,
      "step": 12485
    },
    {
      "entropy": 5.244168138504028,
      "epoch": 1.1998078770413065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004863496850768174,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23009678423404695,
      "num_tokens": 28638046.0,
      "step": 12490
    },
    {
      "entropy": 5.166541051864624,
      "epoch": 1.2002881844380404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004863379230381162,
      "loss": 4.7842,
      "mean_token_accuracy": 0.22876994907855988,
      "num_tokens": 28647923.0,
      "step": 12495
    },
    {
      "entropy": 5.22263503074646,
      "epoch": 1.2007684918347743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000486326156092776,
      "loss": 4.8326,
      "mean_token_accuracy": 0.2350337475538254,
      "num_tokens": 28660023.0,
      "step": 12500
    },
    {
      "entropy": 5.229833364486694,
      "epoch": 1.2012487992315082,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048631438424106985,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22737979888916016,
      "num_tokens": 28671568.0,
      "step": 12505
    },
    {
      "entropy": 5.172854280471801,
      "epoch": 1.201729106628242,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048630260748327124,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2369781032204628,
      "num_tokens": 28682897.0,
      "step": 12510
    },
    {
      "entropy": 5.286155319213867,
      "epoch": 1.202209414024976,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048629082581965355,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22599581331014634,
      "num_tokens": 28694067.0,
      "step": 12515
    },
    {
      "entropy": 5.246472454071045,
      "epoch": 1.2026897214217098,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048627903925049033,
      "loss": 4.8458,
      "mean_token_accuracy": 0.22567730396986008,
      "num_tokens": 28705738.0,
      "step": 12520
    },
    {
      "entropy": 5.167621994018555,
      "epoch": 1.2031700288184437,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048626724777605507,
      "loss": 4.79,
      "mean_token_accuracy": 0.2307256892323494,
      "num_tokens": 28717419.0,
      "step": 12525
    },
    {
      "entropy": 5.2821362018585205,
      "epoch": 1.2036503362151778,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004862554513966217,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21714217066764832,
      "num_tokens": 28728587.0,
      "step": 12530
    },
    {
      "entropy": 5.332984590530396,
      "epoch": 1.2041306436119117,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048624365011246405,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21826449334621428,
      "num_tokens": 28740818.0,
      "step": 12535
    },
    {
      "entropy": 5.204153394699096,
      "epoch": 1.2046109510086456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004862318439238561,
      "loss": 4.8001,
      "mean_token_accuracy": 0.2392582878470421,
      "num_tokens": 28751936.0,
      "step": 12540
    },
    {
      "entropy": 5.227561569213867,
      "epoch": 1.2050912584053795,
      "grad_norm": 1.015625,
      "learning_rate": 0.000486220032831072,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22605605870485307,
      "num_tokens": 28763271.0,
      "step": 12545
    },
    {
      "entropy": 5.3240439891815186,
      "epoch": 1.2055715658021133,
      "grad_norm": 1.0,
      "learning_rate": 0.0004862082168343859,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2303387328982353,
      "num_tokens": 28774282.0,
      "step": 12550
    },
    {
      "entropy": 5.264748287200928,
      "epoch": 1.2060518731988472,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004861963959340722,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22252289354801177,
      "num_tokens": 28785826.0,
      "step": 12555
    },
    {
      "entropy": 5.168811988830567,
      "epoch": 1.206532180595581,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004861845701304053,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23391988277435302,
      "num_tokens": 28797669.0,
      "step": 12560
    },
    {
      "entropy": 5.271500873565674,
      "epoch": 1.2070124879923152,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048617273942365977,
      "loss": 4.9423,
      "mean_token_accuracy": 0.222794246673584,
      "num_tokens": 28808438.0,
      "step": 12565
    },
    {
      "entropy": 5.212741422653198,
      "epoch": 1.207492795389049,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004861609038141103,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23277053982019424,
      "num_tokens": 28819707.0,
      "step": 12570
    },
    {
      "entropy": 5.246720552444458,
      "epoch": 1.207973102785783,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048614906330203165,
      "loss": 4.9058,
      "mean_token_accuracy": 0.22340347915887832,
      "num_tokens": 28831829.0,
      "step": 12575
    },
    {
      "entropy": 5.208948040008545,
      "epoch": 1.2084534101825168,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004861372178876987,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22928400188684464,
      "num_tokens": 28843429.0,
      "step": 12580
    },
    {
      "entropy": 5.210789060592651,
      "epoch": 1.2089337175792507,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048612536757138653,
      "loss": 4.808,
      "mean_token_accuracy": 0.23279453068971634,
      "num_tokens": 28856239.0,
      "step": 12585
    },
    {
      "entropy": 5.275742387771606,
      "epoch": 1.2094140249759846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004861135123533702,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22826707512140273,
      "num_tokens": 28868599.0,
      "step": 12590
    },
    {
      "entropy": 5.22515459060669,
      "epoch": 1.2098943323727185,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048610165223392503,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22514620870351792,
      "num_tokens": 28880234.0,
      "step": 12595
    },
    {
      "entropy": 5.3046191215515135,
      "epoch": 1.2103746397694524,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004860897872133263,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2256513476371765,
      "num_tokens": 28892198.0,
      "step": 12600
    },
    {
      "entropy": 5.2720794677734375,
      "epoch": 1.2108549471661862,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004860779172918496,
      "loss": 4.9692,
      "mean_token_accuracy": 0.22152229100465776,
      "num_tokens": 28904153.0,
      "step": 12605
    },
    {
      "entropy": 5.310212898254394,
      "epoch": 1.2113352545629203,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004860660424697704,
      "loss": 4.9206,
      "mean_token_accuracy": 0.2235573187470436,
      "num_tokens": 28915284.0,
      "step": 12610
    },
    {
      "entropy": 5.227004909515381,
      "epoch": 1.2118155619596542,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048605416274736434,
      "loss": 4.881,
      "mean_token_accuracy": 0.2266954869031906,
      "num_tokens": 28928168.0,
      "step": 12615
    },
    {
      "entropy": 5.184287786483765,
      "epoch": 1.2122958693563881,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048604227812490744,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2337518021464348,
      "num_tokens": 28938548.0,
      "step": 12620
    },
    {
      "entropy": 5.22723422050476,
      "epoch": 1.212776176753122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048603038860267546,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23100861459970473,
      "num_tokens": 28949252.0,
      "step": 12625
    },
    {
      "entropy": 5.269889259338379,
      "epoch": 1.2132564841498559,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004860184941809445,
      "loss": 4.8844,
      "mean_token_accuracy": 0.23126914203166962,
      "num_tokens": 28960758.0,
      "step": 12630
    },
    {
      "entropy": 5.258989906311035,
      "epoch": 1.2137367915465898,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048600659485999073,
      "loss": 4.8517,
      "mean_token_accuracy": 0.22889953553676606,
      "num_tokens": 28972604.0,
      "step": 12635
    },
    {
      "entropy": 5.250363540649414,
      "epoch": 1.2142170989433236,
      "grad_norm": 1.0,
      "learning_rate": 0.00048599469064009027,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2316880017518997,
      "num_tokens": 28983617.0,
      "step": 12640
    },
    {
      "entropy": 5.241665315628052,
      "epoch": 1.2146974063400577,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048598278152151974,
      "loss": 4.891,
      "mean_token_accuracy": 0.2280999541282654,
      "num_tokens": 28994277.0,
      "step": 12645
    },
    {
      "entropy": 5.289932584762573,
      "epoch": 1.2151777137367916,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004859708675045555,
      "loss": 4.9332,
      "mean_token_accuracy": 0.22959504425525665,
      "num_tokens": 29004844.0,
      "step": 12650
    },
    {
      "entropy": 5.154257011413574,
      "epoch": 1.2156580211335255,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004859589485894741,
      "loss": 4.797,
      "mean_token_accuracy": 0.23411002457141877,
      "num_tokens": 29015978.0,
      "step": 12655
    },
    {
      "entropy": 5.319645595550537,
      "epoch": 1.2161383285302594,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004859470247765524,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2253017395734787,
      "num_tokens": 29026966.0,
      "step": 12660
    },
    {
      "entropy": 5.27032241821289,
      "epoch": 1.2166186359269933,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004859350960660671,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23118204325437547,
      "num_tokens": 29037943.0,
      "step": 12665
    },
    {
      "entropy": 5.280789089202881,
      "epoch": 1.2170989433237271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004859231624582953,
      "loss": 4.9033,
      "mean_token_accuracy": 0.21823128461837768,
      "num_tokens": 29049939.0,
      "step": 12670
    },
    {
      "entropy": 5.191416072845459,
      "epoch": 1.217579250720461,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048591122395351394,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22699238508939742,
      "num_tokens": 29062442.0,
      "step": 12675
    },
    {
      "entropy": 5.191320514678955,
      "epoch": 1.218059558117195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004858992805520003,
      "loss": 4.8758,
      "mean_token_accuracy": 0.22837745398283005,
      "num_tokens": 29074240.0,
      "step": 12680
    },
    {
      "entropy": 5.32074761390686,
      "epoch": 1.218539865513929,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048588733225403153,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22534438520669936,
      "num_tokens": 29085550.0,
      "step": 12685
    },
    {
      "entropy": 5.228748321533203,
      "epoch": 1.219020172910663,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004858753790598851,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2293204441666603,
      "num_tokens": 29097092.0,
      "step": 12690
    },
    {
      "entropy": 5.209425735473633,
      "epoch": 1.2195004803073968,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004858634209698386,
      "loss": 4.878,
      "mean_token_accuracy": 0.23070379793643953,
      "num_tokens": 29108119.0,
      "step": 12695
    },
    {
      "entropy": 5.233037662506104,
      "epoch": 1.2199807877041307,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048585145798416956,
      "loss": 4.8577,
      "mean_token_accuracy": 0.2260055348277092,
      "num_tokens": 29120605.0,
      "step": 12700
    },
    {
      "entropy": 5.257398462295532,
      "epoch": 1.2204610951008645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004858394901031558,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22487132102251053,
      "num_tokens": 29131582.0,
      "step": 12705
    },
    {
      "entropy": 5.293846321105957,
      "epoch": 1.2209414024975984,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004858275173270751,
      "loss": 4.892,
      "mean_token_accuracy": 0.22256153225898742,
      "num_tokens": 29143436.0,
      "step": 12710
    },
    {
      "entropy": 5.227873945236206,
      "epoch": 1.2214217098943323,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048581553965620553,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23221997767686844,
      "num_tokens": 29154445.0,
      "step": 12715
    },
    {
      "entropy": 5.277367353439331,
      "epoch": 1.2219020172910664,
      "grad_norm": 1.0,
      "learning_rate": 0.00048580355709082506,
      "loss": 4.9239,
      "mean_token_accuracy": 0.23011967837810515,
      "num_tokens": 29164599.0,
      "step": 12720
    },
    {
      "entropy": 5.213503456115722,
      "epoch": 1.2223823246878003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000485791569631212,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22499535232782364,
      "num_tokens": 29176275.0,
      "step": 12725
    },
    {
      "entropy": 5.303771924972534,
      "epoch": 1.2228626320845342,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004857795772776446,
      "loss": 4.865,
      "mean_token_accuracy": 0.226776422560215,
      "num_tokens": 29189102.0,
      "step": 12730
    },
    {
      "entropy": 5.2937071323394775,
      "epoch": 1.223342939481268,
      "grad_norm": 1.0,
      "learning_rate": 0.00048576758003040127,
      "loss": 4.9196,
      "mean_token_accuracy": 0.22070501148700714,
      "num_tokens": 29200953.0,
      "step": 12735
    },
    {
      "entropy": 5.245256233215332,
      "epoch": 1.223823246878002,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048575557788976066,
      "loss": 4.9171,
      "mean_token_accuracy": 0.21777532398700714,
      "num_tokens": 29212942.0,
      "step": 12740
    },
    {
      "entropy": 5.164741325378418,
      "epoch": 1.2243035542747358,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004857435708560013,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23270945847034455,
      "num_tokens": 29224949.0,
      "step": 12745
    },
    {
      "entropy": 5.347541618347168,
      "epoch": 1.2247838616714697,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048573155892940204,
      "loss": 4.9645,
      "mean_token_accuracy": 0.22189656943082808,
      "num_tokens": 29236044.0,
      "step": 12750
    },
    {
      "entropy": 5.250304412841797,
      "epoch": 1.2252641690682036,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048571954211024164,
      "loss": 4.8653,
      "mean_token_accuracy": 0.2273030161857605,
      "num_tokens": 29248084.0,
      "step": 12755
    },
    {
      "entropy": 5.241931056976318,
      "epoch": 1.2257444764649374,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048570752039879924,
      "loss": 4.7701,
      "mean_token_accuracy": 0.23451704829931258,
      "num_tokens": 29258710.0,
      "step": 12760
    },
    {
      "entropy": 5.292201662063599,
      "epoch": 1.2262247838616716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004856954937953539,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21470081061124802,
      "num_tokens": 29270173.0,
      "step": 12765
    },
    {
      "entropy": 5.312425422668457,
      "epoch": 1.2267050912584054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856834623001848,
      "loss": 4.9348,
      "mean_token_accuracy": 0.22782095968723298,
      "num_tokens": 29280407.0,
      "step": 12770
    },
    {
      "entropy": 5.216037750244141,
      "epoch": 1.2271853986551393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004856714259135713,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23695090115070344,
      "num_tokens": 29292287.0,
      "step": 12775
    },
    {
      "entropy": 5.244294261932373,
      "epoch": 1.2276657060518732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004856593846357929,
      "loss": 4.937,
      "mean_token_accuracy": 0.22032152414321898,
      "num_tokens": 29303099.0,
      "step": 12780
    },
    {
      "entropy": 5.349213981628418,
      "epoch": 1.228146013448607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004856473384671291,
      "loss": 4.9769,
      "mean_token_accuracy": 0.22503996640443802,
      "num_tokens": 29314445.0,
      "step": 12785
    },
    {
      "entropy": 5.235841941833496,
      "epoch": 1.228626320845341,
      "grad_norm": 1.0,
      "learning_rate": 0.00048563528740785955,
      "loss": 4.846,
      "mean_token_accuracy": 0.226559641957283,
      "num_tokens": 29325309.0,
      "step": 12790
    },
    {
      "entropy": 5.223233222961426,
      "epoch": 1.2291066282420748,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048562323145826414,
      "loss": 4.7954,
      "mean_token_accuracy": 0.23909534364938737,
      "num_tokens": 29338582.0,
      "step": 12795
    },
    {
      "entropy": 5.250769090652466,
      "epoch": 1.229586935638809,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004856111706186227,
      "loss": 4.8627,
      "mean_token_accuracy": 0.2299795016646385,
      "num_tokens": 29349875.0,
      "step": 12800
    },
    {
      "entropy": 5.324054002761841,
      "epoch": 1.2300672430355428,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048559910488921534,
      "loss": 4.9695,
      "mean_token_accuracy": 0.215803799033165,
      "num_tokens": 29361800.0,
      "step": 12805
    },
    {
      "entropy": 5.214760780334473,
      "epoch": 1.2305475504322767,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000485587034270322,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22895121574401855,
      "num_tokens": 29372795.0,
      "step": 12810
    },
    {
      "entropy": 5.2836075782775875,
      "epoch": 1.2310278578290106,
      "grad_norm": 0.96875,
      "learning_rate": 0.000485574958762223,
      "loss": 4.9469,
      "mean_token_accuracy": 0.2247391402721405,
      "num_tokens": 29385391.0,
      "step": 12815
    },
    {
      "entropy": 5.224454259872436,
      "epoch": 1.2315081652257445,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048556287836519886,
      "loss": 4.79,
      "mean_token_accuracy": 0.2364763453602791,
      "num_tokens": 29397113.0,
      "step": 12820
    },
    {
      "entropy": 5.324073648452758,
      "epoch": 1.2319884726224783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004855507930795299,
      "loss": 4.9907,
      "mean_token_accuracy": 0.22689314931631088,
      "num_tokens": 29407552.0,
      "step": 12825
    },
    {
      "entropy": 5.247876214981079,
      "epoch": 1.2324687800192122,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048553870290549665,
      "loss": 4.8971,
      "mean_token_accuracy": 0.22883429527282714,
      "num_tokens": 29418500.0,
      "step": 12830
    },
    {
      "entropy": 5.3370579242706295,
      "epoch": 1.232949087415946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048552660784338,
      "loss": 4.89,
      "mean_token_accuracy": 0.22561165541410447,
      "num_tokens": 29430335.0,
      "step": 12835
    },
    {
      "entropy": 5.286272573471069,
      "epoch": 1.23342939481268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004855145078934606,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22115739285945893,
      "num_tokens": 29441435.0,
      "step": 12840
    },
    {
      "entropy": 5.242755031585693,
      "epoch": 1.233909702209414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004855024030560195,
      "loss": 4.8811,
      "mean_token_accuracy": 0.22561461180448533,
      "num_tokens": 29453140.0,
      "step": 12845
    },
    {
      "entropy": 5.198592662811279,
      "epoch": 1.234390009606148,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004854902933313376,
      "loss": 4.8166,
      "mean_token_accuracy": 0.23384505808353423,
      "num_tokens": 29464572.0,
      "step": 12850
    },
    {
      "entropy": 5.360685014724732,
      "epoch": 1.2348703170028819,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048547817871969607,
      "loss": 4.9987,
      "mean_token_accuracy": 0.22083649337291716,
      "num_tokens": 29477069.0,
      "step": 12855
    },
    {
      "entropy": 5.352365732192993,
      "epoch": 1.2353506243996157,
      "grad_norm": 1.0,
      "learning_rate": 0.00048546605922137633,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22908525466918944,
      "num_tokens": 29486860.0,
      "step": 12860
    },
    {
      "entropy": 5.361484336853027,
      "epoch": 1.2358309317963496,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004854539348366596,
      "loss": 4.9556,
      "mean_token_accuracy": 0.22584578692913054,
      "num_tokens": 29499129.0,
      "step": 12865
    },
    {
      "entropy": 5.189972639083862,
      "epoch": 1.2363112391930835,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004854418055658274,
      "loss": 4.8216,
      "mean_token_accuracy": 0.2348661467432976,
      "num_tokens": 29510764.0,
      "step": 12870
    },
    {
      "entropy": 5.211154222488403,
      "epoch": 1.2367915465898176,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048542967140916134,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22735048979520797,
      "num_tokens": 29522882.0,
      "step": 12875
    },
    {
      "entropy": 5.28466763496399,
      "epoch": 1.2372718539865515,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004854175323669432,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2294978216290474,
      "num_tokens": 29533348.0,
      "step": 12880
    },
    {
      "entropy": 5.235966062545776,
      "epoch": 1.2377521613832854,
      "grad_norm": 0.875,
      "learning_rate": 0.0004854053884394547,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23199325799942017,
      "num_tokens": 29545649.0,
      "step": 12885
    },
    {
      "entropy": 5.303869771957397,
      "epoch": 1.2382324687800192,
      "grad_norm": 0.875,
      "learning_rate": 0.00048539323962697796,
      "loss": 4.8797,
      "mean_token_accuracy": 0.22824977040290834,
      "num_tokens": 29558252.0,
      "step": 12890
    },
    {
      "entropy": 5.3528844833374025,
      "epoch": 1.2387127761767531,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004853810859297949,
      "loss": 4.9944,
      "mean_token_accuracy": 0.22341640889644623,
      "num_tokens": 29569495.0,
      "step": 12895
    },
    {
      "entropy": 5.2402684688568115,
      "epoch": 1.239193083573487,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048536892734818773,
      "loss": 4.8686,
      "mean_token_accuracy": 0.23319740742444992,
      "num_tokens": 29582167.0,
      "step": 12900
    },
    {
      "entropy": 5.291449117660522,
      "epoch": 1.239673390970221,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004853567638824387,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22693177461624145,
      "num_tokens": 29593964.0,
      "step": 12905
    },
    {
      "entropy": 5.28184757232666,
      "epoch": 1.2401536983669548,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048534459553283026,
      "loss": 4.8516,
      "mean_token_accuracy": 0.23259321898221968,
      "num_tokens": 29604844.0,
      "step": 12910
    },
    {
      "entropy": 5.279412174224854,
      "epoch": 1.2406340057636887,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004853324222996449,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23385276049375534,
      "num_tokens": 29615219.0,
      "step": 12915
    },
    {
      "entropy": 5.36742262840271,
      "epoch": 1.2411143131604228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048532024418316525,
      "loss": 4.9745,
      "mean_token_accuracy": 0.22143367677927017,
      "num_tokens": 29626472.0,
      "step": 12920
    },
    {
      "entropy": 5.220947647094727,
      "epoch": 1.2415946205571566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004853080611836741,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2353065550327301,
      "num_tokens": 29637966.0,
      "step": 12925
    },
    {
      "entropy": 5.180609941482544,
      "epoch": 1.2420749279538905,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048529587330145427,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23085880875587464,
      "num_tokens": 29648730.0,
      "step": 12930
    },
    {
      "entropy": 5.208882427215576,
      "epoch": 1.2425552353506244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048528368053678863,
      "loss": 4.8186,
      "mean_token_accuracy": 0.22901165783405303,
      "num_tokens": 29660576.0,
      "step": 12935
    },
    {
      "entropy": 5.2119590759277346,
      "epoch": 1.2430355427473583,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004852714828899604,
      "loss": 4.8397,
      "mean_token_accuracy": 0.23153468519449233,
      "num_tokens": 29672906.0,
      "step": 12940
    },
    {
      "entropy": 5.215831899642945,
      "epoch": 1.2435158501440922,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048525928036125264,
      "loss": 4.835,
      "mean_token_accuracy": 0.23862196803092955,
      "num_tokens": 29685360.0,
      "step": 12945
    },
    {
      "entropy": 5.255987310409546,
      "epoch": 1.243996157540826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048524707295094884,
      "loss": 4.8501,
      "mean_token_accuracy": 0.2212822362780571,
      "num_tokens": 29697257.0,
      "step": 12950
    },
    {
      "entropy": 5.210432004928589,
      "epoch": 1.2444764649375601,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004852348606593322,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23418079167604447,
      "num_tokens": 29707877.0,
      "step": 12955
    },
    {
      "entropy": 5.256563949584961,
      "epoch": 1.244956772334294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048522264348668646,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22492487579584122,
      "num_tokens": 29719358.0,
      "step": 12960
    },
    {
      "entropy": 5.175535392761231,
      "epoch": 1.245437079731028,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004852104214332951,
      "loss": 4.7413,
      "mean_token_accuracy": 0.2445659816265106,
      "num_tokens": 29730383.0,
      "step": 12965
    },
    {
      "entropy": 5.27660346031189,
      "epoch": 1.2459173871277618,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048519819449944205,
      "loss": 4.8699,
      "mean_token_accuracy": 0.23206330388784407,
      "num_tokens": 29741142.0,
      "step": 12970
    },
    {
      "entropy": 5.209225845336914,
      "epoch": 1.2463976945244957,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000485185962685411,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22909057140350342,
      "num_tokens": 29754618.0,
      "step": 12975
    },
    {
      "entropy": 5.25170111656189,
      "epoch": 1.2468780019212296,
      "grad_norm": 1.015625,
      "learning_rate": 0.000485173725991486,
      "loss": 4.8513,
      "mean_token_accuracy": 0.22596676349639894,
      "num_tokens": 29767115.0,
      "step": 12980
    },
    {
      "entropy": 5.299691486358642,
      "epoch": 1.2473583093179634,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048516148441795124,
      "loss": 4.9224,
      "mean_token_accuracy": 0.22219357937574385,
      "num_tokens": 29778165.0,
      "step": 12985
    },
    {
      "entropy": 5.298619651794434,
      "epoch": 1.2478386167146973,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851492379650908,
      "loss": 4.9947,
      "mean_token_accuracy": 0.22215568870306016,
      "num_tokens": 29790528.0,
      "step": 12990
    },
    {
      "entropy": 5.2471170902252195,
      "epoch": 1.2483189241114312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004851369866331891,
      "loss": 4.8048,
      "mean_token_accuracy": 0.2278981477022171,
      "num_tokens": 29801709.0,
      "step": 12995
    },
    {
      "entropy": 5.2794050693511965,
      "epoch": 1.2487992315081653,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004851247304225306,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22001395523548126,
      "num_tokens": 29812963.0,
      "step": 13000
    },
    {
      "entropy": 5.303523874282837,
      "epoch": 1.2492795389048992,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851124693333997,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22509943544864655,
      "num_tokens": 29823711.0,
      "step": 13005
    },
    {
      "entropy": 5.240002632141113,
      "epoch": 1.249759846301633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004851002033660812,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23124948143959045,
      "num_tokens": 29834038.0,
      "step": 13010
    },
    {
      "entropy": 5.269131517410278,
      "epoch": 1.250240153698367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048508793252085994,
      "loss": 4.8505,
      "mean_token_accuracy": 0.2326391726732254,
      "num_tokens": 29844759.0,
      "step": 13015
    },
    {
      "entropy": 5.152782154083252,
      "epoch": 1.2507204610951008,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850756567980206,
      "loss": 4.7225,
      "mean_token_accuracy": 0.2361448213458061,
      "num_tokens": 29855643.0,
      "step": 13020
    },
    {
      "entropy": 5.288322019577026,
      "epoch": 1.2512007684918347,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048506337619784836,
      "loss": 4.935,
      "mean_token_accuracy": 0.22868784964084626,
      "num_tokens": 29866917.0,
      "step": 13025
    },
    {
      "entropy": 5.337522220611572,
      "epoch": 1.2516810758885688,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004850510907206283,
      "loss": 5.0083,
      "mean_token_accuracy": 0.2173992022871971,
      "num_tokens": 29878937.0,
      "step": 13030
    },
    {
      "entropy": 5.205234003067017,
      "epoch": 1.2521613832853027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048503880036664555,
      "loss": 4.8133,
      "mean_token_accuracy": 0.2338010624051094,
      "num_tokens": 29889544.0,
      "step": 13035
    },
    {
      "entropy": 5.216482877731323,
      "epoch": 1.2526416906820366,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004850265051361857,
      "loss": 4.8442,
      "mean_token_accuracy": 0.228147554397583,
      "num_tokens": 29901919.0,
      "step": 13040
    },
    {
      "entropy": 5.262098932266236,
      "epoch": 1.2531219980787704,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004850142050295339,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22450302243232728,
      "num_tokens": 29913870.0,
      "step": 13045
    },
    {
      "entropy": 5.217760372161865,
      "epoch": 1.2536023054755043,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048500190004697595,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2358544200658798,
      "num_tokens": 29925180.0,
      "step": 13050
    },
    {
      "entropy": 5.150117588043213,
      "epoch": 1.2540826128722382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004849895901887974,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2373497247695923,
      "num_tokens": 29936433.0,
      "step": 13055
    },
    {
      "entropy": 5.303596639633179,
      "epoch": 1.254562920268972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004849772754552842,
      "loss": 4.9266,
      "mean_token_accuracy": 0.2223336711525917,
      "num_tokens": 29948891.0,
      "step": 13060
    },
    {
      "entropy": 5.2072181224823,
      "epoch": 1.255043227665706,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048496495584672214,
      "loss": 4.7476,
      "mean_token_accuracy": 0.23764556050300598,
      "num_tokens": 29960113.0,
      "step": 13065
    },
    {
      "entropy": 5.24822301864624,
      "epoch": 1.2555235350624399,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048495263136339725,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22293685674667357,
      "num_tokens": 29972168.0,
      "step": 13070
    },
    {
      "entropy": 5.254476881027221,
      "epoch": 1.2560038424591737,
      "grad_norm": 1.0,
      "learning_rate": 0.0004849403020055956,
      "loss": 4.9031,
      "mean_token_accuracy": 0.23039231449365616,
      "num_tokens": 29982996.0,
      "step": 13075
    },
    {
      "entropy": 5.235622930526733,
      "epoch": 1.2564841498559078,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048492796777360373,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22832941710948945,
      "num_tokens": 29994088.0,
      "step": 13080
    },
    {
      "entropy": 5.355026626586914,
      "epoch": 1.2569644572526417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048491562866770767,
      "loss": 4.951,
      "mean_token_accuracy": 0.22545912265777587,
      "num_tokens": 30005403.0,
      "step": 13085
    },
    {
      "entropy": 5.254620122909546,
      "epoch": 1.2574447646493756,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048490328468819404,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23776559680700302,
      "num_tokens": 30015961.0,
      "step": 13090
    },
    {
      "entropy": 5.2098313331604,
      "epoch": 1.2579250720461095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048489093583534945,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23235996961593627,
      "num_tokens": 30026670.0,
      "step": 13095
    },
    {
      "entropy": 5.23477692604065,
      "epoch": 1.2584053794428434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004848785821094606,
      "loss": 4.8627,
      "mean_token_accuracy": 0.23102135211229324,
      "num_tokens": 30036711.0,
      "step": 13100
    },
    {
      "entropy": 5.255034589767456,
      "epoch": 1.2588856868395775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004848662235108142,
      "loss": 4.927,
      "mean_token_accuracy": 0.22417750507593154,
      "num_tokens": 30047587.0,
      "step": 13105
    },
    {
      "entropy": 5.241963481903076,
      "epoch": 1.2593659942363113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004848538600396973,
      "loss": 4.8713,
      "mean_token_accuracy": 0.22664868086576462,
      "num_tokens": 30059348.0,
      "step": 13110
    },
    {
      "entropy": 5.306854152679444,
      "epoch": 1.2598463016330452,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048484149169639694,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2283305749297142,
      "num_tokens": 30070485.0,
      "step": 13115
    },
    {
      "entropy": 5.235982751846313,
      "epoch": 1.260326609029779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004848291184812003,
      "loss": 4.8025,
      "mean_token_accuracy": 0.23823789805173873,
      "num_tokens": 30081114.0,
      "step": 13120
    },
    {
      "entropy": 5.252454042434692,
      "epoch": 1.260806916426513,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004848167403943945,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2243156760931015,
      "num_tokens": 30092634.0,
      "step": 13125
    },
    {
      "entropy": 5.240860414505005,
      "epoch": 1.2612872238232469,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048480435743626703,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23426947891712188,
      "num_tokens": 30104205.0,
      "step": 13130
    },
    {
      "entropy": 5.251623821258545,
      "epoch": 1.2617675312199808,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004847919696071054,
      "loss": 4.887,
      "mean_token_accuracy": 0.2259741246700287,
      "num_tokens": 30116978.0,
      "step": 13135
    },
    {
      "entropy": 5.230146312713623,
      "epoch": 1.2622478386167146,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048477957690719716,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23129130452871322,
      "num_tokens": 30128549.0,
      "step": 13140
    },
    {
      "entropy": 5.281802797317505,
      "epoch": 1.2627281460134485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004847671793368301,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22936152219772338,
      "num_tokens": 30139492.0,
      "step": 13145
    },
    {
      "entropy": 5.257601928710938,
      "epoch": 1.2632084534101824,
      "grad_norm": 0.9375,
      "learning_rate": 0.000484754776896292,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23105533719062804,
      "num_tokens": 30150450.0,
      "step": 13150
    },
    {
      "entropy": 5.314513349533081,
      "epoch": 1.2636887608069165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004847423695858708,
      "loss": 4.9156,
      "mean_token_accuracy": 0.22781710028648378,
      "num_tokens": 30162204.0,
      "step": 13155
    },
    {
      "entropy": 5.309820127487183,
      "epoch": 1.2641690682036504,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048472995740585456,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22963863760232925,
      "num_tokens": 30172574.0,
      "step": 13160
    },
    {
      "entropy": 5.151241397857666,
      "epoch": 1.2646493756003843,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004847175403565316,
      "loss": 4.7783,
      "mean_token_accuracy": 0.2357943907380104,
      "num_tokens": 30183957.0,
      "step": 13165
    },
    {
      "entropy": 5.242318391799927,
      "epoch": 1.2651296829971181,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048470511843818996,
      "loss": 4.8406,
      "mean_token_accuracy": 0.23869679719209672,
      "num_tokens": 30194207.0,
      "step": 13170
    },
    {
      "entropy": 5.21481671333313,
      "epoch": 1.265609990393852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004846926916511182,
      "loss": 4.8114,
      "mean_token_accuracy": 0.23948516994714736,
      "num_tokens": 30205180.0,
      "step": 13175
    },
    {
      "entropy": 5.235358810424804,
      "epoch": 1.266090297790586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004846802599956048,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22988174706697465,
      "num_tokens": 30217734.0,
      "step": 13180
    },
    {
      "entropy": 5.349527645111084,
      "epoch": 1.26657060518732,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048466782347193847,
      "loss": 5.0838,
      "mean_token_accuracy": 0.21450426280498505,
      "num_tokens": 30229835.0,
      "step": 13185
    },
    {
      "entropy": 5.298935747146606,
      "epoch": 1.267050912584054,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048465538208040775,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22609591037034987,
      "num_tokens": 30241932.0,
      "step": 13190
    },
    {
      "entropy": 5.312345123291015,
      "epoch": 1.2675312199807878,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048464293582130166,
      "loss": 4.9262,
      "mean_token_accuracy": 0.22402614802122117,
      "num_tokens": 30253149.0,
      "step": 13195
    },
    {
      "entropy": 5.285872268676758,
      "epoch": 1.2680115273775217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004846304846949091,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23213185071945192,
      "num_tokens": 30264083.0,
      "step": 13200
    },
    {
      "entropy": 5.222285223007202,
      "epoch": 1.2684918347742555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048461802870151916,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23633942157030105,
      "num_tokens": 30274832.0,
      "step": 13205
    },
    {
      "entropy": 5.182074880599975,
      "epoch": 1.2689721421709894,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048460556784142106,
      "loss": 4.8383,
      "mean_token_accuracy": 0.22933159470558168,
      "num_tokens": 30284945.0,
      "step": 13210
    },
    {
      "entropy": 5.297265100479126,
      "epoch": 1.2694524495677233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048459310211490406,
      "loss": 4.8681,
      "mean_token_accuracy": 0.23247205913066865,
      "num_tokens": 30295133.0,
      "step": 13215
    },
    {
      "entropy": 5.295944404602051,
      "epoch": 1.2699327569644572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004845806315222576,
      "loss": 4.9148,
      "mean_token_accuracy": 0.2254437819123268,
      "num_tokens": 30305268.0,
      "step": 13220
    },
    {
      "entropy": 5.181505298614502,
      "epoch": 1.270413064361191,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004845681560637711,
      "loss": 4.8021,
      "mean_token_accuracy": 0.22981556057929992,
      "num_tokens": 30317118.0,
      "step": 13225
    },
    {
      "entropy": 5.2919535636901855,
      "epoch": 1.270893371757925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004845556757397344,
      "loss": 4.9467,
      "mean_token_accuracy": 0.2195856213569641,
      "num_tokens": 30328684.0,
      "step": 13230
    },
    {
      "entropy": 5.317254829406738,
      "epoch": 1.271373679154659,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004845431905504372,
      "loss": 4.9648,
      "mean_token_accuracy": 0.220146481692791,
      "num_tokens": 30341433.0,
      "step": 13235
    },
    {
      "entropy": 5.256598711013794,
      "epoch": 1.271853986551393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048453070049616926,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23655469268560408,
      "num_tokens": 30353159.0,
      "step": 13240
    },
    {
      "entropy": 5.230399942398071,
      "epoch": 1.2723342939481268,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048451820557722064,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22237383723258972,
      "num_tokens": 30363251.0,
      "step": 13245
    },
    {
      "entropy": 5.343213748931885,
      "epoch": 1.2728146013448607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004845057057938815,
      "loss": 4.9472,
      "mean_token_accuracy": 0.23122985661029816,
      "num_tokens": 30375850.0,
      "step": 13250
    },
    {
      "entropy": 5.322688198089599,
      "epoch": 1.2732949087415946,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048449320114644185,
      "loss": 4.9418,
      "mean_token_accuracy": 0.22166601717472076,
      "num_tokens": 30386839.0,
      "step": 13255
    },
    {
      "entropy": 5.265662002563476,
      "epoch": 1.2737752161383284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004844806916351922,
      "loss": 4.933,
      "mean_token_accuracy": 0.22135305553674697,
      "num_tokens": 30398872.0,
      "step": 13260
    },
    {
      "entropy": 5.282140827178955,
      "epoch": 1.2742555235350626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004844681772604229,
      "loss": 4.854,
      "mean_token_accuracy": 0.23193912208080292,
      "num_tokens": 30409581.0,
      "step": 13265
    },
    {
      "entropy": 5.265312576293946,
      "epoch": 1.2747358309317964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048445565802242454,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22648796886205674,
      "num_tokens": 30420209.0,
      "step": 13270
    },
    {
      "entropy": 5.302470874786377,
      "epoch": 1.2752161383285303,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004844431339214878,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2334171548485756,
      "num_tokens": 30432093.0,
      "step": 13275
    },
    {
      "entropy": 5.265841484069824,
      "epoch": 1.2756964457252642,
      "grad_norm": 1.0,
      "learning_rate": 0.0004844306049579034,
      "loss": 4.821,
      "mean_token_accuracy": 0.22953782826662064,
      "num_tokens": 30442796.0,
      "step": 13280
    },
    {
      "entropy": 5.250163173675537,
      "epoch": 1.276176753121998,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048441807113196216,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22767444998025893,
      "num_tokens": 30455226.0,
      "step": 13285
    },
    {
      "entropy": 5.3058452129364015,
      "epoch": 1.276657060518732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048440553244395517,
      "loss": 4.9622,
      "mean_token_accuracy": 0.22073904424905777,
      "num_tokens": 30467082.0,
      "step": 13290
    },
    {
      "entropy": 5.307753419876098,
      "epoch": 1.2771373679154658,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048439298889417357,
      "loss": 4.867,
      "mean_token_accuracy": 0.22216540873050689,
      "num_tokens": 30479051.0,
      "step": 13295
    },
    {
      "entropy": 5.288737821578979,
      "epoch": 1.2776176753121997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048438044048290847,
      "loss": 4.9098,
      "mean_token_accuracy": 0.22596427500247956,
      "num_tokens": 30489989.0,
      "step": 13300
    },
    {
      "entropy": 5.254896259307861,
      "epoch": 1.2780979827089336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048436788721045135,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23151204288005828,
      "num_tokens": 30501533.0,
      "step": 13305
    },
    {
      "entropy": 5.2646644115448,
      "epoch": 1.2785782901056677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843553290770935,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2267972230911255,
      "num_tokens": 30512231.0,
      "step": 13310
    },
    {
      "entropy": 5.240894603729248,
      "epoch": 1.2790585975024016,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004843427660831266,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22594929188489915,
      "num_tokens": 30523204.0,
      "step": 13315
    },
    {
      "entropy": 5.311908149719239,
      "epoch": 1.2795389048991355,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048433019822884235,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21735631823539733,
      "num_tokens": 30534956.0,
      "step": 13320
    },
    {
      "entropy": 5.255427312850952,
      "epoch": 1.2800192122958693,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843176255145325,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23683699667453767,
      "num_tokens": 30545938.0,
      "step": 13325
    },
    {
      "entropy": 5.245936298370362,
      "epoch": 1.2804995196926032,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843050479404888,
      "loss": 4.7814,
      "mean_token_accuracy": 0.24090643525123595,
      "num_tokens": 30557323.0,
      "step": 13330
    },
    {
      "entropy": 5.263416814804077,
      "epoch": 1.280979827089337,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048429246550700343,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23212177008390428,
      "num_tokens": 30569960.0,
      "step": 13335
    },
    {
      "entropy": 5.222912359237671,
      "epoch": 1.2814601344860712,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004842798782143686,
      "loss": 4.8925,
      "mean_token_accuracy": 0.23066345751285552,
      "num_tokens": 30581904.0,
      "step": 13340
    },
    {
      "entropy": 5.289132833480835,
      "epoch": 1.281940441882805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048426728606287627,
      "loss": 4.869,
      "mean_token_accuracy": 0.23117629438638687,
      "num_tokens": 30592955.0,
      "step": 13345
    },
    {
      "entropy": 5.306263446807861,
      "epoch": 1.282420749279539,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004842546890528191,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22776635140180587,
      "num_tokens": 30604020.0,
      "step": 13350
    },
    {
      "entropy": 5.229012727737427,
      "epoch": 1.2829010566762729,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004842420871844893,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22665013372898102,
      "num_tokens": 30615623.0,
      "step": 13355
    },
    {
      "entropy": 5.292292833328247,
      "epoch": 1.2833813640730067,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842294804581796,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23146562278270721,
      "num_tokens": 30626258.0,
      "step": 13360
    },
    {
      "entropy": 5.306481266021729,
      "epoch": 1.2838616714697406,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048421686887418266,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22079545855522156,
      "num_tokens": 30637861.0,
      "step": 13365
    },
    {
      "entropy": 5.3031126976013185,
      "epoch": 1.2843419788664745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004842042524327912,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23415923118591309,
      "num_tokens": 30648835.0,
      "step": 13370
    },
    {
      "entropy": 5.3428877830505375,
      "epoch": 1.2848222862632084,
      "grad_norm": 1.0,
      "learning_rate": 0.0004841916311342983,
      "loss": 4.9548,
      "mean_token_accuracy": 0.21763100028038024,
      "num_tokens": 30659117.0,
      "step": 13375
    },
    {
      "entropy": 5.252786874771118,
      "epoch": 1.2853025936599423,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004841790049789969,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23366536796092988,
      "num_tokens": 30670282.0,
      "step": 13380
    },
    {
      "entropy": 5.259763097763061,
      "epoch": 1.2857829010566761,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048416637396718004,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22193025201559066,
      "num_tokens": 30681967.0,
      "step": 13385
    },
    {
      "entropy": 5.35748610496521,
      "epoch": 1.2862632084534102,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004841537380991411,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22481971979141235,
      "num_tokens": 30692803.0,
      "step": 13390
    },
    {
      "entropy": 5.212879991531372,
      "epoch": 1.2867435158501441,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048414109737517346,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23666589260101317,
      "num_tokens": 30704097.0,
      "step": 13395
    },
    {
      "entropy": 5.252661848068238,
      "epoch": 1.287223823246878,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004841284517955706,
      "loss": 4.9313,
      "mean_token_accuracy": 0.21789587438106536,
      "num_tokens": 30717983.0,
      "step": 13400
    },
    {
      "entropy": 5.188103437423706,
      "epoch": 1.2877041306436119,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000484115801360626,
      "loss": 4.7301,
      "mean_token_accuracy": 0.24137166738510132,
      "num_tokens": 30728446.0,
      "step": 13405
    },
    {
      "entropy": 5.1747394561767575,
      "epoch": 1.2881844380403458,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004841031460706335,
      "loss": 4.6955,
      "mean_token_accuracy": 0.2381485566496849,
      "num_tokens": 30739587.0,
      "step": 13410
    },
    {
      "entropy": 5.205757093429566,
      "epoch": 1.2886647454370797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048409048592588683,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23580324351787568,
      "num_tokens": 30750093.0,
      "step": 13415
    },
    {
      "entropy": 5.253301048278809,
      "epoch": 1.2891450528338138,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004840778209266799,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2294408783316612,
      "num_tokens": 30761692.0,
      "step": 13420
    },
    {
      "entropy": 5.21611475944519,
      "epoch": 1.2896253602305476,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048406515107330685,
      "loss": 4.8061,
      "mean_token_accuracy": 0.2366125300526619,
      "num_tokens": 30773474.0,
      "step": 13425
    },
    {
      "entropy": 5.360275554656982,
      "epoch": 1.2901056676272815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048405247636606173,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22020308673381805,
      "num_tokens": 30785464.0,
      "step": 13430
    },
    {
      "entropy": 5.279406499862671,
      "epoch": 1.2905859750240154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048403979680523894,
      "loss": 4.8855,
      "mean_token_accuracy": 0.22523798793554306,
      "num_tokens": 30796343.0,
      "step": 13435
    },
    {
      "entropy": 5.255763006210327,
      "epoch": 1.2910662824207493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004840271123911328,
      "loss": 4.8616,
      "mean_token_accuracy": 0.22892071455717086,
      "num_tokens": 30807795.0,
      "step": 13440
    },
    {
      "entropy": 5.242343091964722,
      "epoch": 1.2915465898174832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004840144231240377,
      "loss": 4.862,
      "mean_token_accuracy": 0.22744656950235367,
      "num_tokens": 30819629.0,
      "step": 13445
    },
    {
      "entropy": 5.247685194015503,
      "epoch": 1.292026897214217,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004840017290042484,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23356058448553085,
      "num_tokens": 30831545.0,
      "step": 13450
    },
    {
      "entropy": 5.257065486907959,
      "epoch": 1.292507204610951,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048398903003205957,
      "loss": 4.869,
      "mean_token_accuracy": 0.2316291391849518,
      "num_tokens": 30843614.0,
      "step": 13455
    },
    {
      "entropy": 5.222712469100952,
      "epoch": 1.2929875120076848,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048397632620776604,
      "loss": 4.7826,
      "mean_token_accuracy": 0.2333214536309242,
      "num_tokens": 30853749.0,
      "step": 13460
    },
    {
      "entropy": 5.214094829559326,
      "epoch": 1.293467819404419,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048396361753166276,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23285606056451796,
      "num_tokens": 30865065.0,
      "step": 13465
    },
    {
      "entropy": 5.25653190612793,
      "epoch": 1.2939481268011528,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048395090400404466,
      "loss": 4.8963,
      "mean_token_accuracy": 0.2295358806848526,
      "num_tokens": 30876746.0,
      "step": 13470
    },
    {
      "entropy": 5.29546012878418,
      "epoch": 1.2944284341978867,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048393818562520715,
      "loss": 4.9233,
      "mean_token_accuracy": 0.22934138476848603,
      "num_tokens": 30889543.0,
      "step": 13475
    },
    {
      "entropy": 5.248726654052734,
      "epoch": 1.2949087415946205,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048392546239544535,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22508079558610916,
      "num_tokens": 30901949.0,
      "step": 13480
    },
    {
      "entropy": 5.292910957336426,
      "epoch": 1.2953890489913544,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004839127343150547,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2238858550786972,
      "num_tokens": 30912660.0,
      "step": 13485
    },
    {
      "entropy": 5.258834314346314,
      "epoch": 1.2958693563880883,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004839000013843307,
      "loss": 4.871,
      "mean_token_accuracy": 0.23140785843133926,
      "num_tokens": 30924276.0,
      "step": 13490
    },
    {
      "entropy": 5.2450279712677,
      "epoch": 1.2963496637848224,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048388726360356894,
      "loss": 4.8094,
      "mean_token_accuracy": 0.23429520577192306,
      "num_tokens": 30935169.0,
      "step": 13495
    },
    {
      "entropy": 5.347187852859497,
      "epoch": 1.2968299711815563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004838745209730653,
      "loss": 4.9641,
      "mean_token_accuracy": 0.2227373331785202,
      "num_tokens": 30946116.0,
      "step": 13500
    },
    {
      "entropy": 5.294888973236084,
      "epoch": 1.2973102785782902,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048386177349311535,
      "loss": 4.9002,
      "mean_token_accuracy": 0.23050967752933502,
      "num_tokens": 30958509.0,
      "step": 13505
    },
    {
      "entropy": 5.278872632980347,
      "epoch": 1.297790585975024,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004838490211640153,
      "loss": 4.9048,
      "mean_token_accuracy": 0.22922322452068328,
      "num_tokens": 30969501.0,
      "step": 13510
    },
    {
      "entropy": 5.227163410186767,
      "epoch": 1.298270893371758,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004838362639860611,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23079610913991927,
      "num_tokens": 30979768.0,
      "step": 13515
    },
    {
      "entropy": 5.194260883331299,
      "epoch": 1.2987512007684918,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000483823501959549,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2361488714814186,
      "num_tokens": 30992020.0,
      "step": 13520
    },
    {
      "entropy": 5.335788106918335,
      "epoch": 1.2992315081652257,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048381073508477527,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22250716388225555,
      "num_tokens": 31004081.0,
      "step": 13525
    },
    {
      "entropy": 5.368961668014526,
      "epoch": 1.2997118155619596,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048379796336203625,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21917013376951217,
      "num_tokens": 31015776.0,
      "step": 13530
    },
    {
      "entropy": 5.225141382217407,
      "epoch": 1.3001921229586935,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004837851867916286,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23062217980623245,
      "num_tokens": 31028066.0,
      "step": 13535
    },
    {
      "entropy": 5.198830461502075,
      "epoch": 1.3006724303554273,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837724053738488,
      "loss": 4.8475,
      "mean_token_accuracy": 0.23482015132904052,
      "num_tokens": 31040234.0,
      "step": 13540
    },
    {
      "entropy": 5.2827095031738285,
      "epoch": 1.3011527377521614,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048375961910899373,
      "loss": 4.9009,
      "mean_token_accuracy": 0.22418154776096344,
      "num_tokens": 31051158.0,
      "step": 13545
    },
    {
      "entropy": 5.326676607131958,
      "epoch": 1.3016330451488953,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004837468279973602,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22995789349079132,
      "num_tokens": 31063404.0,
      "step": 13550
    },
    {
      "entropy": 5.287232112884522,
      "epoch": 1.3021133525456292,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004837340320392451,
      "loss": 4.857,
      "mean_token_accuracy": 0.22912510186433793,
      "num_tokens": 31073845.0,
      "step": 13555
    },
    {
      "entropy": 5.225528001785278,
      "epoch": 1.302593659942363,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048372123123494563,
      "loss": 4.8849,
      "mean_token_accuracy": 0.22522135823965073,
      "num_tokens": 31086914.0,
      "step": 13560
    },
    {
      "entropy": 5.255580902099609,
      "epoch": 1.303073967339097,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000483708425584759,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2346838429570198,
      "num_tokens": 31098627.0,
      "step": 13565
    },
    {
      "entropy": 5.381430196762085,
      "epoch": 1.3035542747358309,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004836956150889825,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21991396099328994,
      "num_tokens": 31110064.0,
      "step": 13570
    },
    {
      "entropy": 5.283044862747192,
      "epoch": 1.304034582132565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004836827997479134,
      "loss": 4.8263,
      "mean_token_accuracy": 0.23393401354551316,
      "num_tokens": 31121572.0,
      "step": 13575
    },
    {
      "entropy": 5.357141399383545,
      "epoch": 1.3045148895292988,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004836699795618494,
      "loss": 4.9462,
      "mean_token_accuracy": 0.2240598753094673,
      "num_tokens": 31133276.0,
      "step": 13580
    },
    {
      "entropy": 5.31301736831665,
      "epoch": 1.3049951969260327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004836571545310881,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22213945388793946,
      "num_tokens": 31144524.0,
      "step": 13585
    },
    {
      "entropy": 5.207334804534912,
      "epoch": 1.3054755043227666,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048364432465592723,
      "loss": 4.8113,
      "mean_token_accuracy": 0.2349862888455391,
      "num_tokens": 31155775.0,
      "step": 13590
    },
    {
      "entropy": 5.2668884754180905,
      "epoch": 1.3059558117195005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004836314899366647,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23196518868207933,
      "num_tokens": 31166176.0,
      "step": 13595
    },
    {
      "entropy": 5.289650392532349,
      "epoch": 1.3064361191162344,
      "grad_norm": 1.0,
      "learning_rate": 0.00048361865037359846,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23234157562255858,
      "num_tokens": 31178568.0,
      "step": 13600
    },
    {
      "entropy": 5.309096717834473,
      "epoch": 1.3069164265129682,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048360580596702664,
      "loss": 4.9152,
      "mean_token_accuracy": 0.22689673751592637,
      "num_tokens": 31188897.0,
      "step": 13605
    },
    {
      "entropy": 5.246697521209716,
      "epoch": 1.3073967339097021,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048359295671724744,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22784806191921234,
      "num_tokens": 31201370.0,
      "step": 13610
    },
    {
      "entropy": 5.2118549823760985,
      "epoch": 1.307877041306436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004835801026245592,
      "loss": 4.9412,
      "mean_token_accuracy": 0.22532319873571396,
      "num_tokens": 31212704.0,
      "step": 13615
    },
    {
      "entropy": 5.287687826156616,
      "epoch": 1.30835734870317,
      "grad_norm": 1.0,
      "learning_rate": 0.00048356724368926035,
      "loss": 4.9059,
      "mean_token_accuracy": 0.23144568502902985,
      "num_tokens": 31224447.0,
      "step": 13620
    },
    {
      "entropy": 5.353122568130493,
      "epoch": 1.308837656099904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048355437991164937,
      "loss": 4.9301,
      "mean_token_accuracy": 0.227942855656147,
      "num_tokens": 31235648.0,
      "step": 13625
    },
    {
      "entropy": 5.227401685714722,
      "epoch": 1.3093179634966379,
      "grad_norm": 1.0,
      "learning_rate": 0.000483541511292025,
      "loss": 4.9264,
      "mean_token_accuracy": 0.2249067783355713,
      "num_tokens": 31248330.0,
      "step": 13630
    },
    {
      "entropy": 5.300617694854736,
      "epoch": 1.3097982708933718,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048352863783068594,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22665570825338363,
      "num_tokens": 31260036.0,
      "step": 13635
    },
    {
      "entropy": 5.214733552932739,
      "epoch": 1.3102785782901056,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048351575952793117,
      "loss": 4.7256,
      "mean_token_accuracy": 0.2379601091146469,
      "num_tokens": 31271473.0,
      "step": 13640
    },
    {
      "entropy": 5.181400108337402,
      "epoch": 1.3107588856868395,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004835028763840595,
      "loss": 4.7295,
      "mean_token_accuracy": 0.23898615837097167,
      "num_tokens": 31282045.0,
      "step": 13645
    },
    {
      "entropy": 5.283769035339356,
      "epoch": 1.3112391930835736,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004834899883993703,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21971488744020462,
      "num_tokens": 31292644.0,
      "step": 13650
    },
    {
      "entropy": 5.291715621948242,
      "epoch": 1.3117195004803075,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048347709557416263,
      "loss": 4.8973,
      "mean_token_accuracy": 0.2369466617703438,
      "num_tokens": 31304422.0,
      "step": 13655
    },
    {
      "entropy": 5.232975149154663,
      "epoch": 1.3121998078770414,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004834641979087359,
      "loss": 4.808,
      "mean_token_accuracy": 0.22477449774742125,
      "num_tokens": 31314845.0,
      "step": 13660
    },
    {
      "entropy": 5.33367805480957,
      "epoch": 1.3126801152737753,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004834512954033894,
      "loss": 4.9582,
      "mean_token_accuracy": 0.2273569718003273,
      "num_tokens": 31325669.0,
      "step": 13665
    },
    {
      "entropy": 5.270903301239014,
      "epoch": 1.3131604226705091,
      "grad_norm": 1.0,
      "learning_rate": 0.00048343838805842284,
      "loss": 4.8509,
      "mean_token_accuracy": 0.23772657066583633,
      "num_tokens": 31336023.0,
      "step": 13670
    },
    {
      "entropy": 5.282229566574097,
      "epoch": 1.313640730067243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048342547587413583,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22155426293611527,
      "num_tokens": 31347146.0,
      "step": 13675
    },
    {
      "entropy": 5.297471761703491,
      "epoch": 1.314121037463977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004834125588508282,
      "loss": 4.9256,
      "mean_token_accuracy": 0.22759185433387757,
      "num_tokens": 31357616.0,
      "step": 13680
    },
    {
      "entropy": 5.336565637588501,
      "epoch": 1.3146013448607108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004833996369887998,
      "loss": 4.9375,
      "mean_token_accuracy": 0.22856993228197098,
      "num_tokens": 31369189.0,
      "step": 13685
    },
    {
      "entropy": 5.247689723968506,
      "epoch": 1.3150816522574447,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048338671028835063,
      "loss": 4.7954,
      "mean_token_accuracy": 0.2368406504392624,
      "num_tokens": 31380841.0,
      "step": 13690
    },
    {
      "entropy": 5.219300365447998,
      "epoch": 1.3155619596541785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048337377874978086,
      "loss": 4.8141,
      "mean_token_accuracy": 0.2332652673125267,
      "num_tokens": 31393092.0,
      "step": 13695
    },
    {
      "entropy": 5.255384922027588,
      "epoch": 1.3160422670509127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048336084237339067,
      "loss": 4.8778,
      "mean_token_accuracy": 0.23192788809537887,
      "num_tokens": 31404228.0,
      "step": 13700
    },
    {
      "entropy": 5.285502672195435,
      "epoch": 1.3165225744476465,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004833479011594804,
      "loss": 4.9731,
      "mean_token_accuracy": 0.21857989132404326,
      "num_tokens": 31416233.0,
      "step": 13705
    },
    {
      "entropy": 5.33919939994812,
      "epoch": 1.3170028818443804,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048333495510835057,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22501098811626435,
      "num_tokens": 31427787.0,
      "step": 13710
    },
    {
      "entropy": 5.132993268966675,
      "epoch": 1.3174831892411143,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048332200422030163,
      "loss": 4.7344,
      "mean_token_accuracy": 0.23716583251953124,
      "num_tokens": 31438736.0,
      "step": 13715
    },
    {
      "entropy": 5.148550176620484,
      "epoch": 1.3179634966378482,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004833090484956345,
      "loss": 4.7488,
      "mean_token_accuracy": 0.23064869046211242,
      "num_tokens": 31449225.0,
      "step": 13720
    },
    {
      "entropy": 5.312323904037475,
      "epoch": 1.318443804034582,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048329608793464966,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22453824579715728,
      "num_tokens": 31461223.0,
      "step": 13725
    },
    {
      "entropy": 5.378889799118042,
      "epoch": 1.3189241114313162,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004832831225376482,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2236689120531082,
      "num_tokens": 31472397.0,
      "step": 13730
    },
    {
      "entropy": 5.22356333732605,
      "epoch": 1.31940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004832701523049312,
      "loss": 4.809,
      "mean_token_accuracy": 0.2394632041454315,
      "num_tokens": 31484079.0,
      "step": 13735
    },
    {
      "entropy": 5.229662132263184,
      "epoch": 1.319884726224784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004832571772367997,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23017798811197282,
      "num_tokens": 31496732.0,
      "step": 13740
    },
    {
      "entropy": 5.288803339004517,
      "epoch": 1.3203650336215178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048324419733355485,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23404137045145035,
      "num_tokens": 31507601.0,
      "step": 13745
    },
    {
      "entropy": 5.217784357070923,
      "epoch": 1.3208453410182517,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048323121259549805,
      "loss": 4.8381,
      "mean_token_accuracy": 0.22916142642498016,
      "num_tokens": 31519551.0,
      "step": 13750
    },
    {
      "entropy": 5.245029163360596,
      "epoch": 1.3213256484149856,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048321822302293095,
      "loss": 4.9015,
      "mean_token_accuracy": 0.23384211361408233,
      "num_tokens": 31531531.0,
      "step": 13755
    },
    {
      "entropy": 5.278189373016358,
      "epoch": 1.3218059558117194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004832052286161549,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22441279590129853,
      "num_tokens": 31542386.0,
      "step": 13760
    },
    {
      "entropy": 5.283987236022949,
      "epoch": 1.3222862632084533,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048319222937547176,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23476779758930205,
      "num_tokens": 31554142.0,
      "step": 13765
    },
    {
      "entropy": 5.271662807464599,
      "epoch": 1.3227665706051872,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048317922530118323,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22670599520206453,
      "num_tokens": 31566687.0,
      "step": 13770
    },
    {
      "entropy": 5.284896945953369,
      "epoch": 1.3232468780019213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004831662163935912,
      "loss": 4.9494,
      "mean_token_accuracy": 0.22496905624866487,
      "num_tokens": 31577893.0,
      "step": 13775
    },
    {
      "entropy": 5.372726917266846,
      "epoch": 1.3237271853986552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048315320265299784,
      "loss": 5.0068,
      "mean_token_accuracy": 0.2185571238398552,
      "num_tokens": 31590483.0,
      "step": 13780
    },
    {
      "entropy": 5.298803377151489,
      "epoch": 1.324207492795389,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048314018407970516,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2291363701224327,
      "num_tokens": 31601373.0,
      "step": 13785
    },
    {
      "entropy": 5.257042932510376,
      "epoch": 1.324687800192123,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048312716067401535,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2186736524105072,
      "num_tokens": 31613981.0,
      "step": 13790
    },
    {
      "entropy": 5.326858615875244,
      "epoch": 1.3251681075888568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483114132436231,
      "loss": 4.9133,
      "mean_token_accuracy": 0.21952528953552247,
      "num_tokens": 31626727.0,
      "step": 13795
    },
    {
      "entropy": 5.232271671295166,
      "epoch": 1.3256484149855907,
      "grad_norm": 1.0,
      "learning_rate": 0.0004831010993666545,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23312344998121262,
      "num_tokens": 31638519.0,
      "step": 13800
    },
    {
      "entropy": 5.189148807525635,
      "epoch": 1.3261287223823248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004830880614655884,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23141148388385774,
      "num_tokens": 31649184.0,
      "step": 13805
    },
    {
      "entropy": 5.29591851234436,
      "epoch": 1.3266090297790587,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048307501873333527,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22616044878959657,
      "num_tokens": 31660266.0,
      "step": 13810
    },
    {
      "entropy": 5.247126674652099,
      "epoch": 1.3270893371757926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004830619711701982,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2310358017683029,
      "num_tokens": 31670841.0,
      "step": 13815
    },
    {
      "entropy": 5.249441146850586,
      "epoch": 1.3275696445725265,
      "grad_norm": 1.125,
      "learning_rate": 0.0004830489187764799,
      "loss": 4.8355,
      "mean_token_accuracy": 0.2253824442625046,
      "num_tokens": 31680786.0,
      "step": 13820
    },
    {
      "entropy": 5.185081672668457,
      "epoch": 1.3280499519692603,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004830358615524835,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22978882491588593,
      "num_tokens": 31692639.0,
      "step": 13825
    },
    {
      "entropy": 5.377322340011597,
      "epoch": 1.3285302593659942,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048302279949851215,
      "loss": 4.933,
      "mean_token_accuracy": 0.2265489086508751,
      "num_tokens": 31703520.0,
      "step": 13830
    },
    {
      "entropy": 5.296588802337647,
      "epoch": 1.329010566762728,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048300973261486906,
      "loss": 4.8002,
      "mean_token_accuracy": 0.24278075695037843,
      "num_tokens": 31715659.0,
      "step": 13835
    },
    {
      "entropy": 5.25448694229126,
      "epoch": 1.329490874159462,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004829966609018577,
      "loss": 4.8723,
      "mean_token_accuracy": 0.2271259382367134,
      "num_tokens": 31727896.0,
      "step": 13840
    },
    {
      "entropy": 5.248024320602417,
      "epoch": 1.3299711815561959,
      "grad_norm": 1.0,
      "learning_rate": 0.00048298358435978146,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23277393728494644,
      "num_tokens": 31738618.0,
      "step": 13845
    },
    {
      "entropy": 5.327314472198486,
      "epoch": 1.3304514889529298,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048297050298894394,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22460584193468094,
      "num_tokens": 31750306.0,
      "step": 13850
    },
    {
      "entropy": 5.369102668762207,
      "epoch": 1.3309317963496639,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004829574167896489,
      "loss": 4.9631,
      "mean_token_accuracy": 0.22483098655939102,
      "num_tokens": 31762015.0,
      "step": 13855
    },
    {
      "entropy": 5.217893934249878,
      "epoch": 1.3314121037463977,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048294432576220027,
      "loss": 4.859,
      "mean_token_accuracy": 0.2286288857460022,
      "num_tokens": 31774120.0,
      "step": 13860
    },
    {
      "entropy": 5.226041841506958,
      "epoch": 1.3318924111431316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004829312299069017,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23711872100830078,
      "num_tokens": 31785485.0,
      "step": 13865
    },
    {
      "entropy": 5.287604427337646,
      "epoch": 1.3323727185398655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048291812922405755,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22598023116588592,
      "num_tokens": 31795621.0,
      "step": 13870
    },
    {
      "entropy": 5.1972571849823,
      "epoch": 1.3328530259365994,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004829050237139717,
      "loss": 4.7858,
      "mean_token_accuracy": 0.2355323016643524,
      "num_tokens": 31806761.0,
      "step": 13875
    },
    {
      "entropy": 5.304302549362182,
      "epoch": 1.3333333333333333,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004828919133769486,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22634188830852509,
      "num_tokens": 31817996.0,
      "step": 13880
    },
    {
      "entropy": 5.3130370616912845,
      "epoch": 1.3338136407300674,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004828787982132926,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22790499031543732,
      "num_tokens": 31829888.0,
      "step": 13885
    },
    {
      "entropy": 5.3215779781341555,
      "epoch": 1.3342939481268012,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048286567822330815,
      "loss": 4.8956,
      "mean_token_accuracy": 0.22926011681556702,
      "num_tokens": 31842582.0,
      "step": 13890
    },
    {
      "entropy": 5.174940061569214,
      "epoch": 1.3347742555235351,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004828525534072999,
      "loss": 4.7718,
      "mean_token_accuracy": 0.22976758778095246,
      "num_tokens": 31854765.0,
      "step": 13895
    },
    {
      "entropy": 5.274209260940552,
      "epoch": 1.335254562920269,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048283942376557254,
      "loss": 4.8341,
      "mean_token_accuracy": 0.22601441591978072,
      "num_tokens": 31865505.0,
      "step": 13900
    },
    {
      "entropy": 5.185961961746216,
      "epoch": 1.3357348703170029,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048282628929843097,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23113172352313996,
      "num_tokens": 31876599.0,
      "step": 13905
    },
    {
      "entropy": 5.285845375061035,
      "epoch": 1.3362151777137368,
      "grad_norm": 1.0,
      "learning_rate": 0.00048281315000617996,
      "loss": 4.8392,
      "mean_token_accuracy": 0.2333289548754692,
      "num_tokens": 31887292.0,
      "step": 13910
    },
    {
      "entropy": 5.287595510482788,
      "epoch": 1.3366954851104706,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004828000058891248,
      "loss": 4.9017,
      "mean_token_accuracy": 0.21990728229284287,
      "num_tokens": 31899018.0,
      "step": 13915
    },
    {
      "entropy": 5.236228132247925,
      "epoch": 1.3371757925072045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004827868569475706,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23429256677627563,
      "num_tokens": 31910402.0,
      "step": 13920
    },
    {
      "entropy": 5.3130988597869875,
      "epoch": 1.3376560999039384,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048277370318182243,
      "loss": 4.954,
      "mean_token_accuracy": 0.21992845833301544,
      "num_tokens": 31921792.0,
      "step": 13925
    },
    {
      "entropy": 5.315318584442139,
      "epoch": 1.3381364073006723,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048276054459218596,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2208926111459732,
      "num_tokens": 31932786.0,
      "step": 13930
    },
    {
      "entropy": 5.304246950149536,
      "epoch": 1.3386167146974064,
      "grad_norm": 1.0,
      "learning_rate": 0.00048274738117896643,
      "loss": 4.896,
      "mean_token_accuracy": 0.22504196614027022,
      "num_tokens": 31944286.0,
      "step": 13935
    },
    {
      "entropy": 5.24188756942749,
      "epoch": 1.3390970220941403,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048273421294246966,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22039035558700562,
      "num_tokens": 31956304.0,
      "step": 13940
    },
    {
      "entropy": 5.234062719345093,
      "epoch": 1.3395773294908742,
      "grad_norm": 0.875,
      "learning_rate": 0.00048272103988300134,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22937825173139573,
      "num_tokens": 31968978.0,
      "step": 13945
    },
    {
      "entropy": 5.299713802337647,
      "epoch": 1.340057636887608,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004827078620008672,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22374628782272338,
      "num_tokens": 31980133.0,
      "step": 13950
    },
    {
      "entropy": 5.240270185470581,
      "epoch": 1.340537944284342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048269467929637337,
      "loss": 4.8312,
      "mean_token_accuracy": 0.24011737555265428,
      "num_tokens": 31990331.0,
      "step": 13955
    },
    {
      "entropy": 5.2799769878387455,
      "epoch": 1.341018251681076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048268149176982576,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22666653692722322,
      "num_tokens": 32001956.0,
      "step": 13960
    },
    {
      "entropy": 5.305698823928833,
      "epoch": 1.34149855907781,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048266829942153055,
      "loss": 4.9348,
      "mean_token_accuracy": 0.23122832626104356,
      "num_tokens": 32013577.0,
      "step": 13965
    },
    {
      "entropy": 5.378722858428955,
      "epoch": 1.3419788664745438,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048265510225179413,
      "loss": 4.9725,
      "mean_token_accuracy": 0.22069113850593566,
      "num_tokens": 32025751.0,
      "step": 13970
    },
    {
      "entropy": 5.261229515075684,
      "epoch": 1.3424591738712777,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004826419002609229,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22630411088466645,
      "num_tokens": 32036592.0,
      "step": 13975
    },
    {
      "entropy": 5.353383874893188,
      "epoch": 1.3429394812680115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048262869344922326,
      "loss": 4.8958,
      "mean_token_accuracy": 0.23175814002752304,
      "num_tokens": 32048122.0,
      "step": 13980
    },
    {
      "entropy": 5.297167730331421,
      "epoch": 1.3434197886647454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048261548181700186,
      "loss": 4.877,
      "mean_token_accuracy": 0.22640518099069595,
      "num_tokens": 32058673.0,
      "step": 13985
    },
    {
      "entropy": 5.161284732818603,
      "epoch": 1.3439000960614793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004826022653645655,
      "loss": 4.8081,
      "mean_token_accuracy": 0.23169436305761337,
      "num_tokens": 32070219.0,
      "step": 13990
    },
    {
      "entropy": 5.305863285064698,
      "epoch": 1.3443804034582132,
      "grad_norm": 0.953125,
      "learning_rate": 0.000482589044092221,
      "loss": 4.9265,
      "mean_token_accuracy": 0.2247641310095787,
      "num_tokens": 32081883.0,
      "step": 13995
    },
    {
      "entropy": 5.377393913269043,
      "epoch": 1.344860710854947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048257581800027527,
      "loss": 4.9532,
      "mean_token_accuracy": 0.22158667147159578,
      "num_tokens": 32094107.0,
      "step": 14000
    },
    {
      "entropy": 5.2598748207092285,
      "epoch": 1.345341018251681,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004825625870890354,
      "loss": 4.8229,
      "mean_token_accuracy": 0.23049319982528688,
      "num_tokens": 32105242.0,
      "step": 14005
    },
    {
      "entropy": 5.247548151016235,
      "epoch": 1.345821325648415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004825493513588086,
      "loss": 4.8187,
      "mean_token_accuracy": 0.23303017020225525,
      "num_tokens": 32117326.0,
      "step": 14010
    },
    {
      "entropy": 5.268412828445435,
      "epoch": 1.346301633045149,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048253611080990226,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22441578805446624,
      "num_tokens": 32129002.0,
      "step": 14015
    },
    {
      "entropy": 5.274739360809326,
      "epoch": 1.3467819404418828,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004825228654426236,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22274742871522904,
      "num_tokens": 32140380.0,
      "step": 14020
    },
    {
      "entropy": 5.2446269512176515,
      "epoch": 1.3472622478386167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004825096152572803,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23532659709453582,
      "num_tokens": 32151806.0,
      "step": 14025
    },
    {
      "entropy": 5.308951950073242,
      "epoch": 1.3477425552353506,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048249636025417974,
      "loss": 4.9825,
      "mean_token_accuracy": 0.22209024876356126,
      "num_tokens": 32163426.0,
      "step": 14030
    },
    {
      "entropy": 5.274008750915527,
      "epoch": 1.3482228626320845,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048248310043362997,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23311594724655152,
      "num_tokens": 32174349.0,
      "step": 14035
    },
    {
      "entropy": 5.347968482971192,
      "epoch": 1.3487031700288186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004824698357959386,
      "loss": 4.9585,
      "mean_token_accuracy": 0.2210630625486374,
      "num_tokens": 32185382.0,
      "step": 14040
    },
    {
      "entropy": 5.293476581573486,
      "epoch": 1.3491834774255524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048245656634141385,
      "loss": 4.844,
      "mean_token_accuracy": 0.23317668586969376,
      "num_tokens": 32195687.0,
      "step": 14045
    },
    {
      "entropy": 5.262878179550171,
      "epoch": 1.3496637848222863,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048244329207036354,
      "loss": 4.7925,
      "mean_token_accuracy": 0.2380603015422821,
      "num_tokens": 32205973.0,
      "step": 14050
    },
    {
      "entropy": 5.242523193359375,
      "epoch": 1.3501440922190202,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048243001298309604,
      "loss": 4.891,
      "mean_token_accuracy": 0.22983876913785933,
      "num_tokens": 32217469.0,
      "step": 14055
    },
    {
      "entropy": 5.245840978622437,
      "epoch": 1.350624399615754,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048241672907991954,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22985116243362427,
      "num_tokens": 32228257.0,
      "step": 14060
    },
    {
      "entropy": 5.299337434768677,
      "epoch": 1.351104707012488,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004824034403611424,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23191396594047547,
      "num_tokens": 32239420.0,
      "step": 14065
    },
    {
      "entropy": 5.312884187698364,
      "epoch": 1.3515850144092219,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004823901468270733,
      "loss": 4.8906,
      "mean_token_accuracy": 0.22693211287260057,
      "num_tokens": 32250962.0,
      "step": 14070
    },
    {
      "entropy": 5.2357707023620605,
      "epoch": 1.3520653218059557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004823768484780209,
      "loss": 4.8938,
      "mean_token_accuracy": 0.23066640198230742,
      "num_tokens": 32262310.0,
      "step": 14075
    },
    {
      "entropy": 5.233581495285034,
      "epoch": 1.3525456292026896,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048236354531429375,
      "loss": 4.78,
      "mean_token_accuracy": 0.23393207937479019,
      "num_tokens": 32273373.0,
      "step": 14080
    },
    {
      "entropy": 5.311652278900146,
      "epoch": 1.3530259365994235,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004823502373362009,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2289963409304619,
      "num_tokens": 32285020.0,
      "step": 14085
    },
    {
      "entropy": 5.271387386322021,
      "epoch": 1.3535062439961576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004823369245440512,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22892376631498337,
      "num_tokens": 32296224.0,
      "step": 14090
    },
    {
      "entropy": 5.248964595794678,
      "epoch": 1.3539865513928915,
      "grad_norm": 1.0,
      "learning_rate": 0.00048232360693815387,
      "loss": 4.87,
      "mean_token_accuracy": 0.2304162785410881,
      "num_tokens": 32306913.0,
      "step": 14095
    },
    {
      "entropy": 5.285278797149658,
      "epoch": 1.3544668587896254,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048231028451881786,
      "loss": 4.8387,
      "mean_token_accuracy": 0.22991230338811874,
      "num_tokens": 32317981.0,
      "step": 14100
    },
    {
      "entropy": 5.1857470035552975,
      "epoch": 1.3549471661863592,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004822969572863527,
      "loss": 4.7569,
      "mean_token_accuracy": 0.23584295809268951,
      "num_tokens": 32329656.0,
      "step": 14105
    },
    {
      "entropy": 5.298352193832398,
      "epoch": 1.3554274735830931,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048228362524106776,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2301860749721527,
      "num_tokens": 32339460.0,
      "step": 14110
    },
    {
      "entropy": 5.211993885040283,
      "epoch": 1.3559077809798272,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048227028838327253,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2284156247973442,
      "num_tokens": 32351237.0,
      "step": 14115
    },
    {
      "entropy": 5.262701034545898,
      "epoch": 1.356388088376561,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048225694671327665,
      "loss": 4.8088,
      "mean_token_accuracy": 0.2309606745839119,
      "num_tokens": 32362368.0,
      "step": 14120
    },
    {
      "entropy": 5.201881742477417,
      "epoch": 1.356868395773295,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004822436002313899,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23756198287010194,
      "num_tokens": 32373738.0,
      "step": 14125
    },
    {
      "entropy": 5.253780269622803,
      "epoch": 1.3573487031700289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004822302489379222,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2264113873243332,
      "num_tokens": 32384274.0,
      "step": 14130
    },
    {
      "entropy": 5.2112266540527346,
      "epoch": 1.3578290105667628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048221689283318335,
      "loss": 4.8,
      "mean_token_accuracy": 0.24014051258563995,
      "num_tokens": 32395692.0,
      "step": 14135
    },
    {
      "entropy": 5.21742205619812,
      "epoch": 1.3583093179634966,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004822035319174837,
      "loss": 4.8722,
      "mean_token_accuracy": 0.22939510345458985,
      "num_tokens": 32407367.0,
      "step": 14140
    },
    {
      "entropy": 5.273572492599487,
      "epoch": 1.3587896253602305,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004821901661911332,
      "loss": 4.8834,
      "mean_token_accuracy": 0.230050291121006,
      "num_tokens": 32418753.0,
      "step": 14145
    },
    {
      "entropy": 5.324258279800415,
      "epoch": 1.3592699327569644,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004821767956544423,
      "loss": 4.885,
      "mean_token_accuracy": 0.22922544032335282,
      "num_tokens": 32429499.0,
      "step": 14150
    },
    {
      "entropy": 5.2189429759979244,
      "epoch": 1.3597502401536983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004821634203077214,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22779567688703536,
      "num_tokens": 32441246.0,
      "step": 14155
    },
    {
      "entropy": 5.293742609024048,
      "epoch": 1.3602305475504322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000482150040151281,
      "loss": 4.909,
      "mean_token_accuracy": 0.22412626445293427,
      "num_tokens": 32452953.0,
      "step": 14160
    },
    {
      "entropy": 5.3198284149169925,
      "epoch": 1.3607108549471663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004821366551854318,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22880239188671112,
      "num_tokens": 32464030.0,
      "step": 14165
    },
    {
      "entropy": 5.215906763076783,
      "epoch": 1.3611911623439001,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004821232654104845,
      "loss": 4.6949,
      "mean_token_accuracy": 0.23983229994773864,
      "num_tokens": 32475091.0,
      "step": 14170
    },
    {
      "entropy": 5.24391016960144,
      "epoch": 1.361671469740634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048210987082675005,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22994298934936525,
      "num_tokens": 32486047.0,
      "step": 14175
    },
    {
      "entropy": 5.215381288528443,
      "epoch": 1.362151777137368,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048209647143453946,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23531037271022798,
      "num_tokens": 32497141.0,
      "step": 14180
    },
    {
      "entropy": 5.279766082763672,
      "epoch": 1.3626320845341018,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048208306723416356,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23049153089523317,
      "num_tokens": 32509282.0,
      "step": 14185
    },
    {
      "entropy": 5.292660140991211,
      "epoch": 1.3631123919308357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004820696582259339,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22351818084716796,
      "num_tokens": 32521383.0,
      "step": 14190
    },
    {
      "entropy": 5.234020090103149,
      "epoch": 1.3635926993275698,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004820562444101616,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22757648527622223,
      "num_tokens": 32533168.0,
      "step": 14195
    },
    {
      "entropy": 5.23175687789917,
      "epoch": 1.3640730067243036,
      "grad_norm": 1.0,
      "learning_rate": 0.0004820428257871581,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2310769811272621,
      "num_tokens": 32545333.0,
      "step": 14200
    },
    {
      "entropy": 5.171838331222534,
      "epoch": 1.3645533141210375,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004820294023572351,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23654371052980422,
      "num_tokens": 32556665.0,
      "step": 14205
    },
    {
      "entropy": 5.316615438461303,
      "epoch": 1.3650336215177714,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000482015974120704,
      "loss": 4.9841,
      "mean_token_accuracy": 0.22598125636577607,
      "num_tokens": 32568318.0,
      "step": 14210
    },
    {
      "entropy": 5.302801561355591,
      "epoch": 1.3655139289145053,
      "grad_norm": 1.0,
      "learning_rate": 0.00048200254107787677,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23048042207956315,
      "num_tokens": 32580010.0,
      "step": 14215
    },
    {
      "entropy": 5.337152338027954,
      "epoch": 1.3659942363112392,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048198910322906516,
      "loss": 4.9618,
      "mean_token_accuracy": 0.2174643263220787,
      "num_tokens": 32592075.0,
      "step": 14220
    },
    {
      "entropy": 5.360256814956665,
      "epoch": 1.366474543707973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048197566057458125,
      "loss": 4.991,
      "mean_token_accuracy": 0.22536295056343078,
      "num_tokens": 32604548.0,
      "step": 14225
    },
    {
      "entropy": 5.317366743087769,
      "epoch": 1.366954851104707,
      "grad_norm": 0.96875,
      "learning_rate": 0.000481962213114737,
      "loss": 4.889,
      "mean_token_accuracy": 0.22539651095867158,
      "num_tokens": 32616036.0,
      "step": 14230
    },
    {
      "entropy": 5.206678676605224,
      "epoch": 1.3674351585014408,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004819487608498448,
      "loss": 4.8528,
      "mean_token_accuracy": 0.2248489037156105,
      "num_tokens": 32628157.0,
      "step": 14235
    },
    {
      "entropy": 5.301694250106811,
      "epoch": 1.3679154658981747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048193530378021687,
      "loss": 4.9188,
      "mean_token_accuracy": 0.23109599351882934,
      "num_tokens": 32640571.0,
      "step": 14240
    },
    {
      "entropy": 5.2953883647918705,
      "epoch": 1.3683957732949088,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048192184190616567,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22575342804193496,
      "num_tokens": 32652005.0,
      "step": 14245
    },
    {
      "entropy": 5.257211589813233,
      "epoch": 1.3688760806916427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004819083752280037,
      "loss": 4.834,
      "mean_token_accuracy": 0.22995568960905075,
      "num_tokens": 32663655.0,
      "step": 14250
    },
    {
      "entropy": 5.2091656684875485,
      "epoch": 1.3693563880883766,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048189490374604373,
      "loss": 4.905,
      "mean_token_accuracy": 0.21928058564662933,
      "num_tokens": 32675419.0,
      "step": 14255
    },
    {
      "entropy": 5.25654673576355,
      "epoch": 1.3698366954851104,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818814274605983,
      "loss": 4.8169,
      "mean_token_accuracy": 0.2306096836924553,
      "num_tokens": 32686680.0,
      "step": 14260
    },
    {
      "entropy": 5.245594644546509,
      "epoch": 1.3703170028818443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818679463719805,
      "loss": 4.7483,
      "mean_token_accuracy": 0.2402739182114601,
      "num_tokens": 32697321.0,
      "step": 14265
    },
    {
      "entropy": 5.215081787109375,
      "epoch": 1.3707973102785782,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004818544604805033,
      "loss": 4.8641,
      "mean_token_accuracy": 0.22428559958934785,
      "num_tokens": 32708885.0,
      "step": 14270
    },
    {
      "entropy": 5.291753673553467,
      "epoch": 1.3712776176753123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004818409697864798,
      "loss": 4.8462,
      "mean_token_accuracy": 0.23593567609786986,
      "num_tokens": 32720517.0,
      "step": 14275
    },
    {
      "entropy": 5.247553014755249,
      "epoch": 1.3717579250720462,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048182747429022303,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22989873737096786,
      "num_tokens": 32731072.0,
      "step": 14280
    },
    {
      "entropy": 5.353033208847046,
      "epoch": 1.37223823246878,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818139739920465,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22635980546474457,
      "num_tokens": 32742117.0,
      "step": 14285
    },
    {
      "entropy": 5.289494132995605,
      "epoch": 1.372718539865514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004818004688922637,
      "loss": 4.91,
      "mean_token_accuracy": 0.22467222809791565,
      "num_tokens": 32753754.0,
      "step": 14290
    },
    {
      "entropy": 5.240292692184449,
      "epoch": 1.3731988472622478,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000481786958991188,
      "loss": 4.8336,
      "mean_token_accuracy": 0.2325671747326851,
      "num_tokens": 32766511.0,
      "step": 14295
    },
    {
      "entropy": 5.3824068069458,
      "epoch": 1.3736791546589817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048177344428913316,
      "loss": 4.9141,
      "mean_token_accuracy": 0.2220743790268898,
      "num_tokens": 32777715.0,
      "step": 14300
    },
    {
      "entropy": 5.286891794204712,
      "epoch": 1.3741594620557156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048175992478641293,
      "loss": 4.8527,
      "mean_token_accuracy": 0.23393717408180237,
      "num_tokens": 32789132.0,
      "step": 14305
    },
    {
      "entropy": 5.361181974411011,
      "epoch": 1.3746397694524495,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004817464004833412,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22809839695692063,
      "num_tokens": 32800439.0,
      "step": 14310
    },
    {
      "entropy": 5.263854074478149,
      "epoch": 1.3751200768491834,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048173287138023204,
      "loss": 4.838,
      "mean_token_accuracy": 0.22990477383136748,
      "num_tokens": 32813605.0,
      "step": 14315
    },
    {
      "entropy": 5.375340604782105,
      "epoch": 1.3756003842459175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004817193374773993,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22738397866487503,
      "num_tokens": 32824225.0,
      "step": 14320
    },
    {
      "entropy": 5.352204942703247,
      "epoch": 1.3760806916426513,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048170579877515753,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21352000087499617,
      "num_tokens": 32836917.0,
      "step": 14325
    },
    {
      "entropy": 5.260381698608398,
      "epoch": 1.3765609990393852,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004816922552738209,
      "loss": 4.8185,
      "mean_token_accuracy": 0.23746367692947387,
      "num_tokens": 32847972.0,
      "step": 14330
    },
    {
      "entropy": 5.184602308273315,
      "epoch": 1.377041306436119,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048167870697370373,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2378396213054657,
      "num_tokens": 32858922.0,
      "step": 14335
    },
    {
      "entropy": 5.240211915969849,
      "epoch": 1.377521613832853,
      "grad_norm": 1.0,
      "learning_rate": 0.0004816651538751207,
      "loss": 4.7704,
      "mean_token_accuracy": 0.2347193643450737,
      "num_tokens": 32869788.0,
      "step": 14340
    },
    {
      "entropy": 5.407274961471558,
      "epoch": 1.3780019212295869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048165159597838664,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21404103338718414,
      "num_tokens": 32881678.0,
      "step": 14345
    },
    {
      "entropy": 5.288625431060791,
      "epoch": 1.378482228626321,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000481638033283816,
      "loss": 4.88,
      "mean_token_accuracy": 0.22211176604032518,
      "num_tokens": 32894183.0,
      "step": 14350
    },
    {
      "entropy": 5.24539041519165,
      "epoch": 1.3789625360230549,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048162446579172387,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2328692451119423,
      "num_tokens": 32906001.0,
      "step": 14355
    },
    {
      "entropy": 5.2817951202392575,
      "epoch": 1.3794428434197887,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004816108935024252,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22786848843097687,
      "num_tokens": 32917236.0,
      "step": 14360
    },
    {
      "entropy": 5.271311616897583,
      "epoch": 1.3799231508165226,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048159731641623507,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2246285244822502,
      "num_tokens": 32929710.0,
      "step": 14365
    },
    {
      "entropy": 5.258879947662353,
      "epoch": 1.3804034582132565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004815837345334687,
      "loss": 4.9362,
      "mean_token_accuracy": 0.223575422167778,
      "num_tokens": 32941565.0,
      "step": 14370
    },
    {
      "entropy": 5.350609540939331,
      "epoch": 1.3808837656099904,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004815701478544415,
      "loss": 4.989,
      "mean_token_accuracy": 0.2222271144390106,
      "num_tokens": 32952730.0,
      "step": 14375
    },
    {
      "entropy": 5.249870491027832,
      "epoch": 1.3813640730067243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048155655637946876,
      "loss": 4.8192,
      "mean_token_accuracy": 0.230657896399498,
      "num_tokens": 32963589.0,
      "step": 14380
    },
    {
      "entropy": 5.247291088104248,
      "epoch": 1.3818443804034581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004815429601088662,
      "loss": 4.7942,
      "mean_token_accuracy": 0.2294119790196419,
      "num_tokens": 32975008.0,
      "step": 14385
    },
    {
      "entropy": 5.286423206329346,
      "epoch": 1.382324687800192,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004815293590429494,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21916555613279343,
      "num_tokens": 32986497.0,
      "step": 14390
    },
    {
      "entropy": 5.210685396194458,
      "epoch": 1.382804995196926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048151575318203417,
      "loss": 4.7122,
      "mean_token_accuracy": 0.2401335507631302,
      "num_tokens": 32998298.0,
      "step": 14395
    },
    {
      "entropy": 5.385661602020264,
      "epoch": 1.38328530259366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048150214252643637,
      "loss": 4.9806,
      "mean_token_accuracy": 0.22190933376550676,
      "num_tokens": 33010943.0,
      "step": 14400
    },
    {
      "entropy": 5.293891429901123,
      "epoch": 1.3837656099903939,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000481488527076472,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22781478464603425,
      "num_tokens": 33022139.0,
      "step": 14405
    },
    {
      "entropy": 5.31260085105896,
      "epoch": 1.3842459173871278,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004814749068324571,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22902003228664397,
      "num_tokens": 33034653.0,
      "step": 14410
    },
    {
      "entropy": 5.198929643630981,
      "epoch": 1.3847262247838616,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048146128179470804,
      "loss": 4.6566,
      "mean_token_accuracy": 0.2501588314771652,
      "num_tokens": 33044339.0,
      "step": 14415
    },
    {
      "entropy": 5.211531162261963,
      "epoch": 1.3852065321805955,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004814476519635411,
      "loss": 4.8781,
      "mean_token_accuracy": 0.23218251764774323,
      "num_tokens": 33055418.0,
      "step": 14420
    },
    {
      "entropy": 5.288378715515137,
      "epoch": 1.3856868395773294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048143401733927274,
      "loss": 4.945,
      "mean_token_accuracy": 0.22810614258050918,
      "num_tokens": 33067239.0,
      "step": 14425
    },
    {
      "entropy": 5.254675197601318,
      "epoch": 1.3861671469740635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048142037792221943,
      "loss": 4.7405,
      "mean_token_accuracy": 0.23895922005176545,
      "num_tokens": 33079101.0,
      "step": 14430
    },
    {
      "entropy": 5.285861349105835,
      "epoch": 1.3866474543707974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004814067337126978,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23541854321956635,
      "num_tokens": 33090265.0,
      "step": 14435
    },
    {
      "entropy": 5.246785831451416,
      "epoch": 1.3871277617675313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004813930847110248,
      "loss": 4.8081,
      "mean_token_accuracy": 0.23784155547618865,
      "num_tokens": 33101724.0,
      "step": 14440
    },
    {
      "entropy": 5.234897232055664,
      "epoch": 1.3876080691642652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004813794309175173,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23652615249156952,
      "num_tokens": 33113111.0,
      "step": 14445
    },
    {
      "entropy": 5.327132511138916,
      "epoch": 1.388088376560999,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048136577233249205,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2180224761366844,
      "num_tokens": 33123925.0,
      "step": 14450
    },
    {
      "entropy": 5.293259525299073,
      "epoch": 1.388568683957733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004813521089562666,
      "loss": 4.9058,
      "mean_token_accuracy": 0.224783031642437,
      "num_tokens": 33137400.0,
      "step": 14455
    },
    {
      "entropy": 5.186555528640747,
      "epoch": 1.3890489913544668,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004813384407891577,
      "loss": 4.7624,
      "mean_token_accuracy": 0.2362146332859993,
      "num_tokens": 33149326.0,
      "step": 14460
    },
    {
      "entropy": 5.334229564666748,
      "epoch": 1.3895292987512007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000481324767831483,
      "loss": 4.9371,
      "mean_token_accuracy": 0.21588520407676698,
      "num_tokens": 33159904.0,
      "step": 14465
    },
    {
      "entropy": 5.231456851959228,
      "epoch": 1.3900096061479346,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813110900835598,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23169203996658325,
      "num_tokens": 33171701.0,
      "step": 14470
    },
    {
      "entropy": 5.258462572097779,
      "epoch": 1.3904899135446687,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004812974075457058,
      "loss": 4.7445,
      "mean_token_accuracy": 0.23411553353071213,
      "num_tokens": 33181598.0,
      "step": 14475
    },
    {
      "entropy": 5.223377370834351,
      "epoch": 1.3909702209414025,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048128372021823845,
      "loss": 4.7824,
      "mean_token_accuracy": 0.234084589779377,
      "num_tokens": 33192674.0,
      "step": 14480
    },
    {
      "entropy": 5.2126545906066895,
      "epoch": 1.3914505283381364,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048127002810147574,
      "loss": 4.8385,
      "mean_token_accuracy": 0.23168370127677917,
      "num_tokens": 33203356.0,
      "step": 14485
    },
    {
      "entropy": 5.273957633972168,
      "epoch": 1.3919308357348703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004812563311957355,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22506984025239946,
      "num_tokens": 33215471.0,
      "step": 14490
    },
    {
      "entropy": 5.252017545700073,
      "epoch": 1.3924111431316042,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004812426295013356,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2267578586935997,
      "num_tokens": 33227656.0,
      "step": 14495
    },
    {
      "entropy": 5.306936883926392,
      "epoch": 1.392891450528338,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048122892301859433,
      "loss": 4.9618,
      "mean_token_accuracy": 0.21984824687242507,
      "num_tokens": 33239752.0,
      "step": 14500
    },
    {
      "entropy": 5.294703960418701,
      "epoch": 1.3933717579250722,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048121521174782983,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22325189858675004,
      "num_tokens": 33251352.0,
      "step": 14505
    },
    {
      "entropy": 5.301400804519654,
      "epoch": 1.393852065321806,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048120149568936044,
      "loss": 4.8987,
      "mean_token_accuracy": 0.22528892904520034,
      "num_tokens": 33262276.0,
      "step": 14510
    },
    {
      "entropy": 5.30625958442688,
      "epoch": 1.39433237271854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004811877748435046,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22357101887464523,
      "num_tokens": 33273615.0,
      "step": 14515
    },
    {
      "entropy": 5.2960858821868895,
      "epoch": 1.3948126801152738,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004811740492105809,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2323189213871956,
      "num_tokens": 33283990.0,
      "step": 14520
    },
    {
      "entropy": 5.236077833175659,
      "epoch": 1.3952929875120077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000481160318790908,
      "loss": 4.8479,
      "mean_token_accuracy": 0.240711709856987,
      "num_tokens": 33295472.0,
      "step": 14525
    },
    {
      "entropy": 5.303311443328857,
      "epoch": 1.3957732949087416,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048114658358480467,
      "loss": 4.9139,
      "mean_token_accuracy": 0.22817054092884065,
      "num_tokens": 33306742.0,
      "step": 14530
    },
    {
      "entropy": 5.336426401138306,
      "epoch": 1.3962536023054755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048113284359258977,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22762095481157302,
      "num_tokens": 33317737.0,
      "step": 14535
    },
    {
      "entropy": 5.289310121536255,
      "epoch": 1.3967339097022093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048111909881458234,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22434562146663667,
      "num_tokens": 33329673.0,
      "step": 14540
    },
    {
      "entropy": 5.172694206237793,
      "epoch": 1.3972142170989432,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048110534925110146,
      "loss": 4.7294,
      "mean_token_accuracy": 0.24283130019903182,
      "num_tokens": 33342047.0,
      "step": 14545
    },
    {
      "entropy": 5.227680253982544,
      "epoch": 1.397694524495677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004810915949024664,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2405730366706848,
      "num_tokens": 33353287.0,
      "step": 14550
    },
    {
      "entropy": 5.299156570434571,
      "epoch": 1.3981748318924112,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004810778357689965,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22985120564699174,
      "num_tokens": 33365465.0,
      "step": 14555
    },
    {
      "entropy": 5.246604633331299,
      "epoch": 1.398655139289145,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048106407185101116,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23307895809412002,
      "num_tokens": 33376680.0,
      "step": 14560
    },
    {
      "entropy": 5.222923421859742,
      "epoch": 1.399135446685879,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048105030314883,
      "loss": 4.8523,
      "mean_token_accuracy": 0.23031921088695526,
      "num_tokens": 33387995.0,
      "step": 14565
    },
    {
      "entropy": 5.252862405776978,
      "epoch": 1.3996157540826129,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004810365296627725,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23433769792318343,
      "num_tokens": 33400455.0,
      "step": 14570
    },
    {
      "entropy": 5.310233640670776,
      "epoch": 1.4000960614793467,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004810227513931587,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22237664163112641,
      "num_tokens": 33413264.0,
      "step": 14575
    },
    {
      "entropy": 5.23611421585083,
      "epoch": 1.4005763688760806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004810089683403084,
      "loss": 4.7632,
      "mean_token_accuracy": 0.23943042010068893,
      "num_tokens": 33423516.0,
      "step": 14580
    },
    {
      "entropy": 5.287558555603027,
      "epoch": 1.4010566762728147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004809951805045415,
      "loss": 4.9102,
      "mean_token_accuracy": 0.22108692079782485,
      "num_tokens": 33434952.0,
      "step": 14585
    },
    {
      "entropy": 5.2549182891845705,
      "epoch": 1.4015369836695486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048098138788617815,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23036137372255325,
      "num_tokens": 33447025.0,
      "step": 14590
    },
    {
      "entropy": 5.300864887237549,
      "epoch": 1.4020172910662825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004809675904855387,
      "loss": 4.8217,
      "mean_token_accuracy": 0.23377874493598938,
      "num_tokens": 33459990.0,
      "step": 14595
    },
    {
      "entropy": 5.282372951507568,
      "epoch": 1.4024975984630164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048095378830294343,
      "loss": 4.7943,
      "mean_token_accuracy": 0.2344907194375992,
      "num_tokens": 33471334.0,
      "step": 14600
    },
    {
      "entropy": 5.300161838531494,
      "epoch": 1.4029779058597502,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048093998133871276,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22837831676006318,
      "num_tokens": 33483744.0,
      "step": 14605
    },
    {
      "entropy": 5.2724371433258055,
      "epoch": 1.4034582132564841,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004809261695931671,
      "loss": 4.9302,
      "mean_token_accuracy": 0.2265966132283211,
      "num_tokens": 33496532.0,
      "step": 14610
    },
    {
      "entropy": 5.204832458496094,
      "epoch": 1.403938520653218,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004809123530666273,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2350610688328743,
      "num_tokens": 33507545.0,
      "step": 14615
    },
    {
      "entropy": 5.304956388473511,
      "epoch": 1.4044188280499519,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004808985317594142,
      "loss": 4.8529,
      "mean_token_accuracy": 0.23266727924346925,
      "num_tokens": 33519232.0,
      "step": 14620
    },
    {
      "entropy": 5.25790958404541,
      "epoch": 1.4048991354466858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048088470567184854,
      "loss": 4.8289,
      "mean_token_accuracy": 0.229205921292305,
      "num_tokens": 33531088.0,
      "step": 14625
    },
    {
      "entropy": 5.298642110824585,
      "epoch": 1.4053794428434199,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048087087480425133,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22187929898500441,
      "num_tokens": 33543991.0,
      "step": 14630
    },
    {
      "entropy": 5.316877317428589,
      "epoch": 1.4058597502401537,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004808570391569437,
      "loss": 4.9472,
      "mean_token_accuracy": 0.22170702666044234,
      "num_tokens": 33555303.0,
      "step": 14635
    },
    {
      "entropy": 5.326555252075195,
      "epoch": 1.4063400576368876,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048084319873024694,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22486514002084732,
      "num_tokens": 33565587.0,
      "step": 14640
    },
    {
      "entropy": 5.280157423019409,
      "epoch": 1.4068203650336215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004808293535244823,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23462168723344803,
      "num_tokens": 33576874.0,
      "step": 14645
    },
    {
      "entropy": 5.2238317966461185,
      "epoch": 1.4073006724303554,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004808155035399712,
      "loss": 4.7996,
      "mean_token_accuracy": 0.2404077023267746,
      "num_tokens": 33587703.0,
      "step": 14650
    },
    {
      "entropy": 5.19550895690918,
      "epoch": 1.4077809798270893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004808016487770354,
      "loss": 4.775,
      "mean_token_accuracy": 0.23514699637889863,
      "num_tokens": 33598487.0,
      "step": 14655
    },
    {
      "entropy": 5.251620531082153,
      "epoch": 1.4082612872238234,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048078778923599637,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22747791856527327,
      "num_tokens": 33610838.0,
      "step": 14660
    },
    {
      "entropy": 5.277871417999267,
      "epoch": 1.4087415946205573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048077392491717593,
      "loss": 4.8564,
      "mean_token_accuracy": 0.2244125559926033,
      "num_tokens": 33622726.0,
      "step": 14665
    },
    {
      "entropy": 5.221473836898804,
      "epoch": 1.4092219020172911,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048076005582089597,
      "loss": 4.8373,
      "mean_token_accuracy": 0.23049210757017136,
      "num_tokens": 33635922.0,
      "step": 14670
    },
    {
      "entropy": 5.38572359085083,
      "epoch": 1.409702209414025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048074618194747845,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22658740729093552,
      "num_tokens": 33648486.0,
      "step": 14675
    },
    {
      "entropy": 5.314838218688965,
      "epoch": 1.410182516810759,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004807323032972456,
      "loss": 4.881,
      "mean_token_accuracy": 0.2283330425620079,
      "num_tokens": 33659812.0,
      "step": 14680
    },
    {
      "entropy": 5.290904188156128,
      "epoch": 1.4106628242074928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004807184198705196,
      "loss": 4.85,
      "mean_token_accuracy": 0.2313144713640213,
      "num_tokens": 33671878.0,
      "step": 14685
    },
    {
      "entropy": 5.20562252998352,
      "epoch": 1.4111431316042267,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004807045316676226,
      "loss": 4.7494,
      "mean_token_accuracy": 0.2372682124376297,
      "num_tokens": 33683759.0,
      "step": 14690
    },
    {
      "entropy": 5.338994312286377,
      "epoch": 1.4116234390009605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004806906386888773,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22186314612627028,
      "num_tokens": 33694085.0,
      "step": 14695
    },
    {
      "entropy": 5.277775478363037,
      "epoch": 1.4121037463976944,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048067674093460607,
      "loss": 4.8076,
      "mean_token_accuracy": 0.23039206266403198,
      "num_tokens": 33705810.0,
      "step": 14700
    },
    {
      "entropy": 5.311179208755493,
      "epoch": 1.4125840537944283,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048066283840513175,
      "loss": 4.8907,
      "mean_token_accuracy": 0.2280463770031929,
      "num_tokens": 33716798.0,
      "step": 14705
    },
    {
      "entropy": 5.220149421691895,
      "epoch": 1.4130643611911624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004806489311007769,
      "loss": 4.8392,
      "mean_token_accuracy": 0.2377402126789093,
      "num_tokens": 33728515.0,
      "step": 14710
    },
    {
      "entropy": 5.267071962356567,
      "epoch": 1.4135446685878963,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048063501902186463,
      "loss": 4.859,
      "mean_token_accuracy": 0.22806317806243898,
      "num_tokens": 33740684.0,
      "step": 14715
    },
    {
      "entropy": 5.289895343780517,
      "epoch": 1.4140249759846302,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048062110216871775,
      "loss": 4.841,
      "mean_token_accuracy": 0.23514548689126968,
      "num_tokens": 33753017.0,
      "step": 14720
    },
    {
      "entropy": 5.262639808654785,
      "epoch": 1.414505283381364,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048060718054165945,
      "loss": 4.8616,
      "mean_token_accuracy": 0.23067759722471237,
      "num_tokens": 33764393.0,
      "step": 14725
    },
    {
      "entropy": 5.324838733673095,
      "epoch": 1.414985590778098,
      "grad_norm": 1.046875,
      "learning_rate": 0.000480593254141013,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2162758842110634,
      "num_tokens": 33774941.0,
      "step": 14730
    },
    {
      "entropy": 5.176590442657471,
      "epoch": 1.4154658981748318,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048057932296710165,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24545035660266876,
      "num_tokens": 33786534.0,
      "step": 14735
    },
    {
      "entropy": 5.373856449127198,
      "epoch": 1.415946205571566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004805653870202489,
      "loss": 4.9853,
      "mean_token_accuracy": 0.22237876504659654,
      "num_tokens": 33798339.0,
      "step": 14740
    },
    {
      "entropy": 5.411990833282471,
      "epoch": 1.4164265129682998,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048055144630077825,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2217285543680191,
      "num_tokens": 33810368.0,
      "step": 14745
    },
    {
      "entropy": 5.297972822189331,
      "epoch": 1.4169068203650337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048053750080901336,
      "loss": 4.8415,
      "mean_token_accuracy": 0.2293008014559746,
      "num_tokens": 33821111.0,
      "step": 14750
    },
    {
      "entropy": 5.241587162017822,
      "epoch": 1.4173871277617676,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048052355054527794,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23233817964792253,
      "num_tokens": 33833629.0,
      "step": 14755
    },
    {
      "entropy": 5.318043899536133,
      "epoch": 1.4178674351585014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048050959550989606,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22736700028181075,
      "num_tokens": 33846531.0,
      "step": 14760
    },
    {
      "entropy": 5.236403894424439,
      "epoch": 1.4183477425552353,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004804956357031916,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23051266074180604,
      "num_tokens": 33857251.0,
      "step": 14765
    },
    {
      "entropy": 5.260496616363525,
      "epoch": 1.4188280499519692,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048048167112548873,
      "loss": 4.8431,
      "mean_token_accuracy": 0.22880603969097138,
      "num_tokens": 33869314.0,
      "step": 14770
    },
    {
      "entropy": 5.366625070571899,
      "epoch": 1.419308357348703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048046770177711157,
      "loss": 4.9621,
      "mean_token_accuracy": 0.22374557852745056,
      "num_tokens": 33880203.0,
      "step": 14775
    },
    {
      "entropy": 5.255201721191407,
      "epoch": 1.419788664745437,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004804537276583844,
      "loss": 4.7747,
      "mean_token_accuracy": 0.23816870003938675,
      "num_tokens": 33891379.0,
      "step": 14780
    },
    {
      "entropy": 5.239676427841187,
      "epoch": 1.420268972142171,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004804397487696319,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23216888457536697,
      "num_tokens": 33902788.0,
      "step": 14785
    },
    {
      "entropy": 5.258911561965943,
      "epoch": 1.420749279538905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004804257651111783,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23529928773641587,
      "num_tokens": 33913609.0,
      "step": 14790
    },
    {
      "entropy": 5.239793729782105,
      "epoch": 1.4212295869356388,
      "grad_norm": 1.0,
      "learning_rate": 0.00048041177668334853,
      "loss": 4.7577,
      "mean_token_accuracy": 0.23409595787525178,
      "num_tokens": 33924379.0,
      "step": 14795
    },
    {
      "entropy": 5.286243724822998,
      "epoch": 1.4217098943323727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004803977834864672,
      "loss": 4.8825,
      "mean_token_accuracy": 0.2294205531477928,
      "num_tokens": 33936209.0,
      "step": 14800
    },
    {
      "entropy": 5.318099641799927,
      "epoch": 1.4221902017291066,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048038378552085927,
      "loss": 4.849,
      "mean_token_accuracy": 0.2209499552845955,
      "num_tokens": 33947679.0,
      "step": 14805
    },
    {
      "entropy": 5.236883735656738,
      "epoch": 1.4226705091258405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048036978278684974,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22910288870334625,
      "num_tokens": 33959474.0,
      "step": 14810
    },
    {
      "entropy": 5.3666339874267575,
      "epoch": 1.4231508165225746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803557752847636,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22008640319108963,
      "num_tokens": 33970831.0,
      "step": 14815
    },
    {
      "entropy": 5.404182863235474,
      "epoch": 1.4236311239193085,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048034176301492616,
      "loss": 4.933,
      "mean_token_accuracy": 0.22103903144598008,
      "num_tokens": 33981765.0,
      "step": 14820
    },
    {
      "entropy": 5.300649929046631,
      "epoch": 1.4241114313160423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004803277459776628,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23103639632463455,
      "num_tokens": 33992435.0,
      "step": 14825
    },
    {
      "entropy": 5.183168840408325,
      "epoch": 1.4245917387127762,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048031372417329875,
      "loss": 4.8108,
      "mean_token_accuracy": 0.23565564155578614,
      "num_tokens": 34004570.0,
      "step": 14830
    },
    {
      "entropy": 5.264811992645264,
      "epoch": 1.42507204610951,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004802996976021598,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23458890467882157,
      "num_tokens": 34015494.0,
      "step": 14835
    },
    {
      "entropy": 5.3279260158538815,
      "epoch": 1.425552353506244,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048028566626457145,
      "loss": 4.8751,
      "mean_token_accuracy": 0.22720762342214584,
      "num_tokens": 34026684.0,
      "step": 14840
    },
    {
      "entropy": 5.257372856140137,
      "epoch": 1.4260326609029779,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048027163016085947,
      "loss": 4.8593,
      "mean_token_accuracy": 0.22624347358942032,
      "num_tokens": 34038948.0,
      "step": 14845
    },
    {
      "entropy": 5.309388113021851,
      "epoch": 1.4265129682997117,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048025758929134976,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21523183286190034,
      "num_tokens": 34052216.0,
      "step": 14850
    },
    {
      "entropy": 5.309312391281128,
      "epoch": 1.4269932756964456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004802435436563684,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23272989839315414,
      "num_tokens": 34062602.0,
      "step": 14855
    },
    {
      "entropy": 5.30966477394104,
      "epoch": 1.4274735830931795,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048022949325624134,
      "loss": 4.8363,
      "mean_token_accuracy": 0.23435741513967515,
      "num_tokens": 34075049.0,
      "step": 14860
    },
    {
      "entropy": 5.300580263137817,
      "epoch": 1.4279538904899136,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048021543809129483,
      "loss": 4.9191,
      "mean_token_accuracy": 0.22533547580242158,
      "num_tokens": 34086719.0,
      "step": 14865
    },
    {
      "entropy": 5.26749963760376,
      "epoch": 1.4284341978866475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004802013781618552,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23508306294679643,
      "num_tokens": 34098439.0,
      "step": 14870
    },
    {
      "entropy": 5.331757736206055,
      "epoch": 1.4289145052833814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048018731346824895,
      "loss": 4.8633,
      "mean_token_accuracy": 0.23592744767665863,
      "num_tokens": 34110711.0,
      "step": 14875
    },
    {
      "entropy": 5.219444704055786,
      "epoch": 1.4293948126801153,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004801732440108026,
      "loss": 4.8079,
      "mean_token_accuracy": 0.22759344130754472,
      "num_tokens": 34122191.0,
      "step": 14880
    },
    {
      "entropy": 5.170486927032471,
      "epoch": 1.4298751200768491,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004801591697898427,
      "loss": 4.7605,
      "mean_token_accuracy": 0.2345541924238205,
      "num_tokens": 34132838.0,
      "step": 14885
    },
    {
      "entropy": 5.296013402938843,
      "epoch": 1.430355427473583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004801450908056961,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22282443791627884,
      "num_tokens": 34143394.0,
      "step": 14890
    },
    {
      "entropy": 5.403216552734375,
      "epoch": 1.4308357348703171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004801310070586896,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22237079441547394,
      "num_tokens": 34155934.0,
      "step": 14895
    },
    {
      "entropy": 5.173088312149048,
      "epoch": 1.431316042267051,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004801169185491503,
      "loss": 4.7395,
      "mean_token_accuracy": 0.24172770082950593,
      "num_tokens": 34167949.0,
      "step": 14900
    },
    {
      "entropy": 5.349506568908692,
      "epoch": 1.4317963496637849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048010282527740516,
      "loss": 5.0032,
      "mean_token_accuracy": 0.21882435381412507,
      "num_tokens": 34179733.0,
      "step": 14905
    },
    {
      "entropy": 5.2824653625488285,
      "epoch": 1.4322766570605188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048008872724378146,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23351285755634307,
      "num_tokens": 34190513.0,
      "step": 14910
    },
    {
      "entropy": 5.253194141387939,
      "epoch": 1.4327569644572526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004800746244486065,
      "loss": 4.8149,
      "mean_token_accuracy": 0.23984325230121611,
      "num_tokens": 34201562.0,
      "step": 14915
    },
    {
      "entropy": 5.164404439926147,
      "epoch": 1.4332372718539865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004800605168922077,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23139239400625228,
      "num_tokens": 34212637.0,
      "step": 14920
    },
    {
      "entropy": 5.239080429077148,
      "epoch": 1.4337175792507204,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048004640457491267,
      "loss": 4.8399,
      "mean_token_accuracy": 0.2338677927851677,
      "num_tokens": 34225394.0,
      "step": 14925
    },
    {
      "entropy": 5.286608552932739,
      "epoch": 1.4341978866474543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000480032287497049,
      "loss": 4.8224,
      "mean_token_accuracy": 0.2320640817284584,
      "num_tokens": 34236977.0,
      "step": 14930
    },
    {
      "entropy": 5.283879280090332,
      "epoch": 1.4346781940441882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048001816565894427,
      "loss": 4.9143,
      "mean_token_accuracy": 0.23253797143697738,
      "num_tokens": 34247486.0,
      "step": 14935
    },
    {
      "entropy": 5.263562631607056,
      "epoch": 1.435158501440922,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004800040390609267,
      "loss": 4.8064,
      "mean_token_accuracy": 0.2363015368580818,
      "num_tokens": 34259404.0,
      "step": 14940
    },
    {
      "entropy": 5.240008735656739,
      "epoch": 1.4356388088376562,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047998990770332396,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23898795545101165,
      "num_tokens": 34270388.0,
      "step": 14945
    },
    {
      "entropy": 5.328236532211304,
      "epoch": 1.43611911623439,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004799757715864643,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2278848722577095,
      "num_tokens": 34281321.0,
      "step": 14950
    },
    {
      "entropy": 5.341649150848388,
      "epoch": 1.436599423631124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004799616307106759,
      "loss": 4.9297,
      "mean_token_accuracy": 0.21883369684219361,
      "num_tokens": 34293177.0,
      "step": 14955
    },
    {
      "entropy": 5.332509517669678,
      "epoch": 1.4370797310278578,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000479947485076287,
      "loss": 4.9113,
      "mean_token_accuracy": 0.2280652940273285,
      "num_tokens": 34305175.0,
      "step": 14960
    },
    {
      "entropy": 5.307195663452148,
      "epoch": 1.4375600384245917,
      "grad_norm": 1.0,
      "learning_rate": 0.00047993333468362607,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22343757152557372,
      "num_tokens": 34317969.0,
      "step": 14965
    },
    {
      "entropy": 5.36025767326355,
      "epoch": 1.4380403458213258,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047991917953302173,
      "loss": 4.9297,
      "mean_token_accuracy": 0.22326131165027618,
      "num_tokens": 34329913.0,
      "step": 14970
    },
    {
      "entropy": 5.353980493545532,
      "epoch": 1.4385206532180597,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047990501962480236,
      "loss": 4.8659,
      "mean_token_accuracy": 0.2304289847612381,
      "num_tokens": 34341656.0,
      "step": 14975
    },
    {
      "entropy": 5.277102279663086,
      "epoch": 1.4390009606147935,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479890854959297,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22762879282236098,
      "num_tokens": 34351767.0,
      "step": 14980
    },
    {
      "entropy": 5.241183805465698,
      "epoch": 1.4394812680115274,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004798766855368344,
      "loss": 4.838,
      "mean_token_accuracy": 0.23126621097326278,
      "num_tokens": 34363437.0,
      "step": 14985
    },
    {
      "entropy": 5.361849975585938,
      "epoch": 1.4399615754082613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047986251135774343,
      "loss": 4.9177,
      "mean_token_accuracy": 0.23086390197277068,
      "num_tokens": 34374991.0,
      "step": 14990
    },
    {
      "entropy": 5.257470178604126,
      "epoch": 1.4404418828049952,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004798483324223533,
      "loss": 4.8385,
      "mean_token_accuracy": 0.2336692363023758,
      "num_tokens": 34386218.0,
      "step": 14995
    },
    {
      "entropy": 5.303475761413575,
      "epoch": 1.440922190201729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004798341487309932,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23103599697351457,
      "num_tokens": 34396287.0,
      "step": 15000
    },
    {
      "epoch": 1.440922190201729,
      "eval_entropy": 5.035100322534607,
      "eval_loss": 4.919432163238525,
      "eval_mean_token_accuracy": 0.23640541044195096,
      "eval_num_tokens": 34396287.0,
      "eval_runtime": 20.6617,
      "eval_samples_per_second": 1588.202,
      "eval_steps_per_second": 198.531,
      "step": 15000
    },
    {
      "entropy": 5.293306636810303,
      "epoch": 1.441402497598463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047981996028399233,
      "loss": 4.88,
      "mean_token_accuracy": 0.22476956695318223,
      "num_tokens": 34407251.0,
      "step": 15005
    },
    {
      "entropy": 5.399342155456543,
      "epoch": 1.4418828049951968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004798057670816802,
      "loss": 5.0089,
      "mean_token_accuracy": 0.22287137061357498,
      "num_tokens": 34419185.0,
      "step": 15010
    },
    {
      "entropy": 5.290394639968872,
      "epoch": 1.4423631123919307,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004797915691243863,
      "loss": 4.855,
      "mean_token_accuracy": 0.229799322783947,
      "num_tokens": 34431159.0,
      "step": 15015
    },
    {
      "entropy": 5.261924123764038,
      "epoch": 1.4428434197886648,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004797773664124403,
      "loss": 4.8304,
      "mean_token_accuracy": 0.23128030747175216,
      "num_tokens": 34442614.0,
      "step": 15020
    },
    {
      "entropy": 5.242826128005982,
      "epoch": 1.4433237271853987,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047976315894617195,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2372859463095665,
      "num_tokens": 34453984.0,
      "step": 15025
    },
    {
      "entropy": 5.240968942642212,
      "epoch": 1.4438040345821326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000479748946725911,
      "loss": 4.8148,
      "mean_token_accuracy": 0.23235574811697007,
      "num_tokens": 34466296.0,
      "step": 15030
    },
    {
      "entropy": 5.388931512832642,
      "epoch": 1.4442843419788665,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004797347297519875,
      "loss": 5.0241,
      "mean_token_accuracy": 0.2227710723876953,
      "num_tokens": 34478088.0,
      "step": 15035
    },
    {
      "entropy": 5.313416242599487,
      "epoch": 1.4447646493756003,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047972050802473154,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21813823878765107,
      "num_tokens": 34491664.0,
      "step": 15040
    },
    {
      "entropy": 5.242147779464721,
      "epoch": 1.4452449567723342,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004797062815444733,
      "loss": 4.8163,
      "mean_token_accuracy": 0.23425178825855256,
      "num_tokens": 34502977.0,
      "step": 15045
    },
    {
      "entropy": 5.2789794445037845,
      "epoch": 1.4457252641690683,
      "grad_norm": 1.109375,
      "learning_rate": 0.000479692050311543,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23795022815465927,
      "num_tokens": 34515067.0,
      "step": 15050
    },
    {
      "entropy": 5.322709083557129,
      "epoch": 1.4462055715658022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004796778143262711,
      "loss": 4.9362,
      "mean_token_accuracy": 0.22488246858119965,
      "num_tokens": 34525012.0,
      "step": 15055
    },
    {
      "entropy": 5.170891141891479,
      "epoch": 1.446685878962536,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004796635735889882,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23996219038963318,
      "num_tokens": 34535789.0,
      "step": 15060
    },
    {
      "entropy": 5.304094886779785,
      "epoch": 1.44716618635927,
      "grad_norm": 1.0,
      "learning_rate": 0.00047964932810002476,
      "loss": 4.8664,
      "mean_token_accuracy": 0.23800635635852813,
      "num_tokens": 34546276.0,
      "step": 15065
    },
    {
      "entropy": 5.282984066009521,
      "epoch": 1.4476464937560038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004796350778597117,
      "loss": 4.8575,
      "mean_token_accuracy": 0.2225933775305748,
      "num_tokens": 34558361.0,
      "step": 15070
    },
    {
      "entropy": 5.227920484542847,
      "epoch": 1.4481268011527377,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004796208228683796,
      "loss": 4.805,
      "mean_token_accuracy": 0.23604972213506697,
      "num_tokens": 34569482.0,
      "step": 15075
    },
    {
      "entropy": 5.325941753387451,
      "epoch": 1.4486071085494716,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047960656312635977,
      "loss": 4.8815,
      "mean_token_accuracy": 0.23136223405599593,
      "num_tokens": 34580128.0,
      "step": 15080
    },
    {
      "entropy": 5.227814388275147,
      "epoch": 1.4490874159462055,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004795922986339831,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22906449884176255,
      "num_tokens": 34591105.0,
      "step": 15085
    },
    {
      "entropy": 5.297535991668701,
      "epoch": 1.4495677233429394,
      "grad_norm": 1.0,
      "learning_rate": 0.00047957802939158057,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22611428052186966,
      "num_tokens": 34602618.0,
      "step": 15090
    },
    {
      "entropy": 5.305563640594483,
      "epoch": 1.4500480307396733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004795637553994838,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22050586640834807,
      "num_tokens": 34614179.0,
      "step": 15095
    },
    {
      "entropy": 5.357950687408447,
      "epoch": 1.4505283381364074,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047954947665802404,
      "loss": 4.968,
      "mean_token_accuracy": 0.2214486226439476,
      "num_tokens": 34625456.0,
      "step": 15100
    },
    {
      "entropy": 5.2753050327301025,
      "epoch": 1.4510086455331412,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004795351931675329,
      "loss": 4.8164,
      "mean_token_accuracy": 0.2386945128440857,
      "num_tokens": 34636268.0,
      "step": 15105
    },
    {
      "entropy": 5.2332155227661135,
      "epoch": 1.4514889529298751,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004795209049283419,
      "loss": 4.7997,
      "mean_token_accuracy": 0.23761892318725586,
      "num_tokens": 34647665.0,
      "step": 15110
    },
    {
      "entropy": 5.281881380081177,
      "epoch": 1.451969260326609,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004795066119407827,
      "loss": 4.8824,
      "mean_token_accuracy": 0.22563790380954743,
      "num_tokens": 34659965.0,
      "step": 15115
    },
    {
      "entropy": 5.279536724090576,
      "epoch": 1.4524495677233429,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794923142051873,
      "loss": 4.9298,
      "mean_token_accuracy": 0.22392759025096892,
      "num_tokens": 34671724.0,
      "step": 15120
    },
    {
      "entropy": 5.280497932434082,
      "epoch": 1.452929875120077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047947801172188755,
      "loss": 4.8075,
      "mean_token_accuracy": 0.2359166517853737,
      "num_tokens": 34682445.0,
      "step": 15125
    },
    {
      "entropy": 5.181373119354248,
      "epoch": 1.4534101825168109,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004794637044912155,
      "loss": 4.7485,
      "mean_token_accuracy": 0.23708086758852004,
      "num_tokens": 34693621.0,
      "step": 15130
    },
    {
      "entropy": 5.223783540725708,
      "epoch": 1.4538904899135447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004794493925135034,
      "loss": 4.74,
      "mean_token_accuracy": 0.2439083218574524,
      "num_tokens": 34703978.0,
      "step": 15135
    },
    {
      "entropy": 5.290994453430176,
      "epoch": 1.4543707973102786,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047943507578908357,
      "loss": 4.9128,
      "mean_token_accuracy": 0.23380369693040848,
      "num_tokens": 34715468.0,
      "step": 15140
    },
    {
      "entropy": 5.302586030960083,
      "epoch": 1.4548511047070125,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004794207543182883,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22884106040000915,
      "num_tokens": 34726383.0,
      "step": 15145
    },
    {
      "entropy": 5.297244071960449,
      "epoch": 1.4553314121037464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047940642810145005,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2315853238105774,
      "num_tokens": 34737123.0,
      "step": 15150
    },
    {
      "entropy": 5.187300109863282,
      "epoch": 1.4558117195004803,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047939209713890156,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23642351627349853,
      "num_tokens": 34749197.0,
      "step": 15155
    },
    {
      "entropy": 5.27302508354187,
      "epoch": 1.4562920268972142,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047937776143097547,
      "loss": 4.8834,
      "mean_token_accuracy": 0.2244852989912033,
      "num_tokens": 34759785.0,
      "step": 15160
    },
    {
      "entropy": 5.182751989364624,
      "epoch": 1.456772334293948,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793634209780047,
      "loss": 4.7378,
      "mean_token_accuracy": 0.23662802577018738,
      "num_tokens": 34770938.0,
      "step": 15165
    },
    {
      "entropy": 5.255116605758667,
      "epoch": 1.457252641690682,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004793490757803221,
      "loss": 4.7814,
      "mean_token_accuracy": 0.23161083608865737,
      "num_tokens": 34782126.0,
      "step": 15170
    },
    {
      "entropy": 5.203011226654053,
      "epoch": 1.457732949087416,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047933472583826063,
      "loss": 4.795,
      "mean_token_accuracy": 0.22943902462720872,
      "num_tokens": 34793802.0,
      "step": 15175
    },
    {
      "entropy": 5.243001222610474,
      "epoch": 1.45821325648415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004793203711521537,
      "loss": 4.8665,
      "mean_token_accuracy": 0.23804147392511368,
      "num_tokens": 34805696.0,
      "step": 15180
    },
    {
      "entropy": 5.362994718551636,
      "epoch": 1.4586935638808838,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047930601172233446,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2218722626566887,
      "num_tokens": 34818679.0,
      "step": 15185
    },
    {
      "entropy": 5.330373382568359,
      "epoch": 1.4591738712776177,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047929164754913624,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23091112971305847,
      "num_tokens": 34830528.0,
      "step": 15190
    },
    {
      "entropy": 5.250917434692383,
      "epoch": 1.4596541786743515,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004792772786328926,
      "loss": 4.7499,
      "mean_token_accuracy": 0.24085568338632585,
      "num_tokens": 34841621.0,
      "step": 15195
    },
    {
      "entropy": 5.186902046203613,
      "epoch": 1.4601344860710854,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047926290497393714,
      "loss": 4.7637,
      "mean_token_accuracy": 0.23436554074287413,
      "num_tokens": 34854448.0,
      "step": 15200
    },
    {
      "entropy": 5.295994520187378,
      "epoch": 1.4606147934678195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004792485265726036,
      "loss": 4.8559,
      "mean_token_accuracy": 0.23443111777305603,
      "num_tokens": 34866492.0,
      "step": 15205
    },
    {
      "entropy": 5.24373722076416,
      "epoch": 1.4610951008645534,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004792341434292257,
      "loss": 4.8552,
      "mean_token_accuracy": 0.22922273725271225,
      "num_tokens": 34876869.0,
      "step": 15210
    },
    {
      "entropy": 5.214378929138183,
      "epoch": 1.4615754082612873,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004792197555441374,
      "loss": 4.7485,
      "mean_token_accuracy": 0.24341776371002197,
      "num_tokens": 34888449.0,
      "step": 15215
    },
    {
      "entropy": 5.236873626708984,
      "epoch": 1.4620557156580212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792053629176729,
      "loss": 4.8069,
      "mean_token_accuracy": 0.2365108847618103,
      "num_tokens": 34898124.0,
      "step": 15220
    },
    {
      "entropy": 5.246282482147217,
      "epoch": 1.462536023054755,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004791909655501662,
      "loss": 4.7701,
      "mean_token_accuracy": 0.24101841896772386,
      "num_tokens": 34909128.0,
      "step": 15225
    },
    {
      "entropy": 5.190614175796509,
      "epoch": 1.463016330451489,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791765634419516,
      "loss": 4.8209,
      "mean_token_accuracy": 0.23349976390600205,
      "num_tokens": 34920541.0,
      "step": 15230
    },
    {
      "entropy": 5.184946346282959,
      "epoch": 1.4634966378482228,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047916215659336343,
      "loss": 4.7752,
      "mean_token_accuracy": 0.23330548107624055,
      "num_tokens": 34931605.0,
      "step": 15235
    },
    {
      "entropy": 5.246107482910157,
      "epoch": 1.4639769452449567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004791477450047363,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23428485840559005,
      "num_tokens": 34943057.0,
      "step": 15240
    },
    {
      "entropy": 5.296639680862427,
      "epoch": 1.4644572526416906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047913332867640464,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21893313974142076,
      "num_tokens": 34954386.0,
      "step": 15245
    },
    {
      "entropy": 5.232596158981323,
      "epoch": 1.4649375600384245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004791189076087033,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23593699634075166,
      "num_tokens": 34965874.0,
      "step": 15250
    },
    {
      "entropy": 5.279146146774292,
      "epoch": 1.4654178674351586,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047910448180196703,
      "loss": 4.902,
      "mean_token_accuracy": 0.22860787212848663,
      "num_tokens": 34977408.0,
      "step": 15255
    },
    {
      "entropy": 5.25410795211792,
      "epoch": 1.4658981748318924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004790900512565307,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23952234238386155,
      "num_tokens": 34987788.0,
      "step": 15260
    },
    {
      "entropy": 5.292865467071533,
      "epoch": 1.4663784822286263,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004790756159727294,
      "loss": 4.8951,
      "mean_token_accuracy": 0.22990466058254241,
      "num_tokens": 35001051.0,
      "step": 15265
    },
    {
      "entropy": 5.2807353973388675,
      "epoch": 1.4668587896253602,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047906117595089835,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23703105002641678,
      "num_tokens": 35012621.0,
      "step": 15270
    },
    {
      "entropy": 5.269022989273071,
      "epoch": 1.467339097022094,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004790467311913727,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23129868805408477,
      "num_tokens": 35023789.0,
      "step": 15275
    },
    {
      "entropy": 5.201593732833862,
      "epoch": 1.4678194044188282,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004790322816944879,
      "loss": 4.8358,
      "mean_token_accuracy": 0.23567743301391603,
      "num_tokens": 35035839.0,
      "step": 15280
    },
    {
      "entropy": 5.3257091522216795,
      "epoch": 1.468299711815562,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004790178274605793,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22430905997753142,
      "num_tokens": 35047823.0,
      "step": 15285
    },
    {
      "entropy": 5.288594961166382,
      "epoch": 1.468780019212296,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047900336848998254,
      "loss": 4.9352,
      "mean_token_accuracy": 0.220192813873291,
      "num_tokens": 35058597.0,
      "step": 15290
    },
    {
      "entropy": 5.190876483917236,
      "epoch": 1.4692603266090298,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004789889047830334,
      "loss": 4.7995,
      "mean_token_accuracy": 0.2339871734380722,
      "num_tokens": 35069822.0,
      "step": 15295
    },
    {
      "entropy": 5.2915503025054935,
      "epoch": 1.4697406340057637,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047897443634006766,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23177099078893662,
      "num_tokens": 35081423.0,
      "step": 15300
    },
    {
      "entropy": 5.331723403930664,
      "epoch": 1.4702209414024976,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004789599631614211,
      "loss": 4.9013,
      "mean_token_accuracy": 0.2262947827577591,
      "num_tokens": 35092565.0,
      "step": 15305
    },
    {
      "entropy": 5.250157260894776,
      "epoch": 1.4707012487992315,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004789454852474298,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23258853405714036,
      "num_tokens": 35103811.0,
      "step": 15310
    },
    {
      "entropy": 5.319003677368164,
      "epoch": 1.4711815561959654,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004789310025984299,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22451530694961547,
      "num_tokens": 35115516.0,
      "step": 15315
    },
    {
      "entropy": 5.258503580093384,
      "epoch": 1.4716618635926992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047891651521475776,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23325894325971602,
      "num_tokens": 35127285.0,
      "step": 15320
    },
    {
      "entropy": 5.262245512008667,
      "epoch": 1.4721421709894331,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047890202309674963,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23608790189027787,
      "num_tokens": 35137884.0,
      "step": 15325
    },
    {
      "entropy": 5.343718242645264,
      "epoch": 1.4726224783861672,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047888752624474195,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2206919878721237,
      "num_tokens": 35149935.0,
      "step": 15330
    },
    {
      "entropy": 5.3225232601165775,
      "epoch": 1.473102785782901,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004788730246590714,
      "loss": 4.9229,
      "mean_token_accuracy": 0.22262202799320222,
      "num_tokens": 35162610.0,
      "step": 15335
    },
    {
      "entropy": 5.263501691818237,
      "epoch": 1.473583093179635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047885851834007456,
      "loss": 4.7898,
      "mean_token_accuracy": 0.2356173112988472,
      "num_tokens": 35174799.0,
      "step": 15340
    },
    {
      "entropy": 5.241328954696655,
      "epoch": 1.4740634005763689,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047884400728808824,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23626696169376374,
      "num_tokens": 35186004.0,
      "step": 15345
    },
    {
      "entropy": 5.264436435699463,
      "epoch": 1.4745437079731027,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004788294915034494,
      "loss": 4.8294,
      "mean_token_accuracy": 0.23771029561758042,
      "num_tokens": 35197310.0,
      "step": 15350
    },
    {
      "entropy": 5.187747669219971,
      "epoch": 1.4750240153698366,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000478814970986495,
      "loss": 4.7878,
      "mean_token_accuracy": 0.23420082032680511,
      "num_tokens": 35208703.0,
      "step": 15355
    },
    {
      "entropy": 5.25417742729187,
      "epoch": 1.4755043227665707,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047880044573756213,
      "loss": 4.8011,
      "mean_token_accuracy": 0.2344979763031006,
      "num_tokens": 35219927.0,
      "step": 15360
    },
    {
      "entropy": 5.237832260131836,
      "epoch": 1.4759846301633046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047878591575698816,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23866041600704194,
      "num_tokens": 35231077.0,
      "step": 15365
    },
    {
      "entropy": 5.305797290802002,
      "epoch": 1.4764649375600385,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004787713810451103,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22529850006103516,
      "num_tokens": 35241984.0,
      "step": 15370
    },
    {
      "entropy": 5.217218923568725,
      "epoch": 1.4769452449567724,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047875684160226606,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2402344048023224,
      "num_tokens": 35252717.0,
      "step": 15375
    },
    {
      "entropy": 5.271824979782105,
      "epoch": 1.4774255523535063,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000478742297428793,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23150755763053893,
      "num_tokens": 35263916.0,
      "step": 15380
    },
    {
      "entropy": 5.259904384613037,
      "epoch": 1.4779058597502401,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047872774852502877,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23040546625852584,
      "num_tokens": 35274772.0,
      "step": 15385
    },
    {
      "entropy": 5.194039726257325,
      "epoch": 1.478386167146974,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004787131948913112,
      "loss": 4.7593,
      "mean_token_accuracy": 0.24456387311220168,
      "num_tokens": 35287150.0,
      "step": 15390
    },
    {
      "entropy": 5.235249805450439,
      "epoch": 1.478866474543708,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047869863652797806,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23730200678110122,
      "num_tokens": 35298538.0,
      "step": 15395
    },
    {
      "entropy": 5.158382844924927,
      "epoch": 1.4793467819404418,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786840734353675,
      "loss": 4.7744,
      "mean_token_accuracy": 0.23720724880695343,
      "num_tokens": 35309395.0,
      "step": 15400
    },
    {
      "entropy": 5.2161377429962155,
      "epoch": 1.4798270893371757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047866950561381756,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23521784991025924,
      "num_tokens": 35320741.0,
      "step": 15405
    },
    {
      "entropy": 5.348979568481445,
      "epoch": 1.4803073967339098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004786549330636665,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22567295879125596,
      "num_tokens": 35331895.0,
      "step": 15410
    },
    {
      "entropy": 5.190139627456665,
      "epoch": 1.4807877041306436,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047864035578525256,
      "loss": 4.718,
      "mean_token_accuracy": 0.24487824141979217,
      "num_tokens": 35343775.0,
      "step": 15415
    },
    {
      "entropy": 5.208282804489135,
      "epoch": 1.4812680115273775,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786257737789143,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22891037464141845,
      "num_tokens": 35355043.0,
      "step": 15420
    },
    {
      "entropy": 5.343260049819946,
      "epoch": 1.4817483189241114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786111870449902,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22711896896362305,
      "num_tokens": 35365387.0,
      "step": 15425
    },
    {
      "entropy": 5.292239618301392,
      "epoch": 1.4822286263208453,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047859659558381894,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23696142733097075,
      "num_tokens": 35376400.0,
      "step": 15430
    },
    {
      "entropy": 5.267343807220459,
      "epoch": 1.4827089337175792,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047858199939573935,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22983453422784805,
      "num_tokens": 35387315.0,
      "step": 15435
    },
    {
      "entropy": 5.2849527359008786,
      "epoch": 1.4831892411143133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047856739848109014,
      "loss": 4.8636,
      "mean_token_accuracy": 0.23588735461235047,
      "num_tokens": 35398666.0,
      "step": 15440
    },
    {
      "entropy": 5.25243616104126,
      "epoch": 1.4836695485110472,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047855279284021046,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23657728880643844,
      "num_tokens": 35409192.0,
      "step": 15445
    },
    {
      "entropy": 5.284908390045166,
      "epoch": 1.484149855907781,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047853818247343933,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22494743019342422,
      "num_tokens": 35419812.0,
      "step": 15450
    },
    {
      "entropy": 5.317013549804687,
      "epoch": 1.484630163304515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047852356738111606,
      "loss": 4.8191,
      "mean_token_accuracy": 0.240191225707531,
      "num_tokens": 35430875.0,
      "step": 15455
    },
    {
      "entropy": 5.237310457229614,
      "epoch": 1.4851104707012488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004785089475635799,
      "loss": 4.7942,
      "mean_token_accuracy": 0.23561248928308487,
      "num_tokens": 35441065.0,
      "step": 15460
    },
    {
      "entropy": 5.209410572052002,
      "epoch": 1.4855907780979827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047849432302117024,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22774945795536042,
      "num_tokens": 35452164.0,
      "step": 15465
    },
    {
      "entropy": 5.304575061798095,
      "epoch": 1.4860710854947166,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047847969375422656,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21905469447374343,
      "num_tokens": 35463158.0,
      "step": 15470
    },
    {
      "entropy": 5.211660289764405,
      "epoch": 1.4865513928914504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004784650597630887,
      "loss": 4.7575,
      "mean_token_accuracy": 0.2474092036485672,
      "num_tokens": 35474153.0,
      "step": 15475
    },
    {
      "entropy": 5.2754570007324215,
      "epoch": 1.4870317002881843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047845042104809635,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22958581447601317,
      "num_tokens": 35485680.0,
      "step": 15480
    },
    {
      "entropy": 5.298479223251343,
      "epoch": 1.4875120076849184,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004784357776095892,
      "loss": 4.8888,
      "mean_token_accuracy": 0.2346413552761078,
      "num_tokens": 35497271.0,
      "step": 15485
    },
    {
      "entropy": 5.2563948154449465,
      "epoch": 1.4879923150816523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004784211294479075,
      "loss": 4.8084,
      "mean_token_accuracy": 0.2373933583498001,
      "num_tokens": 35509166.0,
      "step": 15490
    },
    {
      "entropy": 5.224243021011352,
      "epoch": 1.4884726224783862,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004784064765633912,
      "loss": 4.9029,
      "mean_token_accuracy": 0.23437886387109758,
      "num_tokens": 35521289.0,
      "step": 15495
    },
    {
      "entropy": 5.282890033721924,
      "epoch": 1.48895292987512,
      "grad_norm": 1.0,
      "learning_rate": 0.00047839181895638057,
      "loss": 4.8313,
      "mean_token_accuracy": 0.2355358511209488,
      "num_tokens": 35532179.0,
      "step": 15500
    },
    {
      "entropy": 5.364083957672119,
      "epoch": 1.489433237271854,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047837715662721575,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2223569020628929,
      "num_tokens": 35544703.0,
      "step": 15505
    },
    {
      "entropy": 5.352913951873779,
      "epoch": 1.4899135446685878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004783624895762372,
      "loss": 4.9915,
      "mean_token_accuracy": 0.21763598769903184,
      "num_tokens": 35557853.0,
      "step": 15510
    },
    {
      "entropy": 5.245417213439941,
      "epoch": 1.490393852065322,
      "grad_norm": 1.125,
      "learning_rate": 0.00047834781780378563,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24271586388349534,
      "num_tokens": 35570340.0,
      "step": 15515
    },
    {
      "entropy": 5.290103244781494,
      "epoch": 1.4908741594620558,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004783331413102015,
      "loss": 4.9057,
      "mean_token_accuracy": 0.23224325776100158,
      "num_tokens": 35582387.0,
      "step": 15520
    },
    {
      "entropy": 5.352967548370361,
      "epoch": 1.4913544668587897,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047831846009582557,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22066044807434082,
      "num_tokens": 35595105.0,
      "step": 15525
    },
    {
      "entropy": 5.3005811214447025,
      "epoch": 1.4918347742555236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004783037741609988,
      "loss": 4.9274,
      "mean_token_accuracy": 0.22816851884126663,
      "num_tokens": 35607160.0,
      "step": 15530
    },
    {
      "entropy": 5.21825647354126,
      "epoch": 1.4923150816522575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004782890835060621,
      "loss": 4.7913,
      "mean_token_accuracy": 0.2388398215174675,
      "num_tokens": 35619097.0,
      "step": 15535
    },
    {
      "entropy": 5.161506319046021,
      "epoch": 1.4927953890489913,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004782743881313564,
      "loss": 4.7087,
      "mean_token_accuracy": 0.23903133273124694,
      "num_tokens": 35629868.0,
      "step": 15540
    },
    {
      "entropy": 5.24497013092041,
      "epoch": 1.4932756964457252,
      "grad_norm": 1.0,
      "learning_rate": 0.00047825968803722315,
      "loss": 4.8664,
      "mean_token_accuracy": 0.23258682638406752,
      "num_tokens": 35640622.0,
      "step": 15545
    },
    {
      "entropy": 5.267886543273926,
      "epoch": 1.493756003842459,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004782449832240035,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22740030586719512,
      "num_tokens": 35652383.0,
      "step": 15550
    },
    {
      "entropy": 5.1930908203125,
      "epoch": 1.494236311239193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004782302736920387,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23100065141916276,
      "num_tokens": 35663838.0,
      "step": 15555
    },
    {
      "entropy": 5.249733638763428,
      "epoch": 1.4947166186359269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004782155594416705,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23150836378335954,
      "num_tokens": 35674564.0,
      "step": 15560
    },
    {
      "entropy": 5.346973848342896,
      "epoch": 1.495196926032661,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047820084047324045,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23370860815048217,
      "num_tokens": 35685518.0,
      "step": 15565
    },
    {
      "entropy": 5.344891119003296,
      "epoch": 1.4956772334293948,
      "grad_norm": 1.0,
      "learning_rate": 0.00047818611678709027,
      "loss": 4.8629,
      "mean_token_accuracy": 0.23143185824155807,
      "num_tokens": 35696597.0,
      "step": 15570
    },
    {
      "entropy": 5.207191467285156,
      "epoch": 1.4961575408261287,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004781713883835618,
      "loss": 4.7813,
      "mean_token_accuracy": 0.23726969808340073,
      "num_tokens": 35707229.0,
      "step": 15575
    },
    {
      "entropy": 5.228154706954956,
      "epoch": 1.4966378482228626,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047815665526299695,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22906045168638228,
      "num_tokens": 35719440.0,
      "step": 15580
    },
    {
      "entropy": 5.313417959213257,
      "epoch": 1.4971181556195965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004781419174257378,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22624800503253936,
      "num_tokens": 35731611.0,
      "step": 15585
    },
    {
      "entropy": 5.231529617309571,
      "epoch": 1.4975984630163304,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004781271748721266,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2308451935648918,
      "num_tokens": 35743047.0,
      "step": 15590
    },
    {
      "entropy": 5.275323915481567,
      "epoch": 1.4980787704130645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004781124276025055,
      "loss": 4.7783,
      "mean_token_accuracy": 0.25052398145198823,
      "num_tokens": 35753499.0,
      "step": 15595
    },
    {
      "entropy": 5.29826283454895,
      "epoch": 1.4985590778097984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000478097675617217,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2246858924627304,
      "num_tokens": 35764682.0,
      "step": 15600
    },
    {
      "entropy": 5.330998325347901,
      "epoch": 1.4990393852065322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047808291891660357,
      "loss": 4.8162,
      "mean_token_accuracy": 0.2444232851266861,
      "num_tokens": 35775160.0,
      "step": 15605
    },
    {
      "entropy": 5.326831912994384,
      "epoch": 1.4995196926032661,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047806815750100774,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22930117845535278,
      "num_tokens": 35786089.0,
      "step": 15610
    },
    {
      "entropy": 5.27262921333313,
      "epoch": 1.5,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004780533913707723,
      "loss": 4.8726,
      "mean_token_accuracy": 0.23106993287801741,
      "num_tokens": 35796851.0,
      "step": 15615
    },
    {
      "entropy": 5.280172634124756,
      "epoch": 1.5004803073967339,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047803862052624006,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22755674421787261,
      "num_tokens": 35808553.0,
      "step": 15620
    },
    {
      "entropy": 5.396438407897949,
      "epoch": 1.5009606147934678,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047802384496775397,
      "loss": 5.027,
      "mean_token_accuracy": 0.2271316319704056,
      "num_tokens": 35820108.0,
      "step": 15625
    },
    {
      "entropy": 5.242713403701782,
      "epoch": 1.5014409221902016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004780090646956571,
      "loss": 4.7642,
      "mean_token_accuracy": 0.24122214019298555,
      "num_tokens": 35831672.0,
      "step": 15630
    },
    {
      "entropy": 5.3358818054199215,
      "epoch": 1.5019212295869355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047799427971029245,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22458840906620026,
      "num_tokens": 35843164.0,
      "step": 15635
    },
    {
      "entropy": 5.313462400436402,
      "epoch": 1.5024015369836694,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004779794900120034,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2226031631231308,
      "num_tokens": 35854677.0,
      "step": 15640
    },
    {
      "entropy": 5.299949407577515,
      "epoch": 1.5028818443804035,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004779646956011334,
      "loss": 4.8485,
      "mean_token_accuracy": 0.23317965716123581,
      "num_tokens": 35865956.0,
      "step": 15645
    },
    {
      "entropy": 5.265274953842163,
      "epoch": 1.5033621517771374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047794989647802574,
      "loss": 4.842,
      "mean_token_accuracy": 0.23127577304840088,
      "num_tokens": 35877451.0,
      "step": 15650
    },
    {
      "entropy": 5.24707522392273,
      "epoch": 1.5038424591738713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047793509264302424,
      "loss": 4.8416,
      "mean_token_accuracy": 0.23324955254793167,
      "num_tokens": 35888436.0,
      "step": 15655
    },
    {
      "entropy": 5.26676459312439,
      "epoch": 1.5043227665706052,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047792028409647237,
      "loss": 4.8657,
      "mean_token_accuracy": 0.23046163320541382,
      "num_tokens": 35901010.0,
      "step": 15660
    },
    {
      "entropy": 5.313957214355469,
      "epoch": 1.5048030739673393,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047790547083871414,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22704499810934067,
      "num_tokens": 35912697.0,
      "step": 15665
    },
    {
      "entropy": 5.236976671218872,
      "epoch": 1.5052833813640731,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047789065287009335,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23769291937351228,
      "num_tokens": 35924614.0,
      "step": 15670
    },
    {
      "entropy": 5.284751462936401,
      "epoch": 1.505763688760807,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004778758301909542,
      "loss": 4.8737,
      "mean_token_accuracy": 0.23043350726366044,
      "num_tokens": 35935492.0,
      "step": 15675
    },
    {
      "entropy": 5.193773937225342,
      "epoch": 1.506243996157541,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004778610028016405,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2347308576107025,
      "num_tokens": 35947901.0,
      "step": 15680
    },
    {
      "entropy": 5.199258136749267,
      "epoch": 1.5067243035542748,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004778461707024967,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24339355528354645,
      "num_tokens": 35959690.0,
      "step": 15685
    },
    {
      "entropy": 5.263948965072632,
      "epoch": 1.5072046109510087,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004778313338938672,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2397877350449562,
      "num_tokens": 35971209.0,
      "step": 15690
    },
    {
      "entropy": 5.220327949523925,
      "epoch": 1.5076849183477425,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047781649237609643,
      "loss": 4.7215,
      "mean_token_accuracy": 0.23729157745838164,
      "num_tokens": 35981795.0,
      "step": 15695
    },
    {
      "entropy": 5.241696691513061,
      "epoch": 1.5081652257444764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004778016461495289,
      "loss": 4.8433,
      "mean_token_accuracy": 0.2334246575832367,
      "num_tokens": 35993358.0,
      "step": 15700
    },
    {
      "entropy": 5.269581842422485,
      "epoch": 1.5086455331412103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004777867952145094,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23468463271856307,
      "num_tokens": 36005353.0,
      "step": 15705
    },
    {
      "entropy": 5.2916826725006105,
      "epoch": 1.5091258405379442,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004777719395713826,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23007266223430634,
      "num_tokens": 36017510.0,
      "step": 15710
    },
    {
      "entropy": 5.259425926208496,
      "epoch": 1.509606147934678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047775707922049354,
      "loss": 4.8063,
      "mean_token_accuracy": 0.23540484756231309,
      "num_tokens": 36028687.0,
      "step": 15715
    },
    {
      "entropy": 5.218069267272949,
      "epoch": 1.510086455331412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004777422141621871,
      "loss": 4.7166,
      "mean_token_accuracy": 0.2395255148410797,
      "num_tokens": 36039851.0,
      "step": 15720
    },
    {
      "entropy": 5.305160665512085,
      "epoch": 1.510566762728146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004777273443968085,
      "loss": 4.9345,
      "mean_token_accuracy": 0.23160575181245804,
      "num_tokens": 36050776.0,
      "step": 15725
    },
    {
      "entropy": 5.185805892944336,
      "epoch": 1.51104707012488,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004777124699247029,
      "loss": 4.7454,
      "mean_token_accuracy": 0.23966709077358245,
      "num_tokens": 36062151.0,
      "step": 15730
    },
    {
      "entropy": 5.254261827468872,
      "epoch": 1.5115273775216138,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004776975907462157,
      "loss": 4.8746,
      "mean_token_accuracy": 0.23475838005542754,
      "num_tokens": 36074601.0,
      "step": 15735
    },
    {
      "entropy": 5.2465901374816895,
      "epoch": 1.5120076849183477,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004776827068616924,
      "loss": 4.7637,
      "mean_token_accuracy": 0.2402043178677559,
      "num_tokens": 36085436.0,
      "step": 15740
    },
    {
      "entropy": 5.35146164894104,
      "epoch": 1.5124879923150818,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004776678182714785,
      "loss": 4.9351,
      "mean_token_accuracy": 0.22033382654190065,
      "num_tokens": 36097589.0,
      "step": 15745
    },
    {
      "entropy": 5.2620538711547855,
      "epoch": 1.5129682997118157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047765292497591955,
      "loss": 4.8503,
      "mean_token_accuracy": 0.2342249035835266,
      "num_tokens": 36109993.0,
      "step": 15750
    },
    {
      "entropy": 5.24609341621399,
      "epoch": 1.5134486071085496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047763802697536146,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23118275851011277,
      "num_tokens": 36122439.0,
      "step": 15755
    },
    {
      "entropy": 5.274936246871948,
      "epoch": 1.5139289145052834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047762312427015015,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23140279352664947,
      "num_tokens": 36133867.0,
      "step": 15760
    },
    {
      "entropy": 5.34727373123169,
      "epoch": 1.5144092219020173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047760821686063153,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22632501125335694,
      "num_tokens": 36144202.0,
      "step": 15765
    },
    {
      "entropy": 5.285828018188477,
      "epoch": 1.5148895292987512,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047759330474715173,
      "loss": 4.8276,
      "mean_token_accuracy": 0.23273420482873916,
      "num_tokens": 36154863.0,
      "step": 15770
    },
    {
      "entropy": 5.310642242431641,
      "epoch": 1.515369836695485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047757838793005704,
      "loss": 4.864,
      "mean_token_accuracy": 0.22782388627529143,
      "num_tokens": 36166360.0,
      "step": 15775
    },
    {
      "entropy": 5.215022325515747,
      "epoch": 1.515850144092219,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047756346640969366,
      "loss": 4.7581,
      "mean_token_accuracy": 0.24157615303993224,
      "num_tokens": 36177477.0,
      "step": 15780
    },
    {
      "entropy": 5.292706680297852,
      "epoch": 1.5163304514889528,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047754854018640803,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22950258553028108,
      "num_tokens": 36188510.0,
      "step": 15785
    },
    {
      "entropy": 5.257551288604736,
      "epoch": 1.5168107588856867,
      "grad_norm": 1.0,
      "learning_rate": 0.00047753360926054684,
      "loss": 4.7826,
      "mean_token_accuracy": 0.23785875588655472,
      "num_tokens": 36199084.0,
      "step": 15790
    },
    {
      "entropy": 5.196926784515381,
      "epoch": 1.5172910662824206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047751867363245653,
      "loss": 4.8072,
      "mean_token_accuracy": 0.23104542046785354,
      "num_tokens": 36211265.0,
      "step": 15795
    },
    {
      "entropy": 5.4017181396484375,
      "epoch": 1.5177713736791547,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004775037333024841,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22230798453092576,
      "num_tokens": 36223070.0,
      "step": 15800
    },
    {
      "entropy": 5.303314924240112,
      "epoch": 1.5182516810758886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004774887882709762,
      "loss": 4.8825,
      "mean_token_accuracy": 0.23698771744966507,
      "num_tokens": 36235021.0,
      "step": 15805
    },
    {
      "entropy": 5.264981031417847,
      "epoch": 1.5187319884726225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047747383853827995,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22941047847270965,
      "num_tokens": 36245647.0,
      "step": 15810
    },
    {
      "entropy": 5.303854084014892,
      "epoch": 1.5192122958693564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004774588841047424,
      "loss": 4.9201,
      "mean_token_accuracy": 0.22997910976409913,
      "num_tokens": 36257470.0,
      "step": 15815
    },
    {
      "entropy": 5.250058650970459,
      "epoch": 1.5196926032660905,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004774439249707108,
      "loss": 4.7874,
      "mean_token_accuracy": 0.2370414525270462,
      "num_tokens": 36268839.0,
      "step": 15820
    },
    {
      "entropy": 5.317734146118164,
      "epoch": 1.5201729106628243,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004774289611365323,
      "loss": 4.9079,
      "mean_token_accuracy": 0.2260454222559929,
      "num_tokens": 36280624.0,
      "step": 15825
    },
    {
      "entropy": 5.2899007320404055,
      "epoch": 1.5206532180595582,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047741399260255434,
      "loss": 4.8259,
      "mean_token_accuracy": 0.2321329653263092,
      "num_tokens": 36292696.0,
      "step": 15830
    },
    {
      "entropy": 5.291850280761719,
      "epoch": 1.521133525456292,
      "grad_norm": 1.0,
      "learning_rate": 0.00047739901936912467,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23720038533210755,
      "num_tokens": 36303612.0,
      "step": 15835
    },
    {
      "entropy": 5.202324342727661,
      "epoch": 1.521613832853026,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004773840414365907,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2351069211959839,
      "num_tokens": 36314511.0,
      "step": 15840
    },
    {
      "entropy": 5.345182180404663,
      "epoch": 1.5220941402497599,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047736905880530026,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22437727600336074,
      "num_tokens": 36327276.0,
      "step": 15845
    },
    {
      "entropy": 5.30887360572815,
      "epoch": 1.5225744476464937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004773540714756012,
      "loss": 4.7999,
      "mean_token_accuracy": 0.23397299647331238,
      "num_tokens": 36339373.0,
      "step": 15850
    },
    {
      "entropy": 5.3143352508544925,
      "epoch": 1.5230547550432276,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047733907944784144,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22212078422307968,
      "num_tokens": 36351863.0,
      "step": 15855
    },
    {
      "entropy": 5.256155204772949,
      "epoch": 1.5235350624399615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004773240827223691,
      "loss": 4.8272,
      "mean_token_accuracy": 0.23712152391672134,
      "num_tokens": 36363961.0,
      "step": 15860
    },
    {
      "entropy": 5.235923910140992,
      "epoch": 1.5240153698366954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004773090812995323,
      "loss": 4.8247,
      "mean_token_accuracy": 0.24120467603206636,
      "num_tokens": 36374738.0,
      "step": 15865
    },
    {
      "entropy": 5.241990566253662,
      "epoch": 1.5244956772334293,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047729407517967945,
      "loss": 4.7398,
      "mean_token_accuracy": 0.23975107371807097,
      "num_tokens": 36386472.0,
      "step": 15870
    },
    {
      "entropy": 5.270555782318115,
      "epoch": 1.5249759846301632,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047727906436315884,
      "loss": 4.854,
      "mean_token_accuracy": 0.24118195325136185,
      "num_tokens": 36397042.0,
      "step": 15875
    },
    {
      "entropy": 5.250922060012817,
      "epoch": 1.5254562920268973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047726404885031895,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23758921772241592,
      "num_tokens": 36408720.0,
      "step": 15880
    },
    {
      "entropy": 5.2538457870483395,
      "epoch": 1.5259365994236311,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047724902864150845,
      "loss": 4.8884,
      "mean_token_accuracy": 0.23163739889860152,
      "num_tokens": 36420885.0,
      "step": 15885
    },
    {
      "entropy": 5.3196310043334964,
      "epoch": 1.526416906820365,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047723400373707607,
      "loss": 4.8964,
      "mean_token_accuracy": 0.22379076480865479,
      "num_tokens": 36433678.0,
      "step": 15890
    },
    {
      "entropy": 5.324388599395752,
      "epoch": 1.526897214217099,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004772189741373707,
      "loss": 4.9068,
      "mean_token_accuracy": 0.23037031441926956,
      "num_tokens": 36445143.0,
      "step": 15895
    },
    {
      "entropy": 5.210389947891235,
      "epoch": 1.527377521613833,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047720393984274117,
      "loss": 4.8135,
      "mean_token_accuracy": 0.23132385313510895,
      "num_tokens": 36456214.0,
      "step": 15900
    },
    {
      "entropy": 5.294915199279785,
      "epoch": 1.5278578290105669,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047718890085353654,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23230967819690704,
      "num_tokens": 36466767.0,
      "step": 15905
    },
    {
      "entropy": 5.315571975708008,
      "epoch": 1.5283381364073008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477173857170106,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23581524640321733,
      "num_tokens": 36478150.0,
      "step": 15910
    },
    {
      "entropy": 5.260588884353638,
      "epoch": 1.5288184438040346,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047715880879279894,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22552652806043624,
      "num_tokens": 36488280.0,
      "step": 15915
    },
    {
      "entropy": 5.355570936203003,
      "epoch": 1.5292987512007685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004771437557219646,
      "loss": 4.8871,
      "mean_token_accuracy": 0.22926830500364304,
      "num_tokens": 36498505.0,
      "step": 15920
    },
    {
      "entropy": 5.254315185546875,
      "epoch": 1.5297790585975024,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004771286979579524,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2313305228948593,
      "num_tokens": 36509271.0,
      "step": 15925
    },
    {
      "entropy": 5.283911418914795,
      "epoch": 1.5302593659942363,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004771136355011121,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22682373970746994,
      "num_tokens": 36520022.0,
      "step": 15930
    },
    {
      "entropy": 5.262305498123169,
      "epoch": 1.5307396733909702,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047709856835179333,
      "loss": 4.8265,
      "mean_token_accuracy": 0.23469631224870682,
      "num_tokens": 36532769.0,
      "step": 15935
    },
    {
      "entropy": 5.339139795303344,
      "epoch": 1.531219980787704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047708349651034586,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2389321893453598,
      "num_tokens": 36544454.0,
      "step": 15940
    },
    {
      "entropy": 5.264775705337525,
      "epoch": 1.531700288184438,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047706841997711974,
      "loss": 4.8452,
      "mean_token_accuracy": 0.23283364325761796,
      "num_tokens": 36555916.0,
      "step": 15945
    },
    {
      "entropy": 5.2446037292480465,
      "epoch": 1.5321805955811718,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047705333875246495,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23109405636787414,
      "num_tokens": 36567829.0,
      "step": 15950
    },
    {
      "entropy": 5.312460708618164,
      "epoch": 1.532660902977906,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047703825283673153,
      "loss": 4.9276,
      "mean_token_accuracy": 0.23166270405054093,
      "num_tokens": 36578216.0,
      "step": 15955
    },
    {
      "entropy": 5.31017575263977,
      "epoch": 1.5331412103746398,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004770231622302699,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23334792405366897,
      "num_tokens": 36589945.0,
      "step": 15960
    },
    {
      "entropy": 5.349650478363037,
      "epoch": 1.5336215177713737,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047700806693343016,
      "loss": 4.862,
      "mean_token_accuracy": 0.23086611926555634,
      "num_tokens": 36600724.0,
      "step": 15965
    },
    {
      "entropy": 5.197714614868164,
      "epoch": 1.5341018251681076,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047699296694656316,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24501417130231856,
      "num_tokens": 36611154.0,
      "step": 15970
    },
    {
      "entropy": 5.254950094223022,
      "epoch": 1.5345821325648417,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004769778622700192,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23695746809244156,
      "num_tokens": 36621750.0,
      "step": 15975
    },
    {
      "entropy": 5.313170480728149,
      "epoch": 1.5350624399615755,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047696275290414885,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23439727872610092,
      "num_tokens": 36633294.0,
      "step": 15980
    },
    {
      "entropy": 5.224903392791748,
      "epoch": 1.5355427473583094,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047694763884930324,
      "loss": 4.7779,
      "mean_token_accuracy": 0.23291939944028855,
      "num_tokens": 36646377.0,
      "step": 15985
    },
    {
      "entropy": 5.198040390014649,
      "epoch": 1.5360230547550433,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047693252010583314,
      "loss": 4.7357,
      "mean_token_accuracy": 0.23867324292659758,
      "num_tokens": 36656159.0,
      "step": 15990
    },
    {
      "entropy": 5.202288484573364,
      "epoch": 1.5365033621517772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004769173966740895,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2236504152417183,
      "num_tokens": 36667522.0,
      "step": 15995
    },
    {
      "entropy": 5.354539680480957,
      "epoch": 1.536983669548511,
      "grad_norm": 1.0,
      "learning_rate": 0.00047690226855442346,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22752151638269424,
      "num_tokens": 36678662.0,
      "step": 16000
    },
    {
      "entropy": 5.310584640502929,
      "epoch": 1.537463976945245,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004768871357471863,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22984777837991716,
      "num_tokens": 36689876.0,
      "step": 16005
    },
    {
      "entropy": 5.27028341293335,
      "epoch": 1.5379442843419788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047687199825272936,
      "loss": 4.801,
      "mean_token_accuracy": 0.23628502190113068,
      "num_tokens": 36701140.0,
      "step": 16010
    },
    {
      "entropy": 5.324418020248413,
      "epoch": 1.5384245917387127,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047685685607140403,
      "loss": 4.9457,
      "mean_token_accuracy": 0.22756727039813995,
      "num_tokens": 36711077.0,
      "step": 16015
    },
    {
      "entropy": 5.312882137298584,
      "epoch": 1.5389048991354466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047684170920356185,
      "loss": 4.8254,
      "mean_token_accuracy": 0.2380486324429512,
      "num_tokens": 36722705.0,
      "step": 16020
    },
    {
      "entropy": 5.257938623428345,
      "epoch": 1.5393852065321805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004768265576495546,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23663015961647033,
      "num_tokens": 36733446.0,
      "step": 16025
    },
    {
      "entropy": 5.218319940567016,
      "epoch": 1.5398655139289144,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047681140140973396,
      "loss": 4.8104,
      "mean_token_accuracy": 0.23721818923950194,
      "num_tokens": 36744529.0,
      "step": 16030
    },
    {
      "entropy": 5.160533046722412,
      "epoch": 1.5403458213256485,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004767962404844517,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23624338656663896,
      "num_tokens": 36756248.0,
      "step": 16035
    },
    {
      "entropy": 5.350418138504028,
      "epoch": 1.5408261287223823,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047678107487406015,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2277960315346718,
      "num_tokens": 36768097.0,
      "step": 16040
    },
    {
      "entropy": 5.303984498977661,
      "epoch": 1.5413064361191162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047676590457891116,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23309519588947297,
      "num_tokens": 36780395.0,
      "step": 16045
    },
    {
      "entropy": 5.248359107971192,
      "epoch": 1.54178674351585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004767507295993569,
      "loss": 4.8102,
      "mean_token_accuracy": 0.2298466831445694,
      "num_tokens": 36791897.0,
      "step": 16050
    },
    {
      "entropy": 5.275026082992554,
      "epoch": 1.5422670509125842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004767355499357498,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23677476197481157,
      "num_tokens": 36802716.0,
      "step": 16055
    },
    {
      "entropy": 5.284662389755249,
      "epoch": 1.542747358309318,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004767203655884423,
      "loss": 4.871,
      "mean_token_accuracy": 0.2305197536945343,
      "num_tokens": 36814471.0,
      "step": 16060
    },
    {
      "entropy": 5.236914157867432,
      "epoch": 1.543227665706052,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004767051765577869,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23969156593084334,
      "num_tokens": 36825975.0,
      "step": 16065
    },
    {
      "entropy": 5.315780353546143,
      "epoch": 1.5437079731027858,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047668998284413624,
      "loss": 4.9249,
      "mean_token_accuracy": 0.223564749956131,
      "num_tokens": 36837990.0,
      "step": 16070
    },
    {
      "entropy": 5.300530004501343,
      "epoch": 1.5441882804995197,
      "grad_norm": 1.0,
      "learning_rate": 0.00047667478444784306,
      "loss": 4.9106,
      "mean_token_accuracy": 0.22756724059581757,
      "num_tokens": 36849115.0,
      "step": 16075
    },
    {
      "entropy": 5.286770534515381,
      "epoch": 1.5446685878962536,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004766595813692602,
      "loss": 4.7433,
      "mean_token_accuracy": 0.23503902852535247,
      "num_tokens": 36860626.0,
      "step": 16080
    },
    {
      "entropy": 5.304471826553344,
      "epoch": 1.5451488952929875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047664437360874076,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22475566118955612,
      "num_tokens": 36871926.0,
      "step": 16085
    },
    {
      "entropy": 5.257347249984742,
      "epoch": 1.5456292026897214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047662916116663766,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23676751106977462,
      "num_tokens": 36883511.0,
      "step": 16090
    },
    {
      "entropy": 5.3757298469543455,
      "epoch": 1.5461095100864553,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047661394404330417,
      "loss": 4.9033,
      "mean_token_accuracy": 0.23180662393569945,
      "num_tokens": 36895468.0,
      "step": 16095
    },
    {
      "entropy": 5.266822242736817,
      "epoch": 1.5465898174831891,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047659872223909357,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23456798642873763,
      "num_tokens": 36906957.0,
      "step": 16100
    },
    {
      "entropy": 5.2664590835571286,
      "epoch": 1.547070124879923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004765834957543592,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23553932905197145,
      "num_tokens": 36916495.0,
      "step": 16105
    },
    {
      "entropy": 5.240660238265991,
      "epoch": 1.547550432276657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047656826458945475,
      "loss": 4.8136,
      "mean_token_accuracy": 0.22973976731300355,
      "num_tokens": 36927301.0,
      "step": 16110
    },
    {
      "entropy": 5.257215213775635,
      "epoch": 1.548030739673391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047655302874473365,
      "loss": 4.7851,
      "mean_token_accuracy": 0.23472480177879335,
      "num_tokens": 36938116.0,
      "step": 16115
    },
    {
      "entropy": 5.3371506690979,
      "epoch": 1.5485110470701249,
      "grad_norm": 1.0,
      "learning_rate": 0.0004765377882205498,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23839707672595978,
      "num_tokens": 36950292.0,
      "step": 16120
    },
    {
      "entropy": 5.190141677856445,
      "epoch": 1.5489913544668588,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004765225430172568,
      "loss": 4.859,
      "mean_token_accuracy": 0.2358293801546097,
      "num_tokens": 36962017.0,
      "step": 16125
    },
    {
      "entropy": 5.2972370147705075,
      "epoch": 1.5494716618635929,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004765072931352089,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2212133064866066,
      "num_tokens": 36972831.0,
      "step": 16130
    },
    {
      "entropy": 5.280661678314209,
      "epoch": 1.5499519692603267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047649203857476,
      "loss": 4.7734,
      "mean_token_accuracy": 0.23957034647464753,
      "num_tokens": 36983324.0,
      "step": 16135
    },
    {
      "entropy": 5.3072587013244625,
      "epoch": 1.5504322766570606,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047647677933626423,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21827882826328276,
      "num_tokens": 36996093.0,
      "step": 16140
    },
    {
      "entropy": 5.277344846725464,
      "epoch": 1.5509125840537945,
      "grad_norm": 1.0,
      "learning_rate": 0.00047646151542007583,
      "loss": 4.8231,
      "mean_token_accuracy": 0.23959118276834487,
      "num_tokens": 37006548.0,
      "step": 16145
    },
    {
      "entropy": 5.268341016769409,
      "epoch": 1.5513928914505284,
      "grad_norm": 1.0,
      "learning_rate": 0.0004764462468265494,
      "loss": 4.8333,
      "mean_token_accuracy": 0.23140914887189865,
      "num_tokens": 37019181.0,
      "step": 16150
    },
    {
      "entropy": 5.316660451889038,
      "epoch": 1.5518731988472623,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047643097355603913,
      "loss": 4.8744,
      "mean_token_accuracy": 0.23096304386854172,
      "num_tokens": 37029808.0,
      "step": 16155
    },
    {
      "entropy": 5.287496185302734,
      "epoch": 1.5523535062439962,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004764156956088999,
      "loss": 4.8387,
      "mean_token_accuracy": 0.2293274760246277,
      "num_tokens": 37040352.0,
      "step": 16160
    },
    {
      "entropy": 5.249794912338257,
      "epoch": 1.55283381364073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004764004129854863,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2368505597114563,
      "num_tokens": 37050780.0,
      "step": 16165
    },
    {
      "entropy": 5.26005859375,
      "epoch": 1.553314121037464,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047638512568615307,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23668075501918792,
      "num_tokens": 37062608.0,
      "step": 16170
    },
    {
      "entropy": 5.2944869041442875,
      "epoch": 1.5537944284341978,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004763698337112553,
      "loss": 4.9119,
      "mean_token_accuracy": 0.22292884439229965,
      "num_tokens": 37074578.0,
      "step": 16175
    },
    {
      "entropy": 5.324888753890991,
      "epoch": 1.5542747358309317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004763545370611479,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2288072422146797,
      "num_tokens": 37086487.0,
      "step": 16180
    },
    {
      "entropy": 5.269550323486328,
      "epoch": 1.5547550432276656,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047633923573618605,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23912712931632996,
      "num_tokens": 37097249.0,
      "step": 16185
    },
    {
      "entropy": 5.375278520584106,
      "epoch": 1.5552353506243997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000476323929736725,
      "loss": 5.0065,
      "mean_token_accuracy": 0.21976594030857086,
      "num_tokens": 37109274.0,
      "step": 16190
    },
    {
      "entropy": 5.25685977935791,
      "epoch": 1.5557156580211335,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047630861906312004,
      "loss": 4.8156,
      "mean_token_accuracy": 0.23738760501146317,
      "num_tokens": 37120100.0,
      "step": 16195
    },
    {
      "entropy": 5.252499389648437,
      "epoch": 1.5561959654178674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004762933037157268,
      "loss": 4.8163,
      "mean_token_accuracy": 0.2296218156814575,
      "num_tokens": 37131889.0,
      "step": 16200
    },
    {
      "entropy": 5.212792873382568,
      "epoch": 1.5566762728146013,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047627798369490076,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23912741541862487,
      "num_tokens": 37142538.0,
      "step": 16205
    },
    {
      "entropy": 5.3538085460662845,
      "epoch": 1.5571565802113354,
      "grad_norm": 1.0,
      "learning_rate": 0.00047626265900099757,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22720673978328704,
      "num_tokens": 37153639.0,
      "step": 16210
    },
    {
      "entropy": 5.3181867599487305,
      "epoch": 1.5576368876080693,
      "grad_norm": 1.625,
      "learning_rate": 0.00047624732963437314,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22831140607595443,
      "num_tokens": 37164029.0,
      "step": 16215
    },
    {
      "entropy": 5.2662969589233395,
      "epoch": 1.5581171950048032,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047623199559538324,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22685605138540268,
      "num_tokens": 37175880.0,
      "step": 16220
    },
    {
      "entropy": 5.3094017028808596,
      "epoch": 1.558597502401537,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476216656884384,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23264127522706984,
      "num_tokens": 37187483.0,
      "step": 16225
    },
    {
      "entropy": 5.278099012374878,
      "epoch": 1.559077809798271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047620131350173135,
      "loss": 4.8701,
      "mean_token_accuracy": 0.2326253369450569,
      "num_tokens": 37198978.0,
      "step": 16230
    },
    {
      "entropy": 5.33046326637268,
      "epoch": 1.5595581171950048,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004761859654477817,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2177530899643898,
      "num_tokens": 37210054.0,
      "step": 16235
    },
    {
      "entropy": 5.313302850723266,
      "epoch": 1.5600384245917387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004761706127228914,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23628802448511124,
      "num_tokens": 37221744.0,
      "step": 16240
    },
    {
      "entropy": 5.248558759689331,
      "epoch": 1.5605187319884726,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004761552553274168,
      "loss": 4.8782,
      "mean_token_accuracy": 0.23496740460395812,
      "num_tokens": 37234629.0,
      "step": 16245
    },
    {
      "entropy": 5.359723234176636,
      "epoch": 1.5609990393852065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004761398932617144,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22459383457899093,
      "num_tokens": 37246960.0,
      "step": 16250
    },
    {
      "entropy": 5.268073177337646,
      "epoch": 1.5614793467819403,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000476124526526141,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2322526603937149,
      "num_tokens": 37259326.0,
      "step": 16255
    },
    {
      "entropy": 5.2438233375549315,
      "epoch": 1.5619596541786742,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047610915512105327,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23051901459693908,
      "num_tokens": 37271436.0,
      "step": 16260
    },
    {
      "entropy": 5.304201412200928,
      "epoch": 1.562439961575408,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760937790468082,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22077188342809678,
      "num_tokens": 37283230.0,
      "step": 16265
    },
    {
      "entropy": 5.276275730133056,
      "epoch": 1.5629202689721422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004760783983037627,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23933809846639634,
      "num_tokens": 37295185.0,
      "step": 16270
    },
    {
      "entropy": 5.3030143737792965,
      "epoch": 1.563400576368876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760630128922738,
      "loss": 4.8778,
      "mean_token_accuracy": 0.23150466680526732,
      "num_tokens": 37307573.0,
      "step": 16275
    },
    {
      "entropy": 5.357853984832763,
      "epoch": 1.56388088376561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004760476228126989,
      "loss": 4.916,
      "mean_token_accuracy": 0.22489488273859023,
      "num_tokens": 37319084.0,
      "step": 16280
    },
    {
      "entropy": 5.208237504959106,
      "epoch": 1.5643611911623438,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004760322280653951,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23422920554876328,
      "num_tokens": 37330447.0,
      "step": 16285
    },
    {
      "entropy": 5.243499612808227,
      "epoch": 1.564841498559078,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004760168286507199,
      "loss": 4.7307,
      "mean_token_accuracy": 0.23203796446323394,
      "num_tokens": 37341184.0,
      "step": 16290
    },
    {
      "entropy": 5.3117663860321045,
      "epoch": 1.5653218059558118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047600142456903085,
      "loss": 4.8635,
      "mean_token_accuracy": 0.23007007390260698,
      "num_tokens": 37352712.0,
      "step": 16295
    },
    {
      "entropy": 5.308646774291992,
      "epoch": 1.5658021133525457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047598601582068555,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22739534080028534,
      "num_tokens": 37365001.0,
      "step": 16300
    },
    {
      "entropy": 5.368648290634155,
      "epoch": 1.5662824207492796,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004759706024060418,
      "loss": 4.9895,
      "mean_token_accuracy": 0.2216930478811264,
      "num_tokens": 37377005.0,
      "step": 16305
    },
    {
      "entropy": 5.293943738937378,
      "epoch": 1.5667627281460135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004759551843254575,
      "loss": 4.8264,
      "mean_token_accuracy": 0.23121322989463805,
      "num_tokens": 37387329.0,
      "step": 16310
    },
    {
      "entropy": 5.37990403175354,
      "epoch": 1.5672430355427474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047593976157929034,
      "loss": 4.8974,
      "mean_token_accuracy": 0.2317206934094429,
      "num_tokens": 37398894.0,
      "step": 16315
    },
    {
      "entropy": 5.299487924575805,
      "epoch": 1.5677233429394812,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004759243341678987,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23094695508480073,
      "num_tokens": 37409913.0,
      "step": 16320
    },
    {
      "entropy": 5.298211145401001,
      "epoch": 1.5682036503362151,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004759089020916407,
      "loss": 4.7913,
      "mean_token_accuracy": 0.23022565245628357,
      "num_tokens": 37421859.0,
      "step": 16325
    },
    {
      "entropy": 5.28982367515564,
      "epoch": 1.568683957732949,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047589346535087444,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22880435585975648,
      "num_tokens": 37432827.0,
      "step": 16330
    },
    {
      "entropy": 5.278872919082642,
      "epoch": 1.5691642651296829,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004758780239459586,
      "loss": 4.8672,
      "mean_token_accuracy": 0.23573104441165924,
      "num_tokens": 37444171.0,
      "step": 16335
    },
    {
      "entropy": 5.26825270652771,
      "epoch": 1.5696445725264168,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004758625778772514,
      "loss": 4.8309,
      "mean_token_accuracy": 0.2266126751899719,
      "num_tokens": 37455194.0,
      "step": 16340
    },
    {
      "entropy": 5.1957377910614015,
      "epoch": 1.5701248799231509,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047584712714511166,
      "loss": 4.7626,
      "mean_token_accuracy": 0.2436373308300972,
      "num_tokens": 37465839.0,
      "step": 16345
    },
    {
      "entropy": 5.270568132400513,
      "epoch": 1.5706051873198847,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047583167174989797,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2338510498404503,
      "num_tokens": 37476948.0,
      "step": 16350
    },
    {
      "entropy": 5.30244197845459,
      "epoch": 1.5710854947166186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004758162116919692,
      "loss": 4.8409,
      "mean_token_accuracy": 0.23162637203931807,
      "num_tokens": 37487382.0,
      "step": 16355
    },
    {
      "entropy": 5.233168840408325,
      "epoch": 1.5715658021133525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047580074697168434,
      "loss": 4.7988,
      "mean_token_accuracy": 0.24013985246419906,
      "num_tokens": 37498347.0,
      "step": 16360
    },
    {
      "entropy": 5.287282943725586,
      "epoch": 1.5720461095100866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047578527758940236,
      "loss": 4.903,
      "mean_token_accuracy": 0.22592894583940507,
      "num_tokens": 37509620.0,
      "step": 16365
    },
    {
      "entropy": 5.283927488327026,
      "epoch": 1.5725264169068205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004757698035454825,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23873668015003205,
      "num_tokens": 37519768.0,
      "step": 16370
    },
    {
      "entropy": 5.2395905494689945,
      "epoch": 1.5730067243035544,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004757543248402839,
      "loss": 4.7751,
      "mean_token_accuracy": 0.2303053393959999,
      "num_tokens": 37530476.0,
      "step": 16375
    },
    {
      "entropy": 5.270265722274781,
      "epoch": 1.5734870317002883,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047573884147416597,
      "loss": 4.8774,
      "mean_token_accuracy": 0.23520581275224686,
      "num_tokens": 37541702.0,
      "step": 16380
    },
    {
      "entropy": 5.297644853591919,
      "epoch": 1.5739673390970221,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004757233534474883,
      "loss": 4.8504,
      "mean_token_accuracy": 0.22783696204423903,
      "num_tokens": 37553323.0,
      "step": 16385
    },
    {
      "entropy": 5.306757211685181,
      "epoch": 1.574447646493756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004757078607606103,
      "loss": 4.9282,
      "mean_token_accuracy": 0.2237038493156433,
      "num_tokens": 37564882.0,
      "step": 16390
    },
    {
      "entropy": 5.295416259765625,
      "epoch": 1.57492795389049,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004756923634138918,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23458154499530792,
      "num_tokens": 37575612.0,
      "step": 16395
    },
    {
      "entropy": 5.269225835800171,
      "epoch": 1.5754082612872238,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047567686140769264,
      "loss": 4.8626,
      "mean_token_accuracy": 0.2300908923149109,
      "num_tokens": 37587089.0,
      "step": 16400
    },
    {
      "entropy": 5.307718944549561,
      "epoch": 1.5758885686839577,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047566135474237247,
      "loss": 4.817,
      "mean_token_accuracy": 0.231142820417881,
      "num_tokens": 37598429.0,
      "step": 16405
    },
    {
      "entropy": 5.345110511779785,
      "epoch": 1.5763688760806915,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047564584341829166,
      "loss": 4.9212,
      "mean_token_accuracy": 0.22768387645483018,
      "num_tokens": 37609769.0,
      "step": 16410
    },
    {
      "entropy": 5.323531675338745,
      "epoch": 1.5768491834774254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047563032743581,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22220935821533203,
      "num_tokens": 37622425.0,
      "step": 16415
    },
    {
      "entropy": 5.290588092803955,
      "epoch": 1.5773294908741593,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047561480679528804,
      "loss": 4.8713,
      "mean_token_accuracy": 0.23559848070144654,
      "num_tokens": 37634819.0,
      "step": 16420
    },
    {
      "entropy": 5.279449939727783,
      "epoch": 1.5778097982708934,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004755992814970859,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23601629287004472,
      "num_tokens": 37645531.0,
      "step": 16425
    },
    {
      "entropy": 5.221272754669189,
      "epoch": 1.5782901056676273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475583751541564,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23109989613294601,
      "num_tokens": 37657243.0,
      "step": 16430
    },
    {
      "entropy": 5.317857646942139,
      "epoch": 1.5787704130643612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047556821692908315,
      "loss": 4.8768,
      "mean_token_accuracy": 0.23020142018795015,
      "num_tokens": 37668218.0,
      "step": 16435
    },
    {
      "entropy": 5.288891839981079,
      "epoch": 1.579250720461095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755526776600038,
      "loss": 4.8065,
      "mean_token_accuracy": 0.23194832354784012,
      "num_tokens": 37680173.0,
      "step": 16440
    },
    {
      "entropy": 5.269957780838013,
      "epoch": 1.5797310278578292,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047553713373468684,
      "loss": 4.8514,
      "mean_token_accuracy": 0.22581207007169724,
      "num_tokens": 37691281.0,
      "step": 16445
    },
    {
      "entropy": 5.3202104568481445,
      "epoch": 1.580211335254563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047552158515349306,
      "loss": 4.9073,
      "mean_token_accuracy": 0.22217728644609452,
      "num_tokens": 37702320.0,
      "step": 16450
    },
    {
      "entropy": 5.292507457733154,
      "epoch": 1.580691642651297,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047550603191678356,
      "loss": 4.8286,
      "mean_token_accuracy": 0.23009601086378098,
      "num_tokens": 37713724.0,
      "step": 16455
    },
    {
      "entropy": 5.275936222076416,
      "epoch": 1.5811719500480308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754904740249194,
      "loss": 4.8112,
      "mean_token_accuracy": 0.23357379287481309,
      "num_tokens": 37724532.0,
      "step": 16460
    },
    {
      "entropy": 5.304412317276001,
      "epoch": 1.5816522574447647,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047547491147826156,
      "loss": 4.774,
      "mean_token_accuracy": 0.24305010139942168,
      "num_tokens": 37735174.0,
      "step": 16465
    },
    {
      "entropy": 5.2683953762054445,
      "epoch": 1.5821325648414986,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004754593442771718,
      "loss": 4.9274,
      "mean_token_accuracy": 0.2259184315800667,
      "num_tokens": 37746981.0,
      "step": 16470
    },
    {
      "entropy": 5.323074722290039,
      "epoch": 1.5826128722382324,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047544377242201115,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23403371274471282,
      "num_tokens": 37757576.0,
      "step": 16475
    },
    {
      "entropy": 5.344353723526001,
      "epoch": 1.5830931796349663,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047542819591314136,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2286149114370346,
      "num_tokens": 37769053.0,
      "step": 16480
    },
    {
      "entropy": 5.306644582748413,
      "epoch": 1.5835734870317002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004754126147509241,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22722458988428115,
      "num_tokens": 37780517.0,
      "step": 16485
    },
    {
      "entropy": 5.3736334323883055,
      "epoch": 1.584053794428434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047539702893572086,
      "loss": 4.9328,
      "mean_token_accuracy": 0.22242968678474426,
      "num_tokens": 37791333.0,
      "step": 16490
    },
    {
      "entropy": 5.265167140960694,
      "epoch": 1.584534101825168,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047538143846789376,
      "loss": 4.7941,
      "mean_token_accuracy": 0.24109085500240326,
      "num_tokens": 37802635.0,
      "step": 16495
    },
    {
      "entropy": 5.225461769104004,
      "epoch": 1.585014409221902,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004753658433478047,
      "loss": 4.7011,
      "mean_token_accuracy": 0.24071227610111237,
      "num_tokens": 37814788.0,
      "step": 16500
    },
    {
      "entropy": 5.269275856018067,
      "epoch": 1.585494716618636,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047535024357581564,
      "loss": 4.8092,
      "mean_token_accuracy": 0.2375115931034088,
      "num_tokens": 37826650.0,
      "step": 16505
    },
    {
      "entropy": 5.323819923400879,
      "epoch": 1.5859750240153698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004753346391522889,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22485073506832123,
      "num_tokens": 37839362.0,
      "step": 16510
    },
    {
      "entropy": 5.329894781112671,
      "epoch": 1.5864553314121037,
      "grad_norm": 1.125,
      "learning_rate": 0.00047531903007758667,
      "loss": 4.8719,
      "mean_token_accuracy": 0.23028807789087297,
      "num_tokens": 37851057.0,
      "step": 16515
    },
    {
      "entropy": 5.245165395736694,
      "epoch": 1.5869356388088378,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004753034163520714,
      "loss": 4.7834,
      "mean_token_accuracy": 0.24282555431127548,
      "num_tokens": 37863507.0,
      "step": 16520
    },
    {
      "entropy": 5.318576812744141,
      "epoch": 1.5874159462055717,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047528779797610557,
      "loss": 4.8542,
      "mean_token_accuracy": 0.23055919110774994,
      "num_tokens": 37874439.0,
      "step": 16525
    },
    {
      "entropy": 5.318417978286743,
      "epoch": 1.5878962536023056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047527217495005184,
      "loss": 4.8138,
      "mean_token_accuracy": 0.2345752611756325,
      "num_tokens": 37886720.0,
      "step": 16530
    },
    {
      "entropy": 5.286732912063599,
      "epoch": 1.5883765609990395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047525654727427285,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2297684445977211,
      "num_tokens": 37897919.0,
      "step": 16535
    },
    {
      "entropy": 5.307161331176758,
      "epoch": 1.5888568683957733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004752409149491315,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2312178924679756,
      "num_tokens": 37911061.0,
      "step": 16540
    },
    {
      "entropy": 5.328029537200928,
      "epoch": 1.5893371757925072,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047522527797499075,
      "loss": 4.8872,
      "mean_token_accuracy": 0.22767945230007172,
      "num_tokens": 37922677.0,
      "step": 16545
    },
    {
      "entropy": 5.209262084960938,
      "epoch": 1.589817483189241,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004752096363522135,
      "loss": 4.7109,
      "mean_token_accuracy": 0.23956117182970046,
      "num_tokens": 37932802.0,
      "step": 16550
    },
    {
      "entropy": 5.281701898574829,
      "epoch": 1.590297790585975,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047519399008116305,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23463936299085617,
      "num_tokens": 37944782.0,
      "step": 16555
    },
    {
      "entropy": 5.2755101203918455,
      "epoch": 1.5907780979827089,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004751783391622026,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23424463272094725,
      "num_tokens": 37956577.0,
      "step": 16560
    },
    {
      "entropy": 5.324416303634644,
      "epoch": 1.5912584053794427,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004751626835956955,
      "loss": 4.8915,
      "mean_token_accuracy": 0.22982990592718125,
      "num_tokens": 37967937.0,
      "step": 16565
    },
    {
      "entropy": 5.236428165435791,
      "epoch": 1.5917387127761766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004751470233820053,
      "loss": 4.7881,
      "mean_token_accuracy": 0.24736988991498948,
      "num_tokens": 37978113.0,
      "step": 16570
    },
    {
      "entropy": 5.214314985275268,
      "epoch": 1.5922190201729105,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004751313585214955,
      "loss": 4.7636,
      "mean_token_accuracy": 0.23870517760515214,
      "num_tokens": 37988852.0,
      "step": 16575
    },
    {
      "entropy": 5.329324197769165,
      "epoch": 1.5926993275696446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004751156890145298,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2265935570001602,
      "num_tokens": 38000531.0,
      "step": 16580
    },
    {
      "entropy": 5.267550659179688,
      "epoch": 1.5931796349663785,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000475100014861472,
      "loss": 4.8385,
      "mean_token_accuracy": 0.23517632633447647,
      "num_tokens": 38011818.0,
      "step": 16585
    },
    {
      "entropy": 5.295084857940674,
      "epoch": 1.5936599423631124,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004750843360626861,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23143020868301392,
      "num_tokens": 38022455.0,
      "step": 16590
    },
    {
      "entropy": 5.267300176620483,
      "epoch": 1.5941402497598463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004750686526185359,
      "loss": 4.8224,
      "mean_token_accuracy": 0.2389863446354866,
      "num_tokens": 38033126.0,
      "step": 16595
    },
    {
      "entropy": 5.137936162948608,
      "epoch": 1.5946205571565804,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047505296452938584,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23936698883771895,
      "num_tokens": 38044935.0,
      "step": 16600
    },
    {
      "entropy": 5.301152944564819,
      "epoch": 1.5951008645533142,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047503727179559995,
      "loss": 4.863,
      "mean_token_accuracy": 0.23265578895807265,
      "num_tokens": 38056080.0,
      "step": 16605
    },
    {
      "entropy": 5.298003005981445,
      "epoch": 1.5955811719500481,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047502157441754256,
      "loss": 4.8692,
      "mean_token_accuracy": 0.23027439713478087,
      "num_tokens": 38066788.0,
      "step": 16610
    },
    {
      "entropy": 5.339261388778686,
      "epoch": 1.596061479346782,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004750058723955781,
      "loss": 4.8883,
      "mean_token_accuracy": 0.23016255497932434,
      "num_tokens": 38079529.0,
      "step": 16615
    },
    {
      "entropy": 5.281230640411377,
      "epoch": 1.5965417867435159,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004749901657300713,
      "loss": 4.8493,
      "mean_token_accuracy": 0.2336796298623085,
      "num_tokens": 38091484.0,
      "step": 16620
    },
    {
      "entropy": 5.31980299949646,
      "epoch": 1.5970220941402498,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047497445442138667,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23428179770708085,
      "num_tokens": 38103102.0,
      "step": 16625
    },
    {
      "entropy": 5.271168422698975,
      "epoch": 1.5975024015369836,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047495873846988896,
      "loss": 4.9074,
      "mean_token_accuracy": 0.22833366990089415,
      "num_tokens": 38115565.0,
      "step": 16630
    },
    {
      "entropy": 5.284711170196533,
      "epoch": 1.5979827089337175,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004749430178759431,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23388296365737915,
      "num_tokens": 38126741.0,
      "step": 16635
    },
    {
      "entropy": 5.283800888061523,
      "epoch": 1.5984630163304514,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047492729263991413,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23621572852134703,
      "num_tokens": 38137149.0,
      "step": 16640
    },
    {
      "entropy": 5.259853267669678,
      "epoch": 1.5989433237271853,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047491156276216695,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2413023144006729,
      "num_tokens": 38148537.0,
      "step": 16645
    },
    {
      "entropy": 5.242429447174072,
      "epoch": 1.5994236311239192,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047489582824306704,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23353455662727357,
      "num_tokens": 38159097.0,
      "step": 16650
    },
    {
      "entropy": 5.235528945922852,
      "epoch": 1.5999039385206533,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047488008908297955,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23837292045354844,
      "num_tokens": 38171400.0,
      "step": 16655
    },
    {
      "entropy": 5.346405410766602,
      "epoch": 1.6003842459173871,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004748643452822699,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23825888633728026,
      "num_tokens": 38181876.0,
      "step": 16660
    },
    {
      "entropy": 5.323920392990113,
      "epoch": 1.600864553314121,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004748485968413036,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2359301343560219,
      "num_tokens": 38193680.0,
      "step": 16665
    },
    {
      "entropy": 5.219975328445434,
      "epoch": 1.601344860710855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047483284376044634,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23666229844093323,
      "num_tokens": 38204944.0,
      "step": 16670
    },
    {
      "entropy": 5.247485113143921,
      "epoch": 1.601825168107589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004748170860400638,
      "loss": 4.7539,
      "mean_token_accuracy": 0.2402797594666481,
      "num_tokens": 38216381.0,
      "step": 16675
    },
    {
      "entropy": 5.3091833114624025,
      "epoch": 1.602305475504323,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047480132368052185,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24012972861528398,
      "num_tokens": 38227420.0,
      "step": 16680
    },
    {
      "entropy": 5.291994190216064,
      "epoch": 1.6027857829010568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047478555668218643,
      "loss": 4.8917,
      "mean_token_accuracy": 0.2328664168715477,
      "num_tokens": 38237869.0,
      "step": 16685
    },
    {
      "entropy": 5.2674675464630125,
      "epoch": 1.6032660902977907,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004747697850454237,
      "loss": 4.8651,
      "mean_token_accuracy": 0.22983739227056504,
      "num_tokens": 38250362.0,
      "step": 16690
    },
    {
      "entropy": 5.398574352264404,
      "epoch": 1.6037463976945245,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004747540087705997,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22736509889364243,
      "num_tokens": 38262887.0,
      "step": 16695
    },
    {
      "entropy": 5.2974348068237305,
      "epoch": 1.6042267050912584,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004747382278580808,
      "loss": 4.7901,
      "mean_token_accuracy": 0.24033626168966293,
      "num_tokens": 38273206.0,
      "step": 16700
    },
    {
      "entropy": 5.278647089004517,
      "epoch": 1.6047070124879923,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004747224423082333,
      "loss": 4.8135,
      "mean_token_accuracy": 0.23600702434778215,
      "num_tokens": 38283307.0,
      "step": 16705
    },
    {
      "entropy": 5.3155981540679935,
      "epoch": 1.6051873198847262,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047470665212142384,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2298277959227562,
      "num_tokens": 38293830.0,
      "step": 16710
    },
    {
      "entropy": 5.332856369018555,
      "epoch": 1.60566762728146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047469085729801887,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23344970643520355,
      "num_tokens": 38306523.0,
      "step": 16715
    },
    {
      "entropy": 5.294479131698608,
      "epoch": 1.606147934678194,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047467505783838515,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2326999858021736,
      "num_tokens": 38318468.0,
      "step": 16720
    },
    {
      "entropy": 5.282611846923828,
      "epoch": 1.6066282420749278,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004746592537428895,
      "loss": 4.7838,
      "mean_token_accuracy": 0.2392311856150627,
      "num_tokens": 38329852.0,
      "step": 16725
    },
    {
      "entropy": 5.198703908920288,
      "epoch": 1.6071085494716617,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047464344501189877,
      "loss": 4.7418,
      "mean_token_accuracy": 0.2462858274579048,
      "num_tokens": 38340951.0,
      "step": 16730
    },
    {
      "entropy": 5.328103494644165,
      "epoch": 1.6075888568683958,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047462763164578015,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2261267215013504,
      "num_tokens": 38351490.0,
      "step": 16735
    },
    {
      "entropy": 5.308258676528931,
      "epoch": 1.6080691642651297,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004746118136449007,
      "loss": 4.8051,
      "mean_token_accuracy": 0.23721198737621307,
      "num_tokens": 38362364.0,
      "step": 16740
    },
    {
      "entropy": 5.304068756103516,
      "epoch": 1.6085494716618636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745959910096276,
      "loss": 4.894,
      "mean_token_accuracy": 0.23752813786268234,
      "num_tokens": 38374132.0,
      "step": 16745
    },
    {
      "entropy": 5.294758462905884,
      "epoch": 1.6090297790585975,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047458016374032837,
      "loss": 4.8665,
      "mean_token_accuracy": 0.230617955327034,
      "num_tokens": 38384608.0,
      "step": 16750
    },
    {
      "entropy": 5.267777585983277,
      "epoch": 1.6095100864553316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004745643318373703,
      "loss": 4.8723,
      "mean_token_accuracy": 0.233360093832016,
      "num_tokens": 38396639.0,
      "step": 16755
    },
    {
      "entropy": 5.297470331192017,
      "epoch": 1.6099903938520654,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047454849530112106,
      "loss": 4.889,
      "mean_token_accuracy": 0.22033181935548782,
      "num_tokens": 38407954.0,
      "step": 16760
    },
    {
      "entropy": 5.366159343719483,
      "epoch": 1.6104707012487993,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047453265413194826,
      "loss": 4.8372,
      "mean_token_accuracy": 0.2318975657224655,
      "num_tokens": 38418939.0,
      "step": 16765
    },
    {
      "entropy": 5.299288702011109,
      "epoch": 1.6109510086455332,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047451680833021973,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23406932651996612,
      "num_tokens": 38429717.0,
      "step": 16770
    },
    {
      "entropy": 5.229883098602295,
      "epoch": 1.611431316042267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004745009578963034,
      "loss": 4.782,
      "mean_token_accuracy": 0.23924860060214997,
      "num_tokens": 38441054.0,
      "step": 16775
    },
    {
      "entropy": 5.20950493812561,
      "epoch": 1.611911623439001,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047448510283056716,
      "loss": 4.7981,
      "mean_token_accuracy": 0.23758042752742767,
      "num_tokens": 38451329.0,
      "step": 16780
    },
    {
      "entropy": 5.270791625976562,
      "epoch": 1.6123919308357348,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047446924313337925,
      "loss": 4.8682,
      "mean_token_accuracy": 0.2329186365008354,
      "num_tokens": 38462118.0,
      "step": 16785
    },
    {
      "entropy": 5.29186749458313,
      "epoch": 1.6128722382324687,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047445337880510773,
      "loss": 4.826,
      "mean_token_accuracy": 0.23813817650079727,
      "num_tokens": 38472642.0,
      "step": 16790
    },
    {
      "entropy": 5.199520635604858,
      "epoch": 1.6133525456292026,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004744375098461211,
      "loss": 4.7632,
      "mean_token_accuracy": 0.24078428000211716,
      "num_tokens": 38483866.0,
      "step": 16795
    },
    {
      "entropy": 5.225883054733276,
      "epoch": 1.6138328530259365,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004744216362567876,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2439822182059288,
      "num_tokens": 38494164.0,
      "step": 16800
    },
    {
      "entropy": 5.349162864685058,
      "epoch": 1.6143131604226704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047440575803747595,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22300061881542205,
      "num_tokens": 38505340.0,
      "step": 16805
    },
    {
      "entropy": 5.419594430923462,
      "epoch": 1.6147934678194045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047438987518855463,
      "loss": 4.9589,
      "mean_token_accuracy": 0.22658967822790146,
      "num_tokens": 38516164.0,
      "step": 16810
    },
    {
      "entropy": 5.351446628570557,
      "epoch": 1.6152737752161384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004743739877103926,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2295348584651947,
      "num_tokens": 38526545.0,
      "step": 16815
    },
    {
      "entropy": 5.263201570510864,
      "epoch": 1.6157540826128722,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004743580956033585,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23188736736774446,
      "num_tokens": 38538102.0,
      "step": 16820
    },
    {
      "entropy": 5.288665103912353,
      "epoch": 1.6162343900096061,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047434219886782135,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22566987872123717,
      "num_tokens": 38550594.0,
      "step": 16825
    },
    {
      "entropy": 5.281469297409058,
      "epoch": 1.6167146974063402,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004743262975041504,
      "loss": 4.7963,
      "mean_token_accuracy": 0.2355595663189888,
      "num_tokens": 38561666.0,
      "step": 16830
    },
    {
      "entropy": 5.240443134307862,
      "epoch": 1.617195004803074,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004743103915127146,
      "loss": 4.7631,
      "mean_token_accuracy": 0.2379718616604805,
      "num_tokens": 38572923.0,
      "step": 16835
    },
    {
      "entropy": 5.26913890838623,
      "epoch": 1.617675312199808,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047429448089388336,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2370508760213852,
      "num_tokens": 38584108.0,
      "step": 16840
    },
    {
      "entropy": 5.389714002609253,
      "epoch": 1.6181556195965419,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047427856564802605,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22537550330162048,
      "num_tokens": 38594974.0,
      "step": 16845
    },
    {
      "entropy": 5.321589422225952,
      "epoch": 1.6186359269932757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004742626457755122,
      "loss": 4.8755,
      "mean_token_accuracy": 0.23081362396478652,
      "num_tokens": 38606436.0,
      "step": 16850
    },
    {
      "entropy": 5.262910509109497,
      "epoch": 1.6191162343900096,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004742467212767114,
      "loss": 4.7813,
      "mean_token_accuracy": 0.23807011395692826,
      "num_tokens": 38618760.0,
      "step": 16855
    },
    {
      "entropy": 5.244272756576538,
      "epoch": 1.6195965417867435,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004742307921519933,
      "loss": 4.7506,
      "mean_token_accuracy": 0.24328799545764923,
      "num_tokens": 38629901.0,
      "step": 16860
    },
    {
      "entropy": 5.367648267745972,
      "epoch": 1.6200768491834774,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047421485840172794,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22186616510152818,
      "num_tokens": 38640798.0,
      "step": 16865
    },
    {
      "entropy": 5.332133865356445,
      "epoch": 1.6205571565802113,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004741989200262851,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23155442774295806,
      "num_tokens": 38653048.0,
      "step": 16870
    },
    {
      "entropy": 5.303104496002197,
      "epoch": 1.6210374639769451,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004741829770260347,
      "loss": 4.9125,
      "mean_token_accuracy": 0.2296425998210907,
      "num_tokens": 38666219.0,
      "step": 16875
    },
    {
      "entropy": 5.2867751121521,
      "epoch": 1.621517771373679,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047416702940134714,
      "loss": 4.7781,
      "mean_token_accuracy": 0.24160979688167572,
      "num_tokens": 38678584.0,
      "step": 16880
    },
    {
      "entropy": 5.283935642242431,
      "epoch": 1.621998078770413,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047415107715259255,
      "loss": 4.8154,
      "mean_token_accuracy": 0.22938486337661743,
      "num_tokens": 38690164.0,
      "step": 16885
    },
    {
      "entropy": 5.312966537475586,
      "epoch": 1.622478386167147,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047413512028014125,
      "loss": 4.8907,
      "mean_token_accuracy": 0.23247092515230178,
      "num_tokens": 38702035.0,
      "step": 16890
    },
    {
      "entropy": 5.339834403991699,
      "epoch": 1.622958693563881,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004741191587843638,
      "loss": 4.9182,
      "mean_token_accuracy": 0.23113404214382172,
      "num_tokens": 38713802.0,
      "step": 16895
    },
    {
      "entropy": 5.301450634002686,
      "epoch": 1.6234390009606148,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004741031926656308,
      "loss": 4.7789,
      "mean_token_accuracy": 0.2441946253180504,
      "num_tokens": 38725682.0,
      "step": 16900
    },
    {
      "entropy": 5.246243095397949,
      "epoch": 1.6239193083573487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004740872219243128,
      "loss": 4.799,
      "mean_token_accuracy": 0.23669822812080382,
      "num_tokens": 38737897.0,
      "step": 16905
    },
    {
      "entropy": 5.35952410697937,
      "epoch": 1.6243996157540828,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004740712465607807,
      "loss": 4.8997,
      "mean_token_accuracy": 0.23234787732362747,
      "num_tokens": 38749446.0,
      "step": 16910
    },
    {
      "entropy": 5.3195099353790285,
      "epoch": 1.6248799231508166,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004740552665754054,
      "loss": 4.8585,
      "mean_token_accuracy": 0.23006198704242706,
      "num_tokens": 38761406.0,
      "step": 16915
    },
    {
      "entropy": 5.314310550689697,
      "epoch": 1.6253602305475505,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047403928196855776,
      "loss": 4.8572,
      "mean_token_accuracy": 0.24156680852174758,
      "num_tokens": 38772574.0,
      "step": 16920
    },
    {
      "entropy": 5.292901849746704,
      "epoch": 1.6258405379442844,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047402329274060916,
      "loss": 4.8131,
      "mean_token_accuracy": 0.24698593020439147,
      "num_tokens": 38782932.0,
      "step": 16925
    },
    {
      "entropy": 5.305176496505737,
      "epoch": 1.6263208453410183,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004740072988919306,
      "loss": 4.894,
      "mean_token_accuracy": 0.23321023136377333,
      "num_tokens": 38793799.0,
      "step": 16930
    },
    {
      "entropy": 5.2313446521759035,
      "epoch": 1.6268011527377522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004739913004228936,
      "loss": 4.7205,
      "mean_token_accuracy": 0.24256573021411895,
      "num_tokens": 38804899.0,
      "step": 16935
    },
    {
      "entropy": 5.31002950668335,
      "epoch": 1.627281460134486,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004739752973338694,
      "loss": 4.924,
      "mean_token_accuracy": 0.2220922127366066,
      "num_tokens": 38816455.0,
      "step": 16940
    },
    {
      "entropy": 5.293491458892822,
      "epoch": 1.62776176753122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047395928962522965,
      "loss": 4.8208,
      "mean_token_accuracy": 0.23382614701986312,
      "num_tokens": 38827735.0,
      "step": 16945
    },
    {
      "entropy": 5.315946578979492,
      "epoch": 1.6282420749279538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047394327729734595,
      "loss": 4.8761,
      "mean_token_accuracy": 0.2292775273323059,
      "num_tokens": 38838536.0,
      "step": 16950
    },
    {
      "entropy": 5.366758728027344,
      "epoch": 1.6287223823246877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004739272603505901,
      "loss": 4.8902,
      "mean_token_accuracy": 0.23128196150064467,
      "num_tokens": 38849577.0,
      "step": 16955
    },
    {
      "entropy": 5.365436458587647,
      "epoch": 1.6292026897214216,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000473911238785334,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22778777480125428,
      "num_tokens": 38861370.0,
      "step": 16960
    },
    {
      "entropy": 5.2635413646698,
      "epoch": 1.6296829971181557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004738952126019496,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22942119985818862,
      "num_tokens": 38872885.0,
      "step": 16965
    },
    {
      "entropy": 5.2784277439117435,
      "epoch": 1.6301633045148896,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004738791818008089,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23341765701770784,
      "num_tokens": 38885440.0,
      "step": 16970
    },
    {
      "entropy": 5.281130981445313,
      "epoch": 1.6306436119116234,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004738631463822841,
      "loss": 4.7928,
      "mean_token_accuracy": 0.24120174646377562,
      "num_tokens": 38897108.0,
      "step": 16975
    },
    {
      "entropy": 5.281213331222534,
      "epoch": 1.6311239193083573,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047384710634674766,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22871453911066056,
      "num_tokens": 38908020.0,
      "step": 16980
    },
    {
      "entropy": 5.281463575363159,
      "epoch": 1.6316042267050914,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047383106169457184,
      "loss": 4.868,
      "mean_token_accuracy": 0.23207082003355026,
      "num_tokens": 38919558.0,
      "step": 16985
    },
    {
      "entropy": 5.273614072799683,
      "epoch": 1.6320845341018253,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004738150124261292,
      "loss": 4.8473,
      "mean_token_accuracy": 0.23612669110298157,
      "num_tokens": 38931008.0,
      "step": 16990
    },
    {
      "entropy": 5.2326836585998535,
      "epoch": 1.6325648414985592,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047379895854179226,
      "loss": 4.764,
      "mean_token_accuracy": 0.24143097698688507,
      "num_tokens": 38943474.0,
      "step": 16995
    },
    {
      "entropy": 5.307919597625732,
      "epoch": 1.633045148895293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737829000419338,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22784842401742936,
      "num_tokens": 38954592.0,
      "step": 17000
    },
    {
      "entropy": 5.2810444831848145,
      "epoch": 1.633525456292027,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047376683692692666,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23428010642528535,
      "num_tokens": 38965627.0,
      "step": 17005
    },
    {
      "entropy": 5.224449968338012,
      "epoch": 1.6340057636887608,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004737507691971439,
      "loss": 4.7807,
      "mean_token_accuracy": 0.2387988969683647,
      "num_tokens": 38975731.0,
      "step": 17010
    },
    {
      "entropy": 5.259978246688843,
      "epoch": 1.6344860710854947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047373469685295833,
      "loss": 4.7875,
      "mean_token_accuracy": 0.2431796208024025,
      "num_tokens": 38987101.0,
      "step": 17015
    },
    {
      "entropy": 5.308274507522583,
      "epoch": 1.6349663784822286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047371861989474326,
      "loss": 4.8468,
      "mean_token_accuracy": 0.236659537255764,
      "num_tokens": 38999584.0,
      "step": 17020
    },
    {
      "entropy": 5.245364713668823,
      "epoch": 1.6354466858789625,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004737025383228719,
      "loss": 4.832,
      "mean_token_accuracy": 0.2365382730960846,
      "num_tokens": 39012025.0,
      "step": 17025
    },
    {
      "entropy": 5.181337022781372,
      "epoch": 1.6359269932756964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047368645213771764,
      "loss": 4.7228,
      "mean_token_accuracy": 0.24250809401273726,
      "num_tokens": 39024102.0,
      "step": 17030
    },
    {
      "entropy": 5.307433366775513,
      "epoch": 1.6364073006724302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000473670361339654,
      "loss": 4.849,
      "mean_token_accuracy": 0.22524280697107316,
      "num_tokens": 39035295.0,
      "step": 17035
    },
    {
      "entropy": 5.260975646972656,
      "epoch": 1.6368876080691641,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004736542659290544,
      "loss": 4.7647,
      "mean_token_accuracy": 0.2351788029074669,
      "num_tokens": 39046493.0,
      "step": 17040
    },
    {
      "entropy": 5.2912458896636965,
      "epoch": 1.6373679154658982,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004736381659062927,
      "loss": 4.7917,
      "mean_token_accuracy": 0.24046519547700881,
      "num_tokens": 39056746.0,
      "step": 17045
    },
    {
      "entropy": 5.281490516662598,
      "epoch": 1.637848222862632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047362206127174255,
      "loss": 4.8482,
      "mean_token_accuracy": 0.2265760436654091,
      "num_tokens": 39068263.0,
      "step": 17050
    },
    {
      "entropy": 5.3214210033416744,
      "epoch": 1.638328530259366,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047360595202577786,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23763253539800644,
      "num_tokens": 39080666.0,
      "step": 17055
    },
    {
      "entropy": 5.272362852096558,
      "epoch": 1.6388088376560999,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047358983816877284,
      "loss": 4.8302,
      "mean_token_accuracy": 0.2339204341173172,
      "num_tokens": 39091851.0,
      "step": 17060
    },
    {
      "entropy": 5.332276821136475,
      "epoch": 1.639289145052834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004735737197011015,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23912704437971116,
      "num_tokens": 39102893.0,
      "step": 17065
    },
    {
      "entropy": 5.231080198287964,
      "epoch": 1.6397694524495678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047355759662313793,
      "loss": 4.7458,
      "mean_token_accuracy": 0.23896219581365585,
      "num_tokens": 39113582.0,
      "step": 17070
    },
    {
      "entropy": 5.345025444030762,
      "epoch": 1.6402497598463017,
      "grad_norm": 1.0,
      "learning_rate": 0.0004735414689352566,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2277378648519516,
      "num_tokens": 39124616.0,
      "step": 17075
    },
    {
      "entropy": 5.2194746971130375,
      "epoch": 1.6407300672430356,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735253366378318,
      "loss": 4.7558,
      "mean_token_accuracy": 0.24169106632471085,
      "num_tokens": 39135537.0,
      "step": 17080
    },
    {
      "entropy": 5.303269910812378,
      "epoch": 1.6412103746397695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004735091997312383,
      "loss": 4.8315,
      "mean_token_accuracy": 0.23760396391153335,
      "num_tokens": 39146363.0,
      "step": 17085
    },
    {
      "entropy": 5.299032497406006,
      "epoch": 1.6416906820365034,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047349305821585067,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2418719381093979,
      "num_tokens": 39157319.0,
      "step": 17090
    },
    {
      "entropy": 5.267262315750122,
      "epoch": 1.6421709894332372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004734769120920435,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23554610311985016,
      "num_tokens": 39169352.0,
      "step": 17095
    },
    {
      "entropy": 5.297905206680298,
      "epoch": 1.6426512968299711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004734607613601919,
      "loss": 4.8076,
      "mean_token_accuracy": 0.23866427838802337,
      "num_tokens": 39180989.0,
      "step": 17100
    },
    {
      "entropy": 5.314274549484253,
      "epoch": 1.643131604226705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047344460602067077,
      "loss": 4.8509,
      "mean_token_accuracy": 0.2350212514400482,
      "num_tokens": 39192309.0,
      "step": 17105
    },
    {
      "entropy": 5.241296672821045,
      "epoch": 1.643611911623439,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000473428446073855,
      "loss": 4.798,
      "mean_token_accuracy": 0.2306523159146309,
      "num_tokens": 39203586.0,
      "step": 17110
    },
    {
      "entropy": 5.255983543395996,
      "epoch": 1.6440922190201728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047341228152012003,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23128417134284973,
      "num_tokens": 39216248.0,
      "step": 17115
    },
    {
      "entropy": 5.32506103515625,
      "epoch": 1.6445725264169067,
      "grad_norm": 0.984375,
      "learning_rate": 0.000473396112359841,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2395351603627205,
      "num_tokens": 39227244.0,
      "step": 17120
    },
    {
      "entropy": 5.263369607925415,
      "epoch": 1.6450528338136408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047337993859339334,
      "loss": 4.7477,
      "mean_token_accuracy": 0.24486439526081086,
      "num_tokens": 39238769.0,
      "step": 17125
    },
    {
      "entropy": 5.262567186355591,
      "epoch": 1.6455331412103746,
      "grad_norm": 1.0,
      "learning_rate": 0.00047336376022115255,
      "loss": 4.8454,
      "mean_token_accuracy": 0.23501148968935012,
      "num_tokens": 39251424.0,
      "step": 17130
    },
    {
      "entropy": 5.225255823135376,
      "epoch": 1.6460134486071085,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047334757724349437,
      "loss": 4.8125,
      "mean_token_accuracy": 0.23960975259542466,
      "num_tokens": 39262451.0,
      "step": 17135
    },
    {
      "entropy": 5.258809900283813,
      "epoch": 1.6464937560038426,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004733313896607943,
      "loss": 4.811,
      "mean_token_accuracy": 0.2328306645154953,
      "num_tokens": 39275129.0,
      "step": 17140
    },
    {
      "entropy": 5.384744262695312,
      "epoch": 1.6469740634005765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004733151974734284,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22445586174726487,
      "num_tokens": 39287399.0,
      "step": 17145
    },
    {
      "entropy": 5.303330039978027,
      "epoch": 1.6474543707973104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047329900068177245,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23840391188859938,
      "num_tokens": 39297755.0,
      "step": 17150
    },
    {
      "entropy": 5.325929307937622,
      "epoch": 1.6479346781940443,
      "grad_norm": 1.0,
      "learning_rate": 0.00047328279928620244,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22436654716730117,
      "num_tokens": 39311012.0,
      "step": 17155
    },
    {
      "entropy": 5.316707992553711,
      "epoch": 1.6484149855907781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004732665932870947,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23353114277124404,
      "num_tokens": 39321797.0,
      "step": 17160
    },
    {
      "entropy": 5.297512626647949,
      "epoch": 1.648895292987512,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047325038268482544,
      "loss": 4.7811,
      "mean_token_accuracy": 0.24269960522651673,
      "num_tokens": 39333598.0,
      "step": 17165
    },
    {
      "entropy": 5.34099326133728,
      "epoch": 1.649375600384246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004732341674797709,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2309274435043335,
      "num_tokens": 39345578.0,
      "step": 17170
    },
    {
      "entropy": 5.196650695800781,
      "epoch": 1.6498559077809798,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047321794767230766,
      "loss": 4.7396,
      "mean_token_accuracy": 0.23895098268985748,
      "num_tokens": 39356744.0,
      "step": 17175
    },
    {
      "entropy": 5.287885904312134,
      "epoch": 1.6503362151777137,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047320172326281224,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23571549654006957,
      "num_tokens": 39368138.0,
      "step": 17180
    },
    {
      "entropy": 5.239331007003784,
      "epoch": 1.6508165225744476,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047318549425166134,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24280077069997788,
      "num_tokens": 39380117.0,
      "step": 17185
    },
    {
      "entropy": 5.2853615283966064,
      "epoch": 1.6512968299711814,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004731692606392318,
      "loss": 4.7892,
      "mean_token_accuracy": 0.23440600484609603,
      "num_tokens": 39390913.0,
      "step": 17190
    },
    {
      "entropy": 5.368222618103028,
      "epoch": 1.6517771373679153,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004731530224259004,
      "loss": 4.9323,
      "mean_token_accuracy": 0.2215529128909111,
      "num_tokens": 39401001.0,
      "step": 17195
    },
    {
      "entropy": 5.27427043914795,
      "epoch": 1.6522574447646494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004731367796120442,
      "loss": 4.8127,
      "mean_token_accuracy": 0.23188365995883942,
      "num_tokens": 39411982.0,
      "step": 17200
    },
    {
      "entropy": 5.333657360076904,
      "epoch": 1.6527377521613833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004731205321980404,
      "loss": 4.8381,
      "mean_token_accuracy": 0.23506590276956557,
      "num_tokens": 39423363.0,
      "step": 17205
    },
    {
      "entropy": 5.325553703308105,
      "epoch": 1.6532180595581172,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047310428018426616,
      "loss": 4.8906,
      "mean_token_accuracy": 0.23053719252347946,
      "num_tokens": 39434002.0,
      "step": 17210
    },
    {
      "entropy": 5.253401613235473,
      "epoch": 1.653698366954851,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004730880235710987,
      "loss": 4.7772,
      "mean_token_accuracy": 0.2358376756310463,
      "num_tokens": 39444629.0,
      "step": 17215
    },
    {
      "entropy": 5.2499696731567385,
      "epoch": 1.6541786743515852,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004730717623589155,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2413213461637497,
      "num_tokens": 39454803.0,
      "step": 17220
    },
    {
      "entropy": 5.2881261825561525,
      "epoch": 1.654658981748319,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004730554965480942,
      "loss": 4.8991,
      "mean_token_accuracy": 0.2327059581875801,
      "num_tokens": 39466931.0,
      "step": 17225
    },
    {
      "entropy": 5.308854389190674,
      "epoch": 1.655139289145053,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004730392261390124,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23710836619138717,
      "num_tokens": 39478127.0,
      "step": 17230
    },
    {
      "entropy": 5.298028230667114,
      "epoch": 1.6556195965417868,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004730229511320478,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22708330154418946,
      "num_tokens": 39488188.0,
      "step": 17235
    },
    {
      "entropy": 5.298912525177002,
      "epoch": 1.6560999039385207,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047300667152757827,
      "loss": 4.8105,
      "mean_token_accuracy": 0.2343249037861824,
      "num_tokens": 39500035.0,
      "step": 17240
    },
    {
      "entropy": 5.357331466674805,
      "epoch": 1.6565802113352546,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047299038732598184,
      "loss": 4.8907,
      "mean_token_accuracy": 0.22777570486068727,
      "num_tokens": 39510922.0,
      "step": 17245
    },
    {
      "entropy": 5.305117988586426,
      "epoch": 1.6570605187319885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047297409852763644,
      "loss": 4.8718,
      "mean_token_accuracy": 0.2376135528087616,
      "num_tokens": 39521319.0,
      "step": 17250
    },
    {
      "entropy": 5.312054252624511,
      "epoch": 1.6575408261287223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004729578051329204,
      "loss": 4.8541,
      "mean_token_accuracy": 0.23387885987758636,
      "num_tokens": 39532408.0,
      "step": 17255
    },
    {
      "entropy": 5.32342300415039,
      "epoch": 1.6580211335254562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047294150714221185,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22929502427577972,
      "num_tokens": 39543323.0,
      "step": 17260
    },
    {
      "entropy": 5.329056024551392,
      "epoch": 1.65850144092219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004729252045558894,
      "loss": 4.8587,
      "mean_token_accuracy": 0.2296065166592598,
      "num_tokens": 39554133.0,
      "step": 17265
    },
    {
      "entropy": 5.23650107383728,
      "epoch": 1.658981748318924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047290889737433133,
      "loss": 4.7351,
      "mean_token_accuracy": 0.23968190252780913,
      "num_tokens": 39565990.0,
      "step": 17270
    },
    {
      "entropy": 5.298448324203491,
      "epoch": 1.6594620557156579,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047289258559791633,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23498528599739074,
      "num_tokens": 39577591.0,
      "step": 17275
    },
    {
      "entropy": 5.29054365158081,
      "epoch": 1.659942363112392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047287626922702317,
      "loss": 4.807,
      "mean_token_accuracy": 0.23463153094053268,
      "num_tokens": 39589922.0,
      "step": 17280
    },
    {
      "entropy": 5.281710815429688,
      "epoch": 1.6604226705091258,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047285994826203054,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23684215247631074,
      "num_tokens": 39602183.0,
      "step": 17285
    },
    {
      "entropy": 5.293906021118164,
      "epoch": 1.6609029779058597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004728436227033175,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23370730727910996,
      "num_tokens": 39612922.0,
      "step": 17290
    },
    {
      "entropy": 5.295914602279663,
      "epoch": 1.6613832853025938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047282729255126294,
      "loss": 4.8285,
      "mean_token_accuracy": 0.23624569028615952,
      "num_tokens": 39624475.0,
      "step": 17295
    },
    {
      "entropy": 5.182626247406006,
      "epoch": 1.6618635926993277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004728109578062461,
      "loss": 4.6931,
      "mean_token_accuracy": 0.2408390000462532,
      "num_tokens": 39635230.0,
      "step": 17300
    },
    {
      "entropy": 5.219822025299072,
      "epoch": 1.6623439000960616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047279461846864626,
      "loss": 4.8039,
      "mean_token_accuracy": 0.23505667299032212,
      "num_tokens": 39646788.0,
      "step": 17305
    },
    {
      "entropy": 5.256872892379761,
      "epoch": 1.6628242074927955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047277827453884265,
      "loss": 4.7777,
      "mean_token_accuracy": 0.23859920352697372,
      "num_tokens": 39657531.0,
      "step": 17310
    },
    {
      "entropy": 5.382240247726441,
      "epoch": 1.6633045148895294,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047276192601721477,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21792283952236174,
      "num_tokens": 39669008.0,
      "step": 17315
    },
    {
      "entropy": 5.311697387695313,
      "epoch": 1.6637848222862632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004727455729041422,
      "loss": 4.78,
      "mean_token_accuracy": 0.23815541863441467,
      "num_tokens": 39680772.0,
      "step": 17320
    },
    {
      "entropy": 5.349793291091919,
      "epoch": 1.6642651296829971,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047272921520000465,
      "loss": 4.9551,
      "mean_token_accuracy": 0.2204518958926201,
      "num_tokens": 39694107.0,
      "step": 17325
    },
    {
      "entropy": 5.256176853179932,
      "epoch": 1.664745437079731,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004727128529051819,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23090701252222062,
      "num_tokens": 39705440.0,
      "step": 17330
    },
    {
      "entropy": 5.276105499267578,
      "epoch": 1.6652257444764649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726964860200537,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23572736084461213,
      "num_tokens": 39716246.0,
      "step": 17335
    },
    {
      "entropy": 5.301106071472168,
      "epoch": 1.6657060518731988,
      "grad_norm": 1.0,
      "learning_rate": 0.0004726801145450002,
      "loss": 4.8287,
      "mean_token_accuracy": 0.2313432976603508,
      "num_tokens": 39727233.0,
      "step": 17340
    },
    {
      "entropy": 5.272641324996949,
      "epoch": 1.6661863592699326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004726637384804014,
      "loss": 4.779,
      "mean_token_accuracy": 0.2358853191137314,
      "num_tokens": 39738929.0,
      "step": 17345
    },
    {
      "entropy": 5.235718297958374,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004726473578266375,
      "loss": 4.8273,
      "mean_token_accuracy": 0.23151071220636368,
      "num_tokens": 39750714.0,
      "step": 17350
    },
    {
      "entropy": 5.272091388702393,
      "epoch": 1.6671469740634006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047263097258408893,
      "loss": 4.8651,
      "mean_token_accuracy": 0.24091965705156326,
      "num_tokens": 39762810.0,
      "step": 17355
    },
    {
      "entropy": 5.339130115509033,
      "epoch": 1.6676272814601345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004726145827531359,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22508398890495301,
      "num_tokens": 39774611.0,
      "step": 17360
    },
    {
      "entropy": 5.263414478302002,
      "epoch": 1.6681075888568684,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047259818833415916,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24530019611120224,
      "num_tokens": 39784900.0,
      "step": 17365
    },
    {
      "entropy": 5.3592079162597654,
      "epoch": 1.6685878962536023,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047258178932753917,
      "loss": 4.8662,
      "mean_token_accuracy": 0.23710455596446992,
      "num_tokens": 39794805.0,
      "step": 17370
    },
    {
      "entropy": 5.308233308792114,
      "epoch": 1.6690682036503364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047256538573365675,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23124666810035704,
      "num_tokens": 39806444.0,
      "step": 17375
    },
    {
      "entropy": 5.42754864692688,
      "epoch": 1.6695485110470702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004725489775528928,
      "loss": 5.0007,
      "mean_token_accuracy": 0.21949970126152038,
      "num_tokens": 39817724.0,
      "step": 17380
    },
    {
      "entropy": 5.3036213397979735,
      "epoch": 1.6700288184438041,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047253256478562805,
      "loss": 4.7998,
      "mean_token_accuracy": 0.23325893580913543,
      "num_tokens": 39828944.0,
      "step": 17385
    },
    {
      "entropy": 5.263126802444458,
      "epoch": 1.670509125840538,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047251614743224374,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2347727671265602,
      "num_tokens": 39839978.0,
      "step": 17390
    },
    {
      "entropy": 5.252354860305786,
      "epoch": 1.670989433237272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047249972549312107,
      "loss": 4.8289,
      "mean_token_accuracy": 0.236934694647789,
      "num_tokens": 39850787.0,
      "step": 17395
    },
    {
      "entropy": 5.342466688156128,
      "epoch": 1.6714697406340058,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004724832989686411,
      "loss": 4.8868,
      "mean_token_accuracy": 0.23654668927192687,
      "num_tokens": 39862245.0,
      "step": 17400
    },
    {
      "entropy": 5.262358856201172,
      "epoch": 1.6719500480307397,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047246686785918545,
      "loss": 4.8117,
      "mean_token_accuracy": 0.240486441552639,
      "num_tokens": 39872295.0,
      "step": 17405
    },
    {
      "entropy": 5.231879234313965,
      "epoch": 1.6724303554274735,
      "grad_norm": 1.0,
      "learning_rate": 0.00047245043216513546,
      "loss": 4.7388,
      "mean_token_accuracy": 0.2320246621966362,
      "num_tokens": 39882839.0,
      "step": 17410
    },
    {
      "entropy": 5.213819885253907,
      "epoch": 1.6729106628242074,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004724339918868727,
      "loss": 4.7528,
      "mean_token_accuracy": 0.23954592496156693,
      "num_tokens": 39894767.0,
      "step": 17415
    },
    {
      "entropy": 5.315658521652222,
      "epoch": 1.6733909702209413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472417547024779,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2247718796133995,
      "num_tokens": 39907492.0,
      "step": 17420
    },
    {
      "entropy": 5.1804815292358395,
      "epoch": 1.6738712776176752,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047240109757923593,
      "loss": 4.6796,
      "mean_token_accuracy": 0.24237322360277175,
      "num_tokens": 39919005.0,
      "step": 17425
    },
    {
      "entropy": 5.2563557624816895,
      "epoch": 1.674351585014409,
      "grad_norm": 1.0,
      "learning_rate": 0.0004723846435506256,
      "loss": 4.7961,
      "mean_token_accuracy": 0.2376502439379692,
      "num_tokens": 39930417.0,
      "step": 17430
    },
    {
      "entropy": 5.250257253646851,
      "epoch": 1.6748318924111432,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047236818493932994,
      "loss": 4.8447,
      "mean_token_accuracy": 0.23422068506479263,
      "num_tokens": 39941873.0,
      "step": 17435
    },
    {
      "entropy": 5.3291408061981205,
      "epoch": 1.675312199807877,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004723517217457311,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22622331529855727,
      "num_tokens": 39953280.0,
      "step": 17440
    },
    {
      "entropy": 5.350927734375,
      "epoch": 1.675792507204611,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004723352539702113,
      "loss": 4.9127,
      "mean_token_accuracy": 0.23177622109651566,
      "num_tokens": 39964898.0,
      "step": 17445
    },
    {
      "entropy": 5.2989030361175535,
      "epoch": 1.6762728146013448,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004723187816131529,
      "loss": 4.7688,
      "mean_token_accuracy": 0.23942091464996337,
      "num_tokens": 39976945.0,
      "step": 17450
    },
    {
      "entropy": 5.344297266006469,
      "epoch": 1.676753121998079,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004723023046749383,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2357146993279457,
      "num_tokens": 39987760.0,
      "step": 17455
    },
    {
      "entropy": 5.206570339202881,
      "epoch": 1.6772334293948128,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047228582315595,
      "loss": 4.6988,
      "mean_token_accuracy": 0.24810300767421722,
      "num_tokens": 39998171.0,
      "step": 17460
    },
    {
      "entropy": 5.264092922210693,
      "epoch": 1.6777137367915467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004722693370565708,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2320651337504387,
      "num_tokens": 40010124.0,
      "step": 17465
    },
    {
      "entropy": 5.254520845413208,
      "epoch": 1.6781940441882806,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047225284637718323,
      "loss": 4.8336,
      "mean_token_accuracy": 0.23181586414575578,
      "num_tokens": 40022599.0,
      "step": 17470
    },
    {
      "entropy": 5.300999402999878,
      "epoch": 1.6786743515850144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722363511181703,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2414163291454315,
      "num_tokens": 40033738.0,
      "step": 17475
    },
    {
      "entropy": 5.31995997428894,
      "epoch": 1.6791546589817483,
      "grad_norm": 1.0625,
      "learning_rate": 0.000472219851279915,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23918654918670654,
      "num_tokens": 40044148.0,
      "step": 17480
    },
    {
      "entropy": 5.3219743251800535,
      "epoch": 1.6796349663784822,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004722033468628004,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23576394319534302,
      "num_tokens": 40055430.0,
      "step": 17485
    },
    {
      "entropy": 5.263074684143066,
      "epoch": 1.680115273775216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004721868378672098,
      "loss": 4.8263,
      "mean_token_accuracy": 0.2363945871591568,
      "num_tokens": 40066147.0,
      "step": 17490
    },
    {
      "entropy": 5.279831314086914,
      "epoch": 1.68059558117195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004721703242935261,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23584438115358353,
      "num_tokens": 40077567.0,
      "step": 17495
    },
    {
      "entropy": 5.369070148468017,
      "epoch": 1.6810758885686838,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004721538061421331,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23782893270254135,
      "num_tokens": 40089544.0,
      "step": 17500
    },
    {
      "entropy": 5.233191061019897,
      "epoch": 1.6815561959654177,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047213728341341407,
      "loss": 4.7568,
      "mean_token_accuracy": 0.24242319613695146,
      "num_tokens": 40100557.0,
      "step": 17505
    },
    {
      "entropy": 5.266232109069824,
      "epoch": 1.6820365033621518,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004721207561077527,
      "loss": 4.8331,
      "mean_token_accuracy": 0.2298084944486618,
      "num_tokens": 40112052.0,
      "step": 17510
    },
    {
      "entropy": 5.472871971130371,
      "epoch": 1.6825168107588857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004721042242255327,
      "loss": 4.9971,
      "mean_token_accuracy": 0.22324882596731185,
      "num_tokens": 40123830.0,
      "step": 17515
    },
    {
      "entropy": 5.343978834152222,
      "epoch": 1.6829971181556196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047208768776713805,
      "loss": 4.8857,
      "mean_token_accuracy": 0.2372915655374527,
      "num_tokens": 40134222.0,
      "step": 17520
    },
    {
      "entropy": 5.258941888809204,
      "epoch": 1.6834774255523535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004720711467329523,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23524001538753508,
      "num_tokens": 40145552.0,
      "step": 17525
    },
    {
      "entropy": 5.272229146957398,
      "epoch": 1.6839577329490876,
      "grad_norm": 1.0,
      "learning_rate": 0.0004720546011233599,
      "loss": 4.7112,
      "mean_token_accuracy": 0.2383154660463333,
      "num_tokens": 40156615.0,
      "step": 17530
    },
    {
      "entropy": 5.272588109970092,
      "epoch": 1.6844380403458215,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004720380509387446,
      "loss": 4.8296,
      "mean_token_accuracy": 0.2392334073781967,
      "num_tokens": 40168825.0,
      "step": 17535
    },
    {
      "entropy": 5.286524963378906,
      "epoch": 1.6849183477425553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000472021496179491,
      "loss": 4.8229,
      "mean_token_accuracy": 0.2370854824781418,
      "num_tokens": 40179523.0,
      "step": 17540
    },
    {
      "entropy": 5.265411233901977,
      "epoch": 1.6853986551392892,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047200493684598316,
      "loss": 4.7856,
      "mean_token_accuracy": 0.2351241797208786,
      "num_tokens": 40191362.0,
      "step": 17545
    },
    {
      "entropy": 5.306664752960205,
      "epoch": 1.685878962536023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047198837293860573,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23167840540409088,
      "num_tokens": 40202274.0,
      "step": 17550
    },
    {
      "entropy": 5.325405216217041,
      "epoch": 1.686359269932757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004719718044577432,
      "loss": 4.8826,
      "mean_token_accuracy": 0.22492847740650176,
      "num_tokens": 40213907.0,
      "step": 17555
    },
    {
      "entropy": 5.245599222183228,
      "epoch": 1.6868395773294909,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047195523140378034,
      "loss": 4.8216,
      "mean_token_accuracy": 0.23211388736963273,
      "num_tokens": 40225300.0,
      "step": 17560
    },
    {
      "entropy": 5.241288661956787,
      "epoch": 1.6873198847262247,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047193865377710177,
      "loss": 4.7259,
      "mean_token_accuracy": 0.24264781773090363,
      "num_tokens": 40236197.0,
      "step": 17565
    },
    {
      "entropy": 5.377566289901734,
      "epoch": 1.6878001921229586,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047192207157809246,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22865129858255387,
      "num_tokens": 40247887.0,
      "step": 17570
    },
    {
      "entropy": 5.372284078598023,
      "epoch": 1.6882804995196925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047190548480713736,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23676184564828873,
      "num_tokens": 40258262.0,
      "step": 17575
    },
    {
      "entropy": 5.359549856185913,
      "epoch": 1.6887608069164264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047188889346462163,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22988192290067672,
      "num_tokens": 40268735.0,
      "step": 17580
    },
    {
      "entropy": 5.29575743675232,
      "epoch": 1.6892411143131603,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047187229755093037,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2273739069700241,
      "num_tokens": 40279905.0,
      "step": 17585
    },
    {
      "entropy": 5.306499052047729,
      "epoch": 1.6897214217098944,
      "grad_norm": 1.046875,
      "learning_rate": 0.000471855697066449,
      "loss": 4.8309,
      "mean_token_accuracy": 0.22931000739336013,
      "num_tokens": 40290580.0,
      "step": 17590
    },
    {
      "entropy": 5.302533388137817,
      "epoch": 1.6902017291066282,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047183909201156297,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22898447513580322,
      "num_tokens": 40302472.0,
      "step": 17595
    },
    {
      "entropy": 5.371036195755005,
      "epoch": 1.6906820365033621,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004718224823866576,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22539771795272828,
      "num_tokens": 40314238.0,
      "step": 17600
    },
    {
      "entropy": 5.269080781936646,
      "epoch": 1.691162343900096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004718058681921186,
      "loss": 4.806,
      "mean_token_accuracy": 0.23762340545654298,
      "num_tokens": 40326147.0,
      "step": 17605
    },
    {
      "entropy": 5.2834125518798825,
      "epoch": 1.6916426512968301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047178924942833185,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23668752014636993,
      "num_tokens": 40338210.0,
      "step": 17610
    },
    {
      "entropy": 5.254216480255127,
      "epoch": 1.692122958693564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004717726260956831,
      "loss": 4.7461,
      "mean_token_accuracy": 0.23733062595129012,
      "num_tokens": 40349293.0,
      "step": 17615
    },
    {
      "entropy": 5.255743837356567,
      "epoch": 1.6926032660902979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004717559981945581,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23313410580158234,
      "num_tokens": 40360916.0,
      "step": 17620
    },
    {
      "entropy": 5.290140295028687,
      "epoch": 1.6930835734870318,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004717393657253432,
      "loss": 4.856,
      "mean_token_accuracy": 0.23208979815244674,
      "num_tokens": 40373525.0,
      "step": 17625
    },
    {
      "entropy": 5.353207731246949,
      "epoch": 1.6935638808837656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004717227286884243,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2317865386605263,
      "num_tokens": 40385286.0,
      "step": 17630
    },
    {
      "entropy": 5.30309042930603,
      "epoch": 1.6940441882804995,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004717060870841879,
      "loss": 4.747,
      "mean_token_accuracy": 0.2382659897208214,
      "num_tokens": 40396371.0,
      "step": 17635
    },
    {
      "entropy": 5.230661201477051,
      "epoch": 1.6945244956772334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004716894409130202,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23740303218364717,
      "num_tokens": 40407304.0,
      "step": 17640
    },
    {
      "entropy": 5.281907081604004,
      "epoch": 1.6950048030739673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004716727901753078,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23636680394411086,
      "num_tokens": 40418384.0,
      "step": 17645
    },
    {
      "entropy": 5.346883344650268,
      "epoch": 1.6954851104707012,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004716561348714371,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23025813400745393,
      "num_tokens": 40430603.0,
      "step": 17650
    },
    {
      "entropy": 5.2801093578338625,
      "epoch": 1.695965417867435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047163947500179494,
      "loss": 4.7709,
      "mean_token_accuracy": 0.2375144049525261,
      "num_tokens": 40442597.0,
      "step": 17655
    },
    {
      "entropy": 5.234930467605591,
      "epoch": 1.696445725264169,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004716228105667681,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24814340919256211,
      "num_tokens": 40454078.0,
      "step": 17660
    },
    {
      "entropy": 5.253095245361328,
      "epoch": 1.696926032660903,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004716061415667435,
      "loss": 4.7664,
      "mean_token_accuracy": 0.2439969524741173,
      "num_tokens": 40465561.0,
      "step": 17665
    },
    {
      "entropy": 5.301808786392212,
      "epoch": 1.697406340057637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000471589468002108,
      "loss": 4.8805,
      "mean_token_accuracy": 0.23283839374780654,
      "num_tokens": 40476883.0,
      "step": 17670
    },
    {
      "entropy": 5.321509599685669,
      "epoch": 1.6978866474543708,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004715727898732488,
      "loss": 4.8255,
      "mean_token_accuracy": 0.236427403986454,
      "num_tokens": 40488139.0,
      "step": 17675
    },
    {
      "entropy": 5.3144941329956055,
      "epoch": 1.6983669548511047,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047155610718055315,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23230697959661484,
      "num_tokens": 40499367.0,
      "step": 17680
    },
    {
      "entropy": 5.199106121063233,
      "epoch": 1.6988472622478388,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047153941992440833,
      "loss": 4.7852,
      "mean_token_accuracy": 0.225916551053524,
      "num_tokens": 40510628.0,
      "step": 17685
    },
    {
      "entropy": 5.228081703186035,
      "epoch": 1.6993275696445727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004715227281052018,
      "loss": 4.7796,
      "mean_token_accuracy": 0.23907490223646163,
      "num_tokens": 40522680.0,
      "step": 17690
    },
    {
      "entropy": 5.282255601882935,
      "epoch": 1.6998078770413065,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004715060317233211,
      "loss": 4.7114,
      "mean_token_accuracy": 0.2506225049495697,
      "num_tokens": 40533139.0,
      "step": 17695
    },
    {
      "entropy": 5.327533912658692,
      "epoch": 1.7002881844380404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004714893307791538,
      "loss": 4.8793,
      "mean_token_accuracy": 0.23577014356851578,
      "num_tokens": 40544578.0,
      "step": 17700
    },
    {
      "entropy": 5.361743831634522,
      "epoch": 1.7007684918347743,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047147262527308766,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23196100294589997,
      "num_tokens": 40555667.0,
      "step": 17705
    },
    {
      "entropy": 5.264175224304199,
      "epoch": 1.7012487992315082,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004714559152055106,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2410658612847328,
      "num_tokens": 40567123.0,
      "step": 17710
    },
    {
      "entropy": 5.2780615329742435,
      "epoch": 1.701729106628242,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004714392005768106,
      "loss": 4.8279,
      "mean_token_accuracy": 0.2333737552165985,
      "num_tokens": 40579692.0,
      "step": 17715
    },
    {
      "entropy": 5.277743673324585,
      "epoch": 1.702209414024976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004714224813873756,
      "loss": 4.7987,
      "mean_token_accuracy": 0.2388721689581871,
      "num_tokens": 40590989.0,
      "step": 17720
    },
    {
      "entropy": 5.25553069114685,
      "epoch": 1.7026897214217098,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047140575763759393,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23751717805862427,
      "num_tokens": 40602050.0,
      "step": 17725
    },
    {
      "entropy": 5.250879621505737,
      "epoch": 1.7031700288184437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047138902932785363,
      "loss": 4.7988,
      "mean_token_accuracy": 0.24047496020793915,
      "num_tokens": 40614552.0,
      "step": 17730
    },
    {
      "entropy": 5.3046448707580565,
      "epoch": 1.7036503362151776,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047137229645854333,
      "loss": 4.7772,
      "mean_token_accuracy": 0.24345964789390565,
      "num_tokens": 40625903.0,
      "step": 17735
    },
    {
      "entropy": 5.315084409713745,
      "epoch": 1.7041306436119115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004713555590300513,
      "loss": 4.8427,
      "mean_token_accuracy": 0.23305753916501998,
      "num_tokens": 40638634.0,
      "step": 17740
    },
    {
      "entropy": 5.321714067459107,
      "epoch": 1.7046109510086456,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004713388170427664,
      "loss": 4.8413,
      "mean_token_accuracy": 0.23132963478565216,
      "num_tokens": 40651279.0,
      "step": 17745
    },
    {
      "entropy": 5.307851934432984,
      "epoch": 1.7050912584053795,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004713220704970771,
      "loss": 4.8343,
      "mean_token_accuracy": 0.23759011775255204,
      "num_tokens": 40662306.0,
      "step": 17750
    },
    {
      "entropy": 5.16213846206665,
      "epoch": 1.7055715658021133,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047130531939337236,
      "loss": 4.6673,
      "mean_token_accuracy": 0.24798649102449416,
      "num_tokens": 40672290.0,
      "step": 17755
    },
    {
      "entropy": 5.2923668384552,
      "epoch": 1.7060518731988472,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047128856373204086,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23497513085603713,
      "num_tokens": 40683447.0,
      "step": 17760
    },
    {
      "entropy": 5.396785163879395,
      "epoch": 1.7065321805955813,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047127180351347184,
      "loss": 4.9421,
      "mean_token_accuracy": 0.227732552587986,
      "num_tokens": 40695230.0,
      "step": 17765
    },
    {
      "entropy": 5.286221933364868,
      "epoch": 1.7070124879923152,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004712550387380544,
      "loss": 4.7869,
      "mean_token_accuracy": 0.23917239904403687,
      "num_tokens": 40707311.0,
      "step": 17770
    },
    {
      "entropy": 5.290457916259766,
      "epoch": 1.707492795389049,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004712382694061776,
      "loss": 4.807,
      "mean_token_accuracy": 0.2275317445397377,
      "num_tokens": 40717928.0,
      "step": 17775
    },
    {
      "entropy": 5.270983409881592,
      "epoch": 1.707973102785783,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047122149551823096,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2304789498448372,
      "num_tokens": 40730355.0,
      "step": 17780
    },
    {
      "entropy": 5.321827030181884,
      "epoch": 1.7084534101825168,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004712047170746039,
      "loss": 4.9052,
      "mean_token_accuracy": 0.23561351597309113,
      "num_tokens": 40741412.0,
      "step": 17785
    },
    {
      "entropy": 5.304529333114624,
      "epoch": 1.7089337175792507,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047118793407568586,
      "loss": 4.8075,
      "mean_token_accuracy": 0.23673325926065444,
      "num_tokens": 40753491.0,
      "step": 17790
    },
    {
      "entropy": 5.287667942047119,
      "epoch": 1.7094140249759846,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047117114652186657,
      "loss": 4.7511,
      "mean_token_accuracy": 0.24016797095537185,
      "num_tokens": 40765209.0,
      "step": 17795
    },
    {
      "entropy": 5.240777587890625,
      "epoch": 1.7098943323727185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047115435441353573,
      "loss": 4.8228,
      "mean_token_accuracy": 0.23461499363183974,
      "num_tokens": 40778065.0,
      "step": 17800
    },
    {
      "entropy": 5.259960174560547,
      "epoch": 1.7103746397694524,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047113755775108333,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2380315348505974,
      "num_tokens": 40789149.0,
      "step": 17805
    },
    {
      "entropy": 5.26594729423523,
      "epoch": 1.7108549471661862,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047112075653489913,
      "loss": 4.8075,
      "mean_token_accuracy": 0.24221380949020385,
      "num_tokens": 40800340.0,
      "step": 17810
    },
    {
      "entropy": 5.299975728988647,
      "epoch": 1.7113352545629201,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004711039507653734,
      "loss": 4.8414,
      "mean_token_accuracy": 0.23722921907901764,
      "num_tokens": 40811866.0,
      "step": 17815
    },
    {
      "entropy": 5.277061843872071,
      "epoch": 1.7118155619596542,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004710871404428961,
      "loss": 4.8531,
      "mean_token_accuracy": 0.22797513902187347,
      "num_tokens": 40825850.0,
      "step": 17820
    },
    {
      "entropy": 5.335097026824951,
      "epoch": 1.7122958693563881,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047107032556785786,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23259138464927673,
      "num_tokens": 40836688.0,
      "step": 17825
    },
    {
      "entropy": 5.348273134231567,
      "epoch": 1.712776176753122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047105350614064874,
      "loss": 4.8207,
      "mean_token_accuracy": 0.23576921373605728,
      "num_tokens": 40847803.0,
      "step": 17830
    },
    {
      "entropy": 5.348256587982178,
      "epoch": 1.7132564841498559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047103668216165944,
      "loss": 4.8729,
      "mean_token_accuracy": 0.23162992298603058,
      "num_tokens": 40859099.0,
      "step": 17835
    },
    {
      "entropy": 5.2598409175872805,
      "epoch": 1.71373679154659,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047101985363128045,
      "loss": 4.7391,
      "mean_token_accuracy": 0.24492792338132857,
      "num_tokens": 40870440.0,
      "step": 17840
    },
    {
      "entropy": 5.386252975463867,
      "epoch": 1.7142170989433239,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047100302054990255,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2253631293773651,
      "num_tokens": 40882329.0,
      "step": 17845
    },
    {
      "entropy": 5.321318864822388,
      "epoch": 1.7146974063400577,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004709861829179165,
      "loss": 4.9081,
      "mean_token_accuracy": 0.22894612699747086,
      "num_tokens": 40893458.0,
      "step": 17850
    },
    {
      "entropy": 5.199593496322632,
      "epoch": 1.7151777137367916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047096934073571325,
      "loss": 4.738,
      "mean_token_accuracy": 0.24561861753463746,
      "num_tokens": 40904626.0,
      "step": 17855
    },
    {
      "entropy": 5.217311859130859,
      "epoch": 1.7156580211335255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047095249400368384,
      "loss": 4.7541,
      "mean_token_accuracy": 0.24381903558969498,
      "num_tokens": 40916005.0,
      "step": 17860
    },
    {
      "entropy": 5.232979249954224,
      "epoch": 1.7161383285302594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047093564272221927,
      "loss": 4.7222,
      "mean_token_accuracy": 0.2404430091381073,
      "num_tokens": 40927470.0,
      "step": 17865
    },
    {
      "entropy": 5.3173259735107425,
      "epoch": 1.7166186359269933,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047091878689171105,
      "loss": 4.8603,
      "mean_token_accuracy": 0.239155113697052,
      "num_tokens": 40938968.0,
      "step": 17870
    },
    {
      "entropy": 5.209595966339111,
      "epoch": 1.7170989433237271,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004709019265125502,
      "loss": 4.7509,
      "mean_token_accuracy": 0.2474282369017601,
      "num_tokens": 40952636.0,
      "step": 17875
    },
    {
      "entropy": 5.3298297882080075,
      "epoch": 1.717579250720461,
      "grad_norm": 1.0,
      "learning_rate": 0.00047088506158512837,
      "loss": 4.8638,
      "mean_token_accuracy": 0.23732975423336028,
      "num_tokens": 40964816.0,
      "step": 17880
    },
    {
      "entropy": 5.344283866882324,
      "epoch": 1.718059558117195,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047086819210983714,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2313923329114914,
      "num_tokens": 40977152.0,
      "step": 17885
    },
    {
      "entropy": 5.245797729492187,
      "epoch": 1.7185398655139288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047085131808706813,
      "loss": 4.7984,
      "mean_token_accuracy": 0.24119522869586946,
      "num_tokens": 40987506.0,
      "step": 17890
    },
    {
      "entropy": 5.3131547451019285,
      "epoch": 1.7190201729106627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004708344395172129,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23422798365354539,
      "num_tokens": 40998598.0,
      "step": 17895
    },
    {
      "entropy": 5.2998980522155765,
      "epoch": 1.7195004803073968,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004708175564006636,
      "loss": 4.8416,
      "mean_token_accuracy": 0.234416463971138,
      "num_tokens": 41009644.0,
      "step": 17900
    },
    {
      "entropy": 5.2736091136932375,
      "epoch": 1.7199807877041307,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004708006687378121,
      "loss": 4.7396,
      "mean_token_accuracy": 0.23980935662984848,
      "num_tokens": 41021816.0,
      "step": 17905
    },
    {
      "entropy": 5.218309164047241,
      "epoch": 1.7204610951008645,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004707837765290505,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23915926069021226,
      "num_tokens": 41033482.0,
      "step": 17910
    },
    {
      "entropy": 5.285134267807007,
      "epoch": 1.7209414024975984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707668797747709,
      "loss": 4.7592,
      "mean_token_accuracy": 0.23942258059978486,
      "num_tokens": 41043854.0,
      "step": 17915
    },
    {
      "entropy": 5.278750896453857,
      "epoch": 1.7214217098943325,
      "grad_norm": 1.0,
      "learning_rate": 0.0004707499784753657,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22615695744752884,
      "num_tokens": 41056956.0,
      "step": 17920
    },
    {
      "entropy": 5.373263025283814,
      "epoch": 1.7219020172910664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707330726312273,
      "loss": 4.9153,
      "mean_token_accuracy": 0.22426446825265883,
      "num_tokens": 41068170.0,
      "step": 17925
    },
    {
      "entropy": 5.302338552474976,
      "epoch": 1.7223823246878003,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047071616224274803,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23447800278663636,
      "num_tokens": 41079149.0,
      "step": 17930
    },
    {
      "entropy": 5.355841398239136,
      "epoch": 1.7228626320845342,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004706992473103207,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2313147470355034,
      "num_tokens": 41091039.0,
      "step": 17935
    },
    {
      "entropy": 5.2974122047424315,
      "epoch": 1.723342939481268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047068232783433806,
      "loss": 4.8422,
      "mean_token_accuracy": 0.24009297043085098,
      "num_tokens": 41103318.0,
      "step": 17940
    },
    {
      "entropy": 5.24835205078125,
      "epoch": 1.723823246878002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004706654038151927,
      "loss": 4.7355,
      "mean_token_accuracy": 0.2427227333188057,
      "num_tokens": 41114235.0,
      "step": 17945
    },
    {
      "entropy": 5.237985992431641,
      "epoch": 1.7243035542747358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004706484752532777,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23892873525619507,
      "num_tokens": 41126008.0,
      "step": 17950
    },
    {
      "entropy": 5.232495832443237,
      "epoch": 1.7247838616714697,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004706315421489861,
      "loss": 4.8069,
      "mean_token_accuracy": 0.23547200709581376,
      "num_tokens": 41138819.0,
      "step": 17955
    },
    {
      "entropy": 5.305408716201782,
      "epoch": 1.7252641690682036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004706146045027109,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23612809777259827,
      "num_tokens": 41149389.0,
      "step": 17960
    },
    {
      "entropy": 5.277651453018189,
      "epoch": 1.7257444764649374,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004705976623148455,
      "loss": 4.7806,
      "mean_token_accuracy": 0.22968159019947051,
      "num_tokens": 41161810.0,
      "step": 17965
    },
    {
      "entropy": 5.292064523696899,
      "epoch": 1.7262247838616713,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047058071558578324,
      "loss": 4.7855,
      "mean_token_accuracy": 0.2390703335404396,
      "num_tokens": 41172903.0,
      "step": 17970
    },
    {
      "entropy": 5.317517471313477,
      "epoch": 1.7267050912584054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004705637643159175,
      "loss": 4.8588,
      "mean_token_accuracy": 0.24027514159679414,
      "num_tokens": 41183905.0,
      "step": 17975
    },
    {
      "entropy": 5.288000106811523,
      "epoch": 1.7271853986551393,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047054680850564185,
      "loss": 4.7812,
      "mean_token_accuracy": 0.2398231714963913,
      "num_tokens": 41195921.0,
      "step": 17980
    },
    {
      "entropy": 5.229054069519043,
      "epoch": 1.7276657060518732,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004705298481553499,
      "loss": 4.8222,
      "mean_token_accuracy": 0.23855210095643997,
      "num_tokens": 41208287.0,
      "step": 17985
    },
    {
      "entropy": 5.305472040176392,
      "epoch": 1.728146013448607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047051288326543553,
      "loss": 4.8419,
      "mean_token_accuracy": 0.22867176532745362,
      "num_tokens": 41219864.0,
      "step": 17990
    },
    {
      "entropy": 5.283067464828491,
      "epoch": 1.7286263208453412,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047049591383629247,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23527870178222657,
      "num_tokens": 41230640.0,
      "step": 17995
    },
    {
      "entropy": 5.283354616165161,
      "epoch": 1.729106628242075,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047047893986831493,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23845068067312242,
      "num_tokens": 41242413.0,
      "step": 18000
    },
    {
      "epoch": 1.729106628242075,
      "eval_entropy": 5.079079358650962,
      "eval_loss": 4.8727898597717285,
      "eval_mean_token_accuracy": 0.2415365377465569,
      "eval_num_tokens": 41242413.0,
      "eval_runtime": 20.507,
      "eval_samples_per_second": 1600.184,
      "eval_steps_per_second": 200.029,
      "step": 18000
    },
    {
      "entropy": 5.3202817916870115,
      "epoch": 1.729586935638809,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047046196136189686,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23332198560237885,
      "num_tokens": 41254400.0,
      "step": 18005
    },
    {
      "entropy": 5.282455635070801,
      "epoch": 1.7300672430355428,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004704449783174323,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23446881473064424,
      "num_tokens": 41266849.0,
      "step": 18010
    },
    {
      "entropy": 5.280104112625122,
      "epoch": 1.7305475504322767,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004704279907353158,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2320459306240082,
      "num_tokens": 41278179.0,
      "step": 18015
    },
    {
      "entropy": 5.2924645900726315,
      "epoch": 1.7310278578290106,
      "grad_norm": 1.0,
      "learning_rate": 0.00047041099861594167,
      "loss": 4.8067,
      "mean_token_accuracy": 0.24009425938129425,
      "num_tokens": 41289904.0,
      "step": 18020
    },
    {
      "entropy": 5.398792362213134,
      "epoch": 1.7315081652257445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004703940019597044,
      "loss": 4.9513,
      "mean_token_accuracy": 0.22348021864891052,
      "num_tokens": 41301861.0,
      "step": 18025
    },
    {
      "entropy": 5.288413858413696,
      "epoch": 1.7319884726224783,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047037700076699857,
      "loss": 4.8058,
      "mean_token_accuracy": 0.2350868061184883,
      "num_tokens": 41312420.0,
      "step": 18030
    },
    {
      "entropy": 5.254942560195923,
      "epoch": 1.7324687800192122,
      "grad_norm": 1.09375,
      "learning_rate": 0.000470359995038219,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23995097428560258,
      "num_tokens": 41323264.0,
      "step": 18035
    },
    {
      "entropy": 5.351723146438599,
      "epoch": 1.732949087415946,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004703429847737604,
      "loss": 4.872,
      "mean_token_accuracy": 0.23323709964752198,
      "num_tokens": 41335438.0,
      "step": 18040
    },
    {
      "entropy": 5.31868691444397,
      "epoch": 1.73342939481268,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004703259699740177,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22757987082004547,
      "num_tokens": 41347295.0,
      "step": 18045
    },
    {
      "entropy": 5.296656847000122,
      "epoch": 1.7339097022094139,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047030895063938607,
      "loss": 4.8678,
      "mean_token_accuracy": 0.22630307376384734,
      "num_tokens": 41358886.0,
      "step": 18050
    },
    {
      "entropy": 5.240004348754883,
      "epoch": 1.734390009606148,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047029192677026043,
      "loss": 4.7013,
      "mean_token_accuracy": 0.2471020758152008,
      "num_tokens": 41369402.0,
      "step": 18055
    },
    {
      "entropy": 5.3987809181213375,
      "epoch": 1.7348703170028819,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004702748983670363,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23100990504026414,
      "num_tokens": 41379772.0,
      "step": 18060
    },
    {
      "entropy": 5.3544622421264645,
      "epoch": 1.7353506243996157,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004702578654301088,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22762602418661118,
      "num_tokens": 41389776.0,
      "step": 18065
    },
    {
      "entropy": 5.1944478988647464,
      "epoch": 1.7358309317963496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004702408279598734,
      "loss": 4.7344,
      "mean_token_accuracy": 0.23397865891456604,
      "num_tokens": 41400640.0,
      "step": 18070
    },
    {
      "entropy": 5.205779600143432,
      "epoch": 1.7363112391930837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004702237859567258,
      "loss": 4.7687,
      "mean_token_accuracy": 0.23274911791086197,
      "num_tokens": 41412653.0,
      "step": 18075
    },
    {
      "entropy": 5.382936954498291,
      "epoch": 1.7367915465898176,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004702067394210616,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2256391540169716,
      "num_tokens": 41423815.0,
      "step": 18080
    },
    {
      "entropy": 5.321983909606933,
      "epoch": 1.7372718539865515,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047018968835327643,
      "loss": 4.8201,
      "mean_token_accuracy": 0.23229666650295258,
      "num_tokens": 41436016.0,
      "step": 18085
    },
    {
      "entropy": 5.245569801330566,
      "epoch": 1.7377521613832854,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004701726327537664,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23963469117879868,
      "num_tokens": 41447837.0,
      "step": 18090
    },
    {
      "entropy": 5.324893188476563,
      "epoch": 1.7382324687800192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004701555726229274,
      "loss": 4.8431,
      "mean_token_accuracy": 0.2378482460975647,
      "num_tokens": 41459816.0,
      "step": 18095
    },
    {
      "entropy": 5.3038125991821286,
      "epoch": 1.7387127761767531,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004701385079611555,
      "loss": 4.7896,
      "mean_token_accuracy": 0.2380923420190811,
      "num_tokens": 41471461.0,
      "step": 18100
    },
    {
      "entropy": 5.280284357070923,
      "epoch": 1.739193083573487,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047012143876884677,
      "loss": 4.7667,
      "mean_token_accuracy": 0.24543946385383605,
      "num_tokens": 41481886.0,
      "step": 18105
    },
    {
      "entropy": 5.225103950500488,
      "epoch": 1.739673390970221,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004701043650463977,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23589994609355927,
      "num_tokens": 41493698.0,
      "step": 18110
    },
    {
      "entropy": 5.274262952804565,
      "epoch": 1.7401536983669548,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004700872867942046,
      "loss": 4.7603,
      "mean_token_accuracy": 0.2462306722998619,
      "num_tokens": 41504744.0,
      "step": 18115
    },
    {
      "entropy": 5.330080175399781,
      "epoch": 1.7406340057636887,
      "grad_norm": 1.015625,
      "learning_rate": 0.000470070204012664,
      "loss": 4.8396,
      "mean_token_accuracy": 0.23856029212474822,
      "num_tokens": 41517129.0,
      "step": 18120
    },
    {
      "entropy": 5.365573406219482,
      "epoch": 1.7411143131604225,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047005311670217256,
      "loss": 4.8259,
      "mean_token_accuracy": 0.22922660261392594,
      "num_tokens": 41529969.0,
      "step": 18125
    },
    {
      "entropy": 5.277237272262573,
      "epoch": 1.7415946205571564,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047003602486312687,
      "loss": 4.7647,
      "mean_token_accuracy": 0.24275984913110732,
      "num_tokens": 41541515.0,
      "step": 18130
    },
    {
      "entropy": 5.247265291213989,
      "epoch": 1.7420749279538905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004700189284959238,
      "loss": 4.8486,
      "mean_token_accuracy": 0.2334946021437645,
      "num_tokens": 41553045.0,
      "step": 18135
    },
    {
      "entropy": 5.306149959564209,
      "epoch": 1.7425552353506244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047000182760096037,
      "loss": 4.865,
      "mean_token_accuracy": 0.23304492384195327,
      "num_tokens": 41565376.0,
      "step": 18140
    },
    {
      "entropy": 5.280718040466309,
      "epoch": 1.7430355427473583,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004699847221786335,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22678992301225662,
      "num_tokens": 41577515.0,
      "step": 18145
    },
    {
      "entropy": 5.239706516265869,
      "epoch": 1.7435158501440924,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004699676122293403,
      "loss": 4.7352,
      "mean_token_accuracy": 0.2377943903207779,
      "num_tokens": 41588589.0,
      "step": 18150
    },
    {
      "entropy": 5.335580921173095,
      "epoch": 1.7439961575408263,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004699504977534782,
      "loss": 4.8943,
      "mean_token_accuracy": 0.22800132036209106,
      "num_tokens": 41600727.0,
      "step": 18155
    },
    {
      "entropy": 5.329735612869262,
      "epoch": 1.7444764649375601,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004699333787514444,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23897137343883515,
      "num_tokens": 41612177.0,
      "step": 18160
    },
    {
      "entropy": 5.318365669250488,
      "epoch": 1.744956772334294,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004699162552236363,
      "loss": 4.8202,
      "mean_token_accuracy": 0.234774711728096,
      "num_tokens": 41623749.0,
      "step": 18165
    },
    {
      "entropy": 5.321715545654297,
      "epoch": 1.745437079731028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046989912717045165,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23878899216651917,
      "num_tokens": 41635549.0,
      "step": 18170
    },
    {
      "entropy": 5.330961608886719,
      "epoch": 1.7459173871277618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046988199459228793,
      "loss": 4.9072,
      "mean_token_accuracy": 0.2275061845779419,
      "num_tokens": 41647219.0,
      "step": 18175
    },
    {
      "entropy": 5.374346446990967,
      "epoch": 1.7463976945244957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004698648574895429,
      "loss": 4.8369,
      "mean_token_accuracy": 0.23612894266843795,
      "num_tokens": 41658850.0,
      "step": 18180
    },
    {
      "entropy": 5.366776657104492,
      "epoch": 1.7468780019212296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046984771586261465,
      "loss": 4.8627,
      "mean_token_accuracy": 0.22605449557304383,
      "num_tokens": 41669304.0,
      "step": 18185
    },
    {
      "entropy": 5.349964809417725,
      "epoch": 1.7473583093179634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000469830569711901,
      "loss": 4.8876,
      "mean_token_accuracy": 0.23527871966362,
      "num_tokens": 41681652.0,
      "step": 18190
    },
    {
      "entropy": 5.308848094940186,
      "epoch": 1.7478386167146973,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004698134190377999,
      "loss": 4.8362,
      "mean_token_accuracy": 0.230247201025486,
      "num_tokens": 41692548.0,
      "step": 18195
    },
    {
      "entropy": 5.304030656814575,
      "epoch": 1.7483189241114312,
      "grad_norm": 1.0,
      "learning_rate": 0.00046979626384070983,
      "loss": 4.8159,
      "mean_token_accuracy": 0.231788532435894,
      "num_tokens": 41704528.0,
      "step": 18200
    },
    {
      "entropy": 5.273084115982056,
      "epoch": 1.748799231508165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000469779104121029,
      "loss": 4.7885,
      "mean_token_accuracy": 0.2419068232178688,
      "num_tokens": 41717314.0,
      "step": 18205
    },
    {
      "entropy": 5.385190868377686,
      "epoch": 1.7492795389048992,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046976193987915553,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22901023626327516,
      "num_tokens": 41729193.0,
      "step": 18210
    },
    {
      "entropy": 5.303673982620239,
      "epoch": 1.749759846301633,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004697447711154883,
      "loss": 4.7636,
      "mean_token_accuracy": 0.2396356239914894,
      "num_tokens": 41741208.0,
      "step": 18215
    },
    {
      "entropy": 5.282691097259521,
      "epoch": 1.750240153698367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046972759783042576,
      "loss": 4.7795,
      "mean_token_accuracy": 0.23731310665607452,
      "num_tokens": 41752792.0,
      "step": 18220
    },
    {
      "entropy": 5.319668579101562,
      "epoch": 1.7507204610951008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004697104200243666,
      "loss": 4.7982,
      "mean_token_accuracy": 0.23797611743211747,
      "num_tokens": 41763629.0,
      "step": 18225
    },
    {
      "entropy": 5.243752717971802,
      "epoch": 1.751200768491835,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004696932376977096,
      "loss": 4.7783,
      "mean_token_accuracy": 0.24239349216222764,
      "num_tokens": 41774502.0,
      "step": 18230
    },
    {
      "entropy": 5.301419401168824,
      "epoch": 1.7516810758885688,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004696760508508538,
      "loss": 4.8254,
      "mean_token_accuracy": 0.23149576634168625,
      "num_tokens": 41785700.0,
      "step": 18235
    },
    {
      "entropy": 5.222524499893188,
      "epoch": 1.7521613832853027,
      "grad_norm": 1.0,
      "learning_rate": 0.00046965885948419814,
      "loss": 4.7898,
      "mean_token_accuracy": 0.23379357755184174,
      "num_tokens": 41797347.0,
      "step": 18240
    },
    {
      "entropy": 5.247923135757446,
      "epoch": 1.7526416906820366,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004696416635981418,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23169222921133042,
      "num_tokens": 41810866.0,
      "step": 18245
    },
    {
      "entropy": 5.311535453796386,
      "epoch": 1.7531219980787704,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000469624463193084,
      "loss": 4.8443,
      "mean_token_accuracy": 0.23481549620628356,
      "num_tokens": 41821786.0,
      "step": 18250
    },
    {
      "entropy": 5.334315538406372,
      "epoch": 1.7536023054755043,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000469607258269424,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22994401007890702,
      "num_tokens": 41833591.0,
      "step": 18255
    },
    {
      "entropy": 5.253770065307617,
      "epoch": 1.7540826128722382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004695900488275614,
      "loss": 4.738,
      "mean_token_accuracy": 0.24472519755363464,
      "num_tokens": 41844659.0,
      "step": 18260
    },
    {
      "entropy": 5.252942752838135,
      "epoch": 1.754562920268972,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004695728348678957,
      "loss": 4.796,
      "mean_token_accuracy": 0.23012185990810394,
      "num_tokens": 41856378.0,
      "step": 18265
    },
    {
      "entropy": 5.287821865081787,
      "epoch": 1.755043227665706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695556163908265,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2326225832104683,
      "num_tokens": 41867624.0,
      "step": 18270
    },
    {
      "entropy": 5.298394584655762,
      "epoch": 1.7555235350624399,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004695383933967536,
      "loss": 4.822,
      "mean_token_accuracy": 0.2436641499400139,
      "num_tokens": 41879250.0,
      "step": 18275
    },
    {
      "entropy": 5.3042881965637205,
      "epoch": 1.7560038424591737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046952116588607694,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2357711300253868,
      "num_tokens": 41890713.0,
      "step": 18280
    },
    {
      "entropy": 5.326321458816528,
      "epoch": 1.7564841498559076,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004695039338591963,
      "loss": 4.869,
      "mean_token_accuracy": 0.2332359343767166,
      "num_tokens": 41902924.0,
      "step": 18285
    },
    {
      "entropy": 5.352136468887329,
      "epoch": 1.7569644572526417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000469486697316512,
      "loss": 4.8799,
      "mean_token_accuracy": 0.23191313147544862,
      "num_tokens": 41914498.0,
      "step": 18290
    },
    {
      "entropy": 5.299045276641846,
      "epoch": 1.7574447646493756,
      "grad_norm": 1.03125,
      "learning_rate": 0.000469469456258424,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23328113704919815,
      "num_tokens": 41925258.0,
      "step": 18295
    },
    {
      "entropy": 5.299666881561279,
      "epoch": 1.7579250720461095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004694522106853327,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23197586983442306,
      "num_tokens": 41938176.0,
      "step": 18300
    },
    {
      "entropy": 5.272807216644287,
      "epoch": 1.7584053794428436,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046943496059763845,
      "loss": 4.6885,
      "mean_token_accuracy": 0.24613060802221298,
      "num_tokens": 41949160.0,
      "step": 18305
    },
    {
      "entropy": 5.274253273010254,
      "epoch": 1.7588856868395775,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046941770599574176,
      "loss": 4.8733,
      "mean_token_accuracy": 0.2353225514292717,
      "num_tokens": 41960981.0,
      "step": 18310
    },
    {
      "entropy": 5.215790748596191,
      "epoch": 1.7593659942363113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004694004468800433,
      "loss": 4.823,
      "mean_token_accuracy": 0.23405954092741013,
      "num_tokens": 41971755.0,
      "step": 18315
    },
    {
      "entropy": 5.2936278820037845,
      "epoch": 1.7598463016330452,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004693831832509437,
      "loss": 4.79,
      "mean_token_accuracy": 0.24012883603572846,
      "num_tokens": 41984337.0,
      "step": 18320
    },
    {
      "entropy": 5.266330528259277,
      "epoch": 1.760326609029779,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046936591510884375,
      "loss": 4.7978,
      "mean_token_accuracy": 0.23980507552623748,
      "num_tokens": 41995648.0,
      "step": 18325
    },
    {
      "entropy": 5.252077341079712,
      "epoch": 1.760806916426513,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046934864245414443,
      "loss": 4.774,
      "mean_token_accuracy": 0.24315641075372696,
      "num_tokens": 42006995.0,
      "step": 18330
    },
    {
      "entropy": 5.377856636047364,
      "epoch": 1.7612872238232469,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046933136528724676,
      "loss": 4.9085,
      "mean_token_accuracy": 0.22569063156843186,
      "num_tokens": 42020074.0,
      "step": 18335
    },
    {
      "entropy": 5.254903650283813,
      "epoch": 1.7617675312199808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004693140836085518,
      "loss": 4.7272,
      "mean_token_accuracy": 0.2405659720301628,
      "num_tokens": 42032053.0,
      "step": 18340
    },
    {
      "entropy": 5.259626770019532,
      "epoch": 1.7622478386167146,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046929679741846076,
      "loss": 4.7572,
      "mean_token_accuracy": 0.24445926249027253,
      "num_tokens": 42042223.0,
      "step": 18345
    },
    {
      "entropy": 5.295467853546143,
      "epoch": 1.7627281460134485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046927950671737505,
      "loss": 4.8285,
      "mean_token_accuracy": 0.24241833686828612,
      "num_tokens": 42053988.0,
      "step": 18350
    },
    {
      "entropy": 5.202009820938111,
      "epoch": 1.7632084534101824,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046926221150569617,
      "loss": 4.7268,
      "mean_token_accuracy": 0.2473411276936531,
      "num_tokens": 42064451.0,
      "step": 18355
    },
    {
      "entropy": 5.275821781158447,
      "epoch": 1.7636887608069163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004692449117838255,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2248113676905632,
      "num_tokens": 42076503.0,
      "step": 18360
    },
    {
      "entropy": 5.342512321472168,
      "epoch": 1.7641690682036504,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692276075521648,
      "loss": 4.8294,
      "mean_token_accuracy": 0.22852211594581603,
      "num_tokens": 42088341.0,
      "step": 18365
    },
    {
      "entropy": 5.294155645370483,
      "epoch": 1.7646493756003843,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004692102988111158,
      "loss": 4.7885,
      "mean_token_accuracy": 0.23985744565725325,
      "num_tokens": 42099862.0,
      "step": 18370
    },
    {
      "entropy": 5.3137092113494875,
      "epoch": 1.7651296829971181,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046919298556108023,
      "loss": 4.8999,
      "mean_token_accuracy": 0.23541828691959382,
      "num_tokens": 42112117.0,
      "step": 18375
    },
    {
      "entropy": 5.334254598617553,
      "epoch": 1.765609990393852,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046917566780246036,
      "loss": 4.8226,
      "mean_token_accuracy": 0.232405486702919,
      "num_tokens": 42123093.0,
      "step": 18380
    },
    {
      "entropy": 5.360133266448974,
      "epoch": 1.7660902977905861,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046915834553565793,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22876425087451935,
      "num_tokens": 42135266.0,
      "step": 18385
    },
    {
      "entropy": 5.428660249710083,
      "epoch": 1.76657060518732,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004691410187610753,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22425400465726852,
      "num_tokens": 42145743.0,
      "step": 18390
    },
    {
      "entropy": 5.248659658432007,
      "epoch": 1.767050912584054,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046912368747911465,
      "loss": 4.7321,
      "mean_token_accuracy": 0.24536035358905792,
      "num_tokens": 42157604.0,
      "step": 18395
    },
    {
      "entropy": 5.273485946655273,
      "epoch": 1.7675312199807878,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046910635169017845,
      "loss": 4.7525,
      "mean_token_accuracy": 0.24364032447338105,
      "num_tokens": 42168738.0,
      "step": 18400
    },
    {
      "entropy": 5.32856936454773,
      "epoch": 1.7680115273775217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004690890113946691,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22666967660188675,
      "num_tokens": 42179699.0,
      "step": 18405
    },
    {
      "entropy": 5.284949684143067,
      "epoch": 1.7684918347742555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004690716665929893,
      "loss": 4.8028,
      "mean_token_accuracy": 0.2313702031970024,
      "num_tokens": 42190512.0,
      "step": 18410
    },
    {
      "entropy": 5.336330413818359,
      "epoch": 1.7689721421709894,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046905431728554164,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23195969313383102,
      "num_tokens": 42200996.0,
      "step": 18415
    },
    {
      "entropy": 5.2491919040679935,
      "epoch": 1.7694524495677233,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046903696347272894,
      "loss": 4.7765,
      "mean_token_accuracy": 0.24210900962352752,
      "num_tokens": 42213032.0,
      "step": 18420
    },
    {
      "entropy": 5.2850847244262695,
      "epoch": 1.7699327569644572,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046901960515495413,
      "loss": 4.8546,
      "mean_token_accuracy": 0.23185303062200546,
      "num_tokens": 42224366.0,
      "step": 18425
    },
    {
      "entropy": 5.220268774032593,
      "epoch": 1.770413064361191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004690022423326202,
      "loss": 4.7164,
      "mean_token_accuracy": 0.24618435353040696,
      "num_tokens": 42235540.0,
      "step": 18430
    },
    {
      "entropy": 5.24171633720398,
      "epoch": 1.770893371757925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004689848750061303,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23878249377012253,
      "num_tokens": 42246755.0,
      "step": 18435
    },
    {
      "entropy": 5.257907342910767,
      "epoch": 1.7713736791546588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004689675031758876,
      "loss": 4.7762,
      "mean_token_accuracy": 0.23612100034952163,
      "num_tokens": 42259536.0,
      "step": 18440
    },
    {
      "entropy": 5.274548625946045,
      "epoch": 1.771853986551393,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004689501268422954,
      "loss": 4.7753,
      "mean_token_accuracy": 0.23695267885923385,
      "num_tokens": 42270813.0,
      "step": 18445
    },
    {
      "entropy": 5.4143647193908695,
      "epoch": 1.7723342939481268,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046893274600575725,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2266698345541954,
      "num_tokens": 42281953.0,
      "step": 18450
    },
    {
      "entropy": 5.368619823455811,
      "epoch": 1.7728146013448607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004689153606666765,
      "loss": 4.9283,
      "mean_token_accuracy": 0.22546537220478058,
      "num_tokens": 42294041.0,
      "step": 18455
    },
    {
      "entropy": 5.301262378692627,
      "epoch": 1.7732949087415946,
      "grad_norm": 0.984375,
      "learning_rate": 0.000468897970825457,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24078612476587297,
      "num_tokens": 42305812.0,
      "step": 18460
    },
    {
      "entropy": 5.266234397888184,
      "epoch": 1.7737752161383287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046888057648250233,
      "loss": 4.7262,
      "mean_token_accuracy": 0.24230654686689376,
      "num_tokens": 42317585.0,
      "step": 18465
    },
    {
      "entropy": 5.306941223144531,
      "epoch": 1.7742555235350626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004688631776382164,
      "loss": 4.7866,
      "mean_token_accuracy": 0.23647605776786804,
      "num_tokens": 42328723.0,
      "step": 18470
    },
    {
      "entropy": 5.2429771900177,
      "epoch": 1.7747358309317964,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046884577429300305,
      "loss": 4.8099,
      "mean_token_accuracy": 0.24363051056861879,
      "num_tokens": 42340741.0,
      "step": 18475
    },
    {
      "entropy": 5.285452127456665,
      "epoch": 1.7752161383285303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004688283664472665,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2332512468099594,
      "num_tokens": 42351504.0,
      "step": 18480
    },
    {
      "entropy": 5.316853380203247,
      "epoch": 1.7756964457252642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046881095410141084,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23391344845294954,
      "num_tokens": 42362920.0,
      "step": 18485
    },
    {
      "entropy": 5.286597061157226,
      "epoch": 1.776176753121998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046879353725584036,
      "loss": 4.7477,
      "mean_token_accuracy": 0.23355754911899568,
      "num_tokens": 42374523.0,
      "step": 18490
    },
    {
      "entropy": 5.3182190418243405,
      "epoch": 1.776657060518732,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046877611591095923,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2320338100194931,
      "num_tokens": 42384988.0,
      "step": 18495
    },
    {
      "entropy": 5.22966570854187,
      "epoch": 1.7771373679154658,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046875869006717224,
      "loss": 4.7373,
      "mean_token_accuracy": 0.24271266907453537,
      "num_tokens": 42396503.0,
      "step": 18500
    },
    {
      "entropy": 5.35854139328003,
      "epoch": 1.7776176753121997,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046874125972488375,
      "loss": 4.8494,
      "mean_token_accuracy": 0.23260304927825928,
      "num_tokens": 42406877.0,
      "step": 18505
    },
    {
      "entropy": 5.3048248291015625,
      "epoch": 1.7780979827089336,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046872382488449853,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2403999850153923,
      "num_tokens": 42418010.0,
      "step": 18510
    },
    {
      "entropy": 5.297801828384399,
      "epoch": 1.7785782901056675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046870638554642133,
      "loss": 4.7707,
      "mean_token_accuracy": 0.23542193323373795,
      "num_tokens": 42429056.0,
      "step": 18515
    },
    {
      "entropy": 5.205700349807739,
      "epoch": 1.7790585975024016,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046868894171105704,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23624487072229386,
      "num_tokens": 42442294.0,
      "step": 18520
    },
    {
      "entropy": 5.277652359008789,
      "epoch": 1.7795389048991355,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004686714933788107,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23247092366218566,
      "num_tokens": 42454064.0,
      "step": 18525
    },
    {
      "entropy": 5.2940428256988525,
      "epoch": 1.7800192122958693,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004686540405500873,
      "loss": 4.7859,
      "mean_token_accuracy": 0.23998051285743713,
      "num_tokens": 42465289.0,
      "step": 18530
    },
    {
      "entropy": 5.242935419082642,
      "epoch": 1.7804995196926032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004686365832252922,
      "loss": 4.7131,
      "mean_token_accuracy": 0.240464548766613,
      "num_tokens": 42476662.0,
      "step": 18535
    },
    {
      "entropy": 5.263213491439819,
      "epoch": 1.7809798270893373,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046861912140483056,
      "loss": 4.7916,
      "mean_token_accuracy": 0.2380145773291588,
      "num_tokens": 42486893.0,
      "step": 18540
    },
    {
      "entropy": 5.278890180587768,
      "epoch": 1.7814601344860712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046860165508910787,
      "loss": 4.7886,
      "mean_token_accuracy": 0.2348781481385231,
      "num_tokens": 42498202.0,
      "step": 18545
    },
    {
      "entropy": 5.247362565994263,
      "epoch": 1.781940441882805,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004685841842785296,
      "loss": 4.6723,
      "mean_token_accuracy": 0.2465927168726921,
      "num_tokens": 42509378.0,
      "step": 18550
    },
    {
      "entropy": 5.318032884597779,
      "epoch": 1.782420749279539,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004685667089735014,
      "loss": 4.843,
      "mean_token_accuracy": 0.23445745855569838,
      "num_tokens": 42519955.0,
      "step": 18555
    },
    {
      "entropy": 5.337258148193359,
      "epoch": 1.7829010566762729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046854922917442907,
      "loss": 4.8934,
      "mean_token_accuracy": 0.23056689351797105,
      "num_tokens": 42531896.0,
      "step": 18560
    },
    {
      "entropy": 5.325898504257202,
      "epoch": 1.7833813640730067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004685317448817182,
      "loss": 4.8154,
      "mean_token_accuracy": 0.2406187117099762,
      "num_tokens": 42542123.0,
      "step": 18565
    },
    {
      "entropy": 5.303874397277832,
      "epoch": 1.7838616714697406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004685142560957751,
      "loss": 4.7962,
      "mean_token_accuracy": 0.23223013132810594,
      "num_tokens": 42554034.0,
      "step": 18570
    },
    {
      "entropy": 5.260810613632202,
      "epoch": 1.7843419788664745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004684967628170054,
      "loss": 4.7677,
      "mean_token_accuracy": 0.24101405441761017,
      "num_tokens": 42564532.0,
      "step": 18575
    },
    {
      "entropy": 5.232691717147827,
      "epoch": 1.7848222862632084,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046847926504581553,
      "loss": 4.7581,
      "mean_token_accuracy": 0.23308617621660233,
      "num_tokens": 42576864.0,
      "step": 18580
    },
    {
      "entropy": 5.197420215606689,
      "epoch": 1.7853025936599423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004684617627826116,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24330567419528962,
      "num_tokens": 42588571.0,
      "step": 18585
    },
    {
      "entropy": 5.344934749603271,
      "epoch": 1.7857829010566761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004684442560278001,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2307897314429283,
      "num_tokens": 42600912.0,
      "step": 18590
    },
    {
      "entropy": 5.335548686981201,
      "epoch": 1.78626320845341,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046842674478178727,
      "loss": 4.735,
      "mean_token_accuracy": 0.23970144391059875,
      "num_tokens": 42612929.0,
      "step": 18595
    },
    {
      "entropy": 5.301245260238647,
      "epoch": 1.7867435158501441,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004684092290449798,
      "loss": 4.8269,
      "mean_token_accuracy": 0.23597412109375,
      "num_tokens": 42623688.0,
      "step": 18600
    },
    {
      "entropy": 5.381694316864014,
      "epoch": 1.787223823246878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004683917088177844,
      "loss": 4.9581,
      "mean_token_accuracy": 0.22076814919710158,
      "num_tokens": 42635773.0,
      "step": 18605
    },
    {
      "entropy": 5.304475498199463,
      "epoch": 1.7877041306436119,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004683741841006077,
      "loss": 4.7811,
      "mean_token_accuracy": 0.238293419778347,
      "num_tokens": 42647210.0,
      "step": 18610
    },
    {
      "entropy": 5.325049257278442,
      "epoch": 1.7881844380403458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004683566548938567,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2323343738913536,
      "num_tokens": 42658029.0,
      "step": 18615
    },
    {
      "entropy": 5.270776319503784,
      "epoch": 1.7886647454370799,
      "grad_norm": 1.0,
      "learning_rate": 0.0004683391211979383,
      "loss": 4.8331,
      "mean_token_accuracy": 0.23676474243402482,
      "num_tokens": 42669327.0,
      "step": 18620
    },
    {
      "entropy": 5.31925368309021,
      "epoch": 1.7891450528338138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004683215830132597,
      "loss": 4.7563,
      "mean_token_accuracy": 0.24384283423423767,
      "num_tokens": 42679794.0,
      "step": 18625
    },
    {
      "entropy": 5.326536655426025,
      "epoch": 1.7896253602305476,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046830404034022786,
      "loss": 4.7906,
      "mean_token_accuracy": 0.24541278183460236,
      "num_tokens": 42689838.0,
      "step": 18630
    },
    {
      "entropy": 5.316860914230347,
      "epoch": 1.7901056676272815,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004682864931792502,
      "loss": 4.8467,
      "mean_token_accuracy": 0.2311543643474579,
      "num_tokens": 42700935.0,
      "step": 18635
    },
    {
      "entropy": 5.390247964859009,
      "epoch": 1.7905859750240154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004682689415307342,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22759125977754593,
      "num_tokens": 42712533.0,
      "step": 18640
    },
    {
      "entropy": 5.316840982437133,
      "epoch": 1.7910662824207493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004682513853950872,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23819676637649537,
      "num_tokens": 42723604.0,
      "step": 18645
    },
    {
      "entropy": 5.289678335189819,
      "epoch": 1.7915465898174832,
      "grad_norm": 1.046875,
      "learning_rate": 0.000468233824772717,
      "loss": 4.819,
      "mean_token_accuracy": 0.2378375604748726,
      "num_tokens": 42735957.0,
      "step": 18650
    },
    {
      "entropy": 5.35527548789978,
      "epoch": 1.792026897214217,
      "grad_norm": 1.015625,
      "learning_rate": 0.000468216259664031,
      "loss": 4.9425,
      "mean_token_accuracy": 0.2267468497157097,
      "num_tokens": 42747287.0,
      "step": 18655
    },
    {
      "entropy": 5.292678689956665,
      "epoch": 1.792507204610951,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046819869006943727,
      "loss": 4.7958,
      "mean_token_accuracy": 0.24183426052331924,
      "num_tokens": 42759270.0,
      "step": 18660
    },
    {
      "entropy": 5.218969249725342,
      "epoch": 1.7929875120076848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004681811159893436,
      "loss": 4.7424,
      "mean_token_accuracy": 0.23798127472400665,
      "num_tokens": 42771955.0,
      "step": 18665
    },
    {
      "entropy": 5.384466314315796,
      "epoch": 1.7934678194044187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046816353742415814,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22903222888708114,
      "num_tokens": 42784610.0,
      "step": 18670
    },
    {
      "entropy": 5.352371120452881,
      "epoch": 1.7939481268011528,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046814595437428885,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23625824600458145,
      "num_tokens": 42795221.0,
      "step": 18675
    },
    {
      "entropy": 5.219574785232544,
      "epoch": 1.7944284341978867,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000468128366840144,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2375311404466629,
      "num_tokens": 42807475.0,
      "step": 18680
    },
    {
      "entropy": 5.319950914382934,
      "epoch": 1.7949087415946205,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000468110774822132,
      "loss": 4.7751,
      "mean_token_accuracy": 0.2368447571992874,
      "num_tokens": 42819193.0,
      "step": 18685
    },
    {
      "entropy": 5.336733484268189,
      "epoch": 1.7953890489913544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004680931783206612,
      "loss": 4.8597,
      "mean_token_accuracy": 0.23711795955896378,
      "num_tokens": 42831267.0,
      "step": 18690
    },
    {
      "entropy": 5.35717544555664,
      "epoch": 1.7958693563880885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046807557733614014,
      "loss": 4.8991,
      "mean_token_accuracy": 0.23145951330661774,
      "num_tokens": 42843066.0,
      "step": 18695
    },
    {
      "entropy": 5.274827766418457,
      "epoch": 1.7963496637848224,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046805797186897757,
      "loss": 4.766,
      "mean_token_accuracy": 0.24062791764736174,
      "num_tokens": 42854630.0,
      "step": 18700
    },
    {
      "entropy": 5.388644599914551,
      "epoch": 1.7968299711815563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046804036191958206,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22860129028558732,
      "num_tokens": 42865986.0,
      "step": 18705
    },
    {
      "entropy": 5.382021903991699,
      "epoch": 1.7973102785782902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046802274748836267,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23360093533992768,
      "num_tokens": 42877533.0,
      "step": 18710
    },
    {
      "entropy": 5.273235321044922,
      "epoch": 1.797790585975024,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004680051285757281,
      "loss": 4.7504,
      "mean_token_accuracy": 0.24449097365140915,
      "num_tokens": 42889114.0,
      "step": 18715
    },
    {
      "entropy": 5.367573595046997,
      "epoch": 1.798270893371758,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004679875051820877,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22408441752195357,
      "num_tokens": 42899483.0,
      "step": 18720
    },
    {
      "entropy": 5.305112600326538,
      "epoch": 1.7987512007684918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004679698773078503,
      "loss": 4.7828,
      "mean_token_accuracy": 0.24120235443115234,
      "num_tokens": 42910267.0,
      "step": 18725
    },
    {
      "entropy": 5.258785772323608,
      "epoch": 1.7992315081652257,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046795224495342554,
      "loss": 4.7916,
      "mean_token_accuracy": 0.24203579872846603,
      "num_tokens": 42922440.0,
      "step": 18730
    },
    {
      "entropy": 5.298011207580567,
      "epoch": 1.7997118155619596,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046793460811922255,
      "loss": 4.7674,
      "mean_token_accuracy": 0.2401685893535614,
      "num_tokens": 42933967.0,
      "step": 18735
    },
    {
      "entropy": 5.3096521377563475,
      "epoch": 1.8001921229586935,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046791696680565075,
      "loss": 4.7846,
      "mean_token_accuracy": 0.23967551738023757,
      "num_tokens": 42945049.0,
      "step": 18740
    },
    {
      "entropy": 5.327591466903686,
      "epoch": 1.8006724303554273,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046789932101312003,
      "loss": 4.9006,
      "mean_token_accuracy": 0.2280908405780792,
      "num_tokens": 42956062.0,
      "step": 18745
    },
    {
      "entropy": 5.328690528869629,
      "epoch": 1.8011527377521612,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004678816707420397,
      "loss": 4.8801,
      "mean_token_accuracy": 0.23846964687108993,
      "num_tokens": 42967687.0,
      "step": 18750
    },
    {
      "entropy": 5.3602159976959225,
      "epoch": 1.8016330451488953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004678640159928198,
      "loss": 4.8605,
      "mean_token_accuracy": 0.2210994228720665,
      "num_tokens": 42979916.0,
      "step": 18755
    },
    {
      "entropy": 5.372735214233399,
      "epoch": 1.8021133525456292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004678463567658701,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23284300714731215,
      "num_tokens": 42991639.0,
      "step": 18760
    },
    {
      "entropy": 5.331425666809082,
      "epoch": 1.802593659942363,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004678286930616006,
      "loss": 4.8179,
      "mean_token_accuracy": 0.2431853249669075,
      "num_tokens": 43001843.0,
      "step": 18765
    },
    {
      "entropy": 5.224201440811157,
      "epoch": 1.803073967339097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004678110248804215,
      "loss": 4.762,
      "mean_token_accuracy": 0.2375821590423584,
      "num_tokens": 43013249.0,
      "step": 18770
    },
    {
      "entropy": 5.281113147735596,
      "epoch": 1.803554274735831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046779335222274293,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23980967700481415,
      "num_tokens": 43025867.0,
      "step": 18775
    },
    {
      "entropy": 5.355703115463257,
      "epoch": 1.804034582132565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046777567508897515,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23172237128019332,
      "num_tokens": 43037530.0,
      "step": 18780
    },
    {
      "entropy": 5.309725332260132,
      "epoch": 1.8045148895292988,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046775799347952864,
      "loss": 4.8403,
      "mean_token_accuracy": 0.2352484568953514,
      "num_tokens": 43048675.0,
      "step": 18785
    },
    {
      "entropy": 5.357990884780884,
      "epoch": 1.8049951969260327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004677403073948139,
      "loss": 4.8679,
      "mean_token_accuracy": 0.2369544893503189,
      "num_tokens": 43059592.0,
      "step": 18790
    },
    {
      "entropy": 5.2243246078491214,
      "epoch": 1.8054755043227666,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004677226168352416,
      "loss": 4.691,
      "mean_token_accuracy": 0.23873884826898575,
      "num_tokens": 43071755.0,
      "step": 18795
    },
    {
      "entropy": 5.320462799072265,
      "epoch": 1.8059558117195005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004677049218012223,
      "loss": 4.8534,
      "mean_token_accuracy": 0.236299566924572,
      "num_tokens": 43081801.0,
      "step": 18800
    },
    {
      "entropy": 5.249217224121094,
      "epoch": 1.8064361191162344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000467687222293167,
      "loss": 4.7593,
      "mean_token_accuracy": 0.24274761974811554,
      "num_tokens": 43094160.0,
      "step": 18805
    },
    {
      "entropy": 5.256071662902832,
      "epoch": 1.8069164265129682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004676695183114866,
      "loss": 4.8218,
      "mean_token_accuracy": 0.23502827286720276,
      "num_tokens": 43106082.0,
      "step": 18810
    },
    {
      "entropy": 5.302821016311645,
      "epoch": 1.8073967339097021,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000467651809856592,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23357179313898085,
      "num_tokens": 43117215.0,
      "step": 18815
    },
    {
      "entropy": 5.4054646492004395,
      "epoch": 1.807877041306436,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046763409692889446,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2216537445783615,
      "num_tokens": 43127858.0,
      "step": 18820
    },
    {
      "entropy": 5.34817214012146,
      "epoch": 1.8083573487031699,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046761637952880516,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23880209475755693,
      "num_tokens": 43139952.0,
      "step": 18825
    },
    {
      "entropy": 5.253297138214111,
      "epoch": 1.808837656099904,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046759865765673555,
      "loss": 4.7278,
      "mean_token_accuracy": 0.23873896151781082,
      "num_tokens": 43150900.0,
      "step": 18830
    },
    {
      "entropy": 5.239406776428223,
      "epoch": 1.8093179634966379,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000467580931313097,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23647546023130417,
      "num_tokens": 43163199.0,
      "step": 18835
    },
    {
      "entropy": 5.299229764938355,
      "epoch": 1.8097982708933718,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046756320049830106,
      "loss": 4.7599,
      "mean_token_accuracy": 0.23546827733516693,
      "num_tokens": 43174147.0,
      "step": 18840
    },
    {
      "entropy": 5.272988653182983,
      "epoch": 1.8102785782901056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004675454652127594,
      "loss": 4.744,
      "mean_token_accuracy": 0.2392345130443573,
      "num_tokens": 43186191.0,
      "step": 18845
    },
    {
      "entropy": 5.347779321670532,
      "epoch": 1.8107588856868397,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046752772545688377,
      "loss": 4.9237,
      "mean_token_accuracy": 0.23010241389274597,
      "num_tokens": 43197541.0,
      "step": 18850
    },
    {
      "entropy": 5.3703906536102295,
      "epoch": 1.8112391930835736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004675099812310861,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23175712823867797,
      "num_tokens": 43208792.0,
      "step": 18855
    },
    {
      "entropy": 5.345847702026367,
      "epoch": 1.8117195004803075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004674922325357782,
      "loss": 4.8957,
      "mean_token_accuracy": 0.23193589001893997,
      "num_tokens": 43219914.0,
      "step": 18860
    },
    {
      "entropy": 5.1874096393585205,
      "epoch": 1.8121998078770414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046747447937137235,
      "loss": 4.7447,
      "mean_token_accuracy": 0.24531247913837434,
      "num_tokens": 43232179.0,
      "step": 18865
    },
    {
      "entropy": 5.258132362365723,
      "epoch": 1.8126801152737753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046745672173828057,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23050495684146882,
      "num_tokens": 43243196.0,
      "step": 18870
    },
    {
      "entropy": 5.370373916625977,
      "epoch": 1.8131604226705091,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004674389596369151,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2292989194393158,
      "num_tokens": 43255573.0,
      "step": 18875
    },
    {
      "entropy": 5.2809741497039795,
      "epoch": 1.813640730067243,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046742119306768855,
      "loss": 4.7964,
      "mean_token_accuracy": 0.2351703479886055,
      "num_tokens": 43267679.0,
      "step": 18880
    },
    {
      "entropy": 5.403669643402099,
      "epoch": 1.814121037463977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004674034220310132,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22841188907623292,
      "num_tokens": 43278857.0,
      "step": 18885
    },
    {
      "entropy": 5.371158409118652,
      "epoch": 1.8146013448607108,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046738564652730176,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23753052949905396,
      "num_tokens": 43291099.0,
      "step": 18890
    },
    {
      "entropy": 5.32344069480896,
      "epoch": 1.8150816522574447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004673678665569669,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23103504478931428,
      "num_tokens": 43302863.0,
      "step": 18895
    },
    {
      "entropy": 5.2747414112091064,
      "epoch": 1.8155619596541785,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004673500821204213,
      "loss": 4.7606,
      "mean_token_accuracy": 0.24158115833997726,
      "num_tokens": 43314673.0,
      "step": 18900
    },
    {
      "entropy": 5.28449273109436,
      "epoch": 1.8160422670509124,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467332293218078,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24579012095928193,
      "num_tokens": 43325927.0,
      "step": 18905
    },
    {
      "entropy": 5.298335790634155,
      "epoch": 1.8165225744476465,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046731449985035,
      "loss": 4.833,
      "mean_token_accuracy": 0.2403786912560463,
      "num_tokens": 43337615.0,
      "step": 18910
    },
    {
      "entropy": 5.3472898483276365,
      "epoch": 1.8170028818443804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046729670201765036,
      "loss": 4.8898,
      "mean_token_accuracy": 0.23454538881778716,
      "num_tokens": 43349195.0,
      "step": 18915
    },
    {
      "entropy": 5.374157857894898,
      "epoch": 1.8174831892411143,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046727889972039227,
      "loss": 4.8611,
      "mean_token_accuracy": 0.23599691241979598,
      "num_tokens": 43361274.0,
      "step": 18920
    },
    {
      "entropy": 5.265533399581909,
      "epoch": 1.8179634966378482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046726109295898904,
      "loss": 4.802,
      "mean_token_accuracy": 0.24018369168043135,
      "num_tokens": 43372843.0,
      "step": 18925
    },
    {
      "entropy": 5.294201993942261,
      "epoch": 1.8184438040345823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004672432817338542,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23899088501930238,
      "num_tokens": 43383492.0,
      "step": 18930
    },
    {
      "entropy": 5.301700592041016,
      "epoch": 1.8189241114313162,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046722546604540115,
      "loss": 4.7946,
      "mean_token_accuracy": 0.23428987264633178,
      "num_tokens": 43395669.0,
      "step": 18935
    },
    {
      "entropy": 5.229050493240356,
      "epoch": 1.81940441882805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004672076458940436,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24301036149263383,
      "num_tokens": 43407340.0,
      "step": 18940
    },
    {
      "entropy": 5.267733860015869,
      "epoch": 1.819884726224784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046718982128019534,
      "loss": 4.8334,
      "mean_token_accuracy": 0.2326653152704239,
      "num_tokens": 43418607.0,
      "step": 18945
    },
    {
      "entropy": 5.3898091316223145,
      "epoch": 1.8203650336215178,
      "grad_norm": 1.125,
      "learning_rate": 0.00046717199220427003,
      "loss": 4.9389,
      "mean_token_accuracy": 0.23205358982086183,
      "num_tokens": 43429286.0,
      "step": 18950
    },
    {
      "entropy": 5.2569173812866214,
      "epoch": 1.8208453410182517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046715415866668163,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23424246907234192,
      "num_tokens": 43440794.0,
      "step": 18955
    },
    {
      "entropy": 5.264708471298218,
      "epoch": 1.8213256484149856,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004671363206678443,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2313573569059372,
      "num_tokens": 43452184.0,
      "step": 18960
    },
    {
      "entropy": 5.27362699508667,
      "epoch": 1.8218059558117194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046711847820817215,
      "loss": 4.7826,
      "mean_token_accuracy": 0.2417294129729271,
      "num_tokens": 43463361.0,
      "step": 18965
    },
    {
      "entropy": 5.34563422203064,
      "epoch": 1.8222862632084533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004671006312880794,
      "loss": 4.8206,
      "mean_token_accuracy": 0.23330789804458618,
      "num_tokens": 43474802.0,
      "step": 18970
    },
    {
      "entropy": 5.311386680603027,
      "epoch": 1.8227665706051872,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004670827799079805,
      "loss": 4.7706,
      "mean_token_accuracy": 0.23729595243930818,
      "num_tokens": 43486962.0,
      "step": 18975
    },
    {
      "entropy": 5.307014513015747,
      "epoch": 1.823246878001921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046706492406828966,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2373532086610794,
      "num_tokens": 43498761.0,
      "step": 18980
    },
    {
      "entropy": 5.323840045928955,
      "epoch": 1.8237271853986552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004670470637694217,
      "loss": 4.865,
      "mean_token_accuracy": 0.23514655828475953,
      "num_tokens": 43511294.0,
      "step": 18985
    },
    {
      "entropy": 5.29512996673584,
      "epoch": 1.824207492795389,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004670291990117912,
      "loss": 4.7776,
      "mean_token_accuracy": 0.24325323700904847,
      "num_tokens": 43522858.0,
      "step": 18990
    },
    {
      "entropy": 5.333927392959595,
      "epoch": 1.824687800192123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004670113297958128,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2352076292037964,
      "num_tokens": 43533453.0,
      "step": 18995
    },
    {
      "entropy": 5.322010278701782,
      "epoch": 1.8251681075888568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046699345612190155,
      "loss": 4.7517,
      "mean_token_accuracy": 0.24377527087926865,
      "num_tokens": 43543889.0,
      "step": 19000
    },
    {
      "entropy": 5.263615846633911,
      "epoch": 1.825648414985591,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046697557799047233,
      "loss": 4.7491,
      "mean_token_accuracy": 0.24131178855895996,
      "num_tokens": 43554815.0,
      "step": 19005
    },
    {
      "entropy": 5.222885513305664,
      "epoch": 1.8261287223823248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004669576954019403,
      "loss": 4.6952,
      "mean_token_accuracy": 0.24694344401359558,
      "num_tokens": 43565258.0,
      "step": 19010
    },
    {
      "entropy": 5.305173873901367,
      "epoch": 1.8266090297790587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004669398083567205,
      "loss": 4.822,
      "mean_token_accuracy": 0.23649607747793197,
      "num_tokens": 43576710.0,
      "step": 19015
    },
    {
      "entropy": 5.31352653503418,
      "epoch": 1.8270893371757926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004669219168552284,
      "loss": 4.803,
      "mean_token_accuracy": 0.23646220862865447,
      "num_tokens": 43587590.0,
      "step": 19020
    },
    {
      "entropy": 5.291584539413452,
      "epoch": 1.8275696445725265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046690402089787916,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23813187181949616,
      "num_tokens": 43599675.0,
      "step": 19025
    },
    {
      "entropy": 5.241330146789551,
      "epoch": 1.8280499519692603,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004668861204850884,
      "loss": 4.686,
      "mean_token_accuracy": 0.2485849991440773,
      "num_tokens": 43612369.0,
      "step": 19030
    },
    {
      "entropy": 5.282721996307373,
      "epoch": 1.8285302593659942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046686821561727176,
      "loss": 4.737,
      "mean_token_accuracy": 0.24603454917669296,
      "num_tokens": 43624807.0,
      "step": 19035
    },
    {
      "entropy": 5.279276609420776,
      "epoch": 1.829010566762728,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004668503062948449,
      "loss": 4.8203,
      "mean_token_accuracy": 0.23661645501852036,
      "num_tokens": 43635389.0,
      "step": 19040
    },
    {
      "entropy": 5.285945796966553,
      "epoch": 1.829490874159462,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004668323925182236,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23462207317352296,
      "num_tokens": 43646807.0,
      "step": 19045
    },
    {
      "entropy": 5.316197204589844,
      "epoch": 1.8299711815561959,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046681447428782377,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2304614156484604,
      "num_tokens": 43657910.0,
      "step": 19050
    },
    {
      "entropy": 5.292719459533691,
      "epoch": 1.8304514889529298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004667965516040613,
      "loss": 4.7854,
      "mean_token_accuracy": 0.24015338569879532,
      "num_tokens": 43669376.0,
      "step": 19055
    },
    {
      "entropy": 5.2729720115661625,
      "epoch": 1.8309317963496636,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004667786244673526,
      "loss": 4.7609,
      "mean_token_accuracy": 0.23753412067890167,
      "num_tokens": 43681702.0,
      "step": 19060
    },
    {
      "entropy": 5.337646865844727,
      "epoch": 1.8314121037463977,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046676069287811365,
      "loss": 4.8683,
      "mean_token_accuracy": 0.23899340033531188,
      "num_tokens": 43694795.0,
      "step": 19065
    },
    {
      "entropy": 5.387523603439331,
      "epoch": 1.8318924111431316,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004667427568367607,
      "loss": 4.9009,
      "mean_token_accuracy": 0.2319638431072235,
      "num_tokens": 43707197.0,
      "step": 19070
    },
    {
      "entropy": 5.345732307434082,
      "epoch": 1.8323727185398655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046672481634371047,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22934607565402984,
      "num_tokens": 43719066.0,
      "step": 19075
    },
    {
      "entropy": 5.29883975982666,
      "epoch": 1.8328530259365994,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046670687139937925,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23997736424207688,
      "num_tokens": 43731080.0,
      "step": 19080
    },
    {
      "entropy": 5.347378921508789,
      "epoch": 1.8333333333333335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004666889220041837,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23471244871616365,
      "num_tokens": 43743048.0,
      "step": 19085
    },
    {
      "entropy": 5.305838871002197,
      "epoch": 1.8338136407300674,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046667096815854056,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2292414277791977,
      "num_tokens": 43754701.0,
      "step": 19090
    },
    {
      "entropy": 5.320587205886841,
      "epoch": 1.8342939481268012,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004666530098628667,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23636903017759323,
      "num_tokens": 43765696.0,
      "step": 19095
    },
    {
      "entropy": 5.368696069717407,
      "epoch": 1.8347742555235351,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004666350471175791,
      "loss": 4.9232,
      "mean_token_accuracy": 0.22831243723630906,
      "num_tokens": 43777162.0,
      "step": 19100
    },
    {
      "entropy": 5.324111557006836,
      "epoch": 1.835254562920269,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004666170799230947,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23241520226001738,
      "num_tokens": 43789340.0,
      "step": 19105
    },
    {
      "entropy": 5.248450374603271,
      "epoch": 1.8357348703170029,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004665991082798307,
      "loss": 4.7168,
      "mean_token_accuracy": 0.2514473408460617,
      "num_tokens": 43800098.0,
      "step": 19110
    },
    {
      "entropy": 5.2936155796051025,
      "epoch": 1.8362151777137368,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004665811321882043,
      "loss": 4.7908,
      "mean_token_accuracy": 0.2401132494211197,
      "num_tokens": 43811403.0,
      "step": 19115
    },
    {
      "entropy": 5.347124242782593,
      "epoch": 1.8366954851104706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046656315164863297,
      "loss": 4.8239,
      "mean_token_accuracy": 0.24206000566482544,
      "num_tokens": 43821946.0,
      "step": 19120
    },
    {
      "entropy": 5.217678022384644,
      "epoch": 1.8371757925072045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046654516666153403,
      "loss": 4.7472,
      "mean_token_accuracy": 0.23907105177640914,
      "num_tokens": 43832638.0,
      "step": 19125
    },
    {
      "entropy": 5.265948390960693,
      "epoch": 1.8376560999039384,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004665271772273251,
      "loss": 4.7415,
      "mean_token_accuracy": 0.24128571450710296,
      "num_tokens": 43843616.0,
      "step": 19130
    },
    {
      "entropy": 5.345883131027222,
      "epoch": 1.8381364073006723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004665091833464239,
      "loss": 4.762,
      "mean_token_accuracy": 0.23629057109355928,
      "num_tokens": 43854680.0,
      "step": 19135
    },
    {
      "entropy": 5.293078470230102,
      "epoch": 1.8386167146974062,
      "grad_norm": 1.0,
      "learning_rate": 0.00046649118501924805,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23989685922861098,
      "num_tokens": 43866683.0,
      "step": 19140
    },
    {
      "entropy": 5.335873985290528,
      "epoch": 1.8390970220941403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004664731822462154,
      "loss": 4.8508,
      "mean_token_accuracy": 0.2338727369904518,
      "num_tokens": 43878080.0,
      "step": 19145
    },
    {
      "entropy": 5.329166412353516,
      "epoch": 1.8395773294908742,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046645517502774415,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23923422545194625,
      "num_tokens": 43888633.0,
      "step": 19150
    },
    {
      "entropy": 5.384761810302734,
      "epoch": 1.840057636887608,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046643716336425224,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23723267763853073,
      "num_tokens": 43899678.0,
      "step": 19155
    },
    {
      "entropy": 5.278373908996582,
      "epoch": 1.8405379442843421,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004664191472561578,
      "loss": 4.7962,
      "mean_token_accuracy": 0.24250260293483733,
      "num_tokens": 43910744.0,
      "step": 19160
    },
    {
      "entropy": 5.322403621673584,
      "epoch": 1.841018251681076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004664011267038792,
      "loss": 4.908,
      "mean_token_accuracy": 0.2349931165575981,
      "num_tokens": 43922467.0,
      "step": 19165
    },
    {
      "entropy": 5.290589475631714,
      "epoch": 1.84149855907781,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046638310170783476,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2419314593076706,
      "num_tokens": 43934919.0,
      "step": 19170
    },
    {
      "entropy": 5.298811244964599,
      "epoch": 1.8419788664745438,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000466365072268443,
      "loss": 4.8198,
      "mean_token_accuracy": 0.23383379429578782,
      "num_tokens": 43947737.0,
      "step": 19175
    },
    {
      "entropy": 5.25124683380127,
      "epoch": 1.8424591738712777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004663470383861225,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24163396656513214,
      "num_tokens": 43959656.0,
      "step": 19180
    },
    {
      "entropy": 5.2981404781341555,
      "epoch": 1.8429394812680115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466329000061292,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23979755938053132,
      "num_tokens": 43970732.0,
      "step": 19185
    },
    {
      "entropy": 5.257415151596069,
      "epoch": 1.8434197886647454,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004663109572943702,
      "loss": 4.779,
      "mean_token_accuracy": 0.23165252208709716,
      "num_tokens": 43982369.0,
      "step": 19190
    },
    {
      "entropy": 5.3806356430053714,
      "epoch": 1.8439000960614793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000466292910085776,
      "loss": 4.829,
      "mean_token_accuracy": 0.23587993234395982,
      "num_tokens": 43993751.0,
      "step": 19195
    },
    {
      "entropy": 5.277002668380737,
      "epoch": 1.8443804034582132,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046627485843592854,
      "loss": 4.7856,
      "mean_token_accuracy": 0.2360158696770668,
      "num_tokens": 44005021.0,
      "step": 19200
    },
    {
      "entropy": 5.22818284034729,
      "epoch": 1.844860710854947,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046625680234524674,
      "loss": 4.7621,
      "mean_token_accuracy": 0.23653512746095656,
      "num_tokens": 44015571.0,
      "step": 19205
    },
    {
      "entropy": 5.315073537826538,
      "epoch": 1.845341018251681,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046623874181414993,
      "loss": 4.8316,
      "mean_token_accuracy": 0.2345678001642227,
      "num_tokens": 44026363.0,
      "step": 19210
    },
    {
      "entropy": 5.299896764755249,
      "epoch": 1.8458213256484148,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004662206768430574,
      "loss": 4.7286,
      "mean_token_accuracy": 0.24789835512638092,
      "num_tokens": 44038042.0,
      "step": 19215
    },
    {
      "entropy": 5.3589376449584964,
      "epoch": 1.846301633045149,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004662026074323886,
      "loss": 4.8241,
      "mean_token_accuracy": 0.23154074996709822,
      "num_tokens": 44049086.0,
      "step": 19220
    },
    {
      "entropy": 5.298569965362549,
      "epoch": 1.8467819404418828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046618453358256303,
      "loss": 4.8253,
      "mean_token_accuracy": 0.24008380025625228,
      "num_tokens": 44060816.0,
      "step": 19225
    },
    {
      "entropy": 5.227885675430298,
      "epoch": 1.8472622478386167,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046616645529400026,
      "loss": 4.7003,
      "mean_token_accuracy": 0.23722879141569136,
      "num_tokens": 44071877.0,
      "step": 19230
    },
    {
      "entropy": 5.340013885498047,
      "epoch": 1.8477425552353506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004661483725671201,
      "loss": 4.8089,
      "mean_token_accuracy": 0.23418303877115249,
      "num_tokens": 44082618.0,
      "step": 19235
    },
    {
      "entropy": 5.29099931716919,
      "epoch": 1.8482228626320847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046613028540234226,
      "loss": 4.7583,
      "mean_token_accuracy": 0.24212500154972078,
      "num_tokens": 44093411.0,
      "step": 19240
    },
    {
      "entropy": 5.22143588066101,
      "epoch": 1.8487031700288186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004661121938000867,
      "loss": 4.6787,
      "mean_token_accuracy": 0.24540379494428635,
      "num_tokens": 44103595.0,
      "step": 19245
    },
    {
      "entropy": 5.237143516540527,
      "epoch": 1.8491834774255524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004660940977607736,
      "loss": 4.783,
      "mean_token_accuracy": 0.2458460286259651,
      "num_tokens": 44114740.0,
      "step": 19250
    },
    {
      "entropy": 5.225882339477539,
      "epoch": 1.8496637848222863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046607599728482285,
      "loss": 4.7299,
      "mean_token_accuracy": 0.2435559317469597,
      "num_tokens": 44125415.0,
      "step": 19255
    },
    {
      "entropy": 5.26289176940918,
      "epoch": 1.8501440922190202,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046605789237265496,
      "loss": 4.782,
      "mean_token_accuracy": 0.2383887931704521,
      "num_tokens": 44137399.0,
      "step": 19260
    },
    {
      "entropy": 5.266622400283813,
      "epoch": 1.850624399615754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046603978302469,
      "loss": 4.7201,
      "mean_token_accuracy": 0.23832615315914155,
      "num_tokens": 44148354.0,
      "step": 19265
    },
    {
      "entropy": 5.25102014541626,
      "epoch": 1.851104707012488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004660216692413486,
      "loss": 4.7608,
      "mean_token_accuracy": 0.2413267284631729,
      "num_tokens": 44160496.0,
      "step": 19270
    },
    {
      "entropy": 5.2868000030517575,
      "epoch": 1.8515850144092219,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004660035510230513,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23619239181280136,
      "num_tokens": 44171832.0,
      "step": 19275
    },
    {
      "entropy": 5.338691759109497,
      "epoch": 1.8520653218059557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004659854283702186,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23887517005205156,
      "num_tokens": 44182850.0,
      "step": 19280
    },
    {
      "entropy": 5.300503158569336,
      "epoch": 1.8525456292026896,
      "grad_norm": 1.0,
      "learning_rate": 0.0004659673012832715,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23692278563976288,
      "num_tokens": 44194343.0,
      "step": 19285
    },
    {
      "entropy": 5.360225343704224,
      "epoch": 1.8530259365994235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004659491697626306,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2250534102320671,
      "num_tokens": 44206294.0,
      "step": 19290
    },
    {
      "entropy": 5.309760999679566,
      "epoch": 1.8535062439961574,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046593103380871705,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23867851197719575,
      "num_tokens": 44218576.0,
      "step": 19295
    },
    {
      "entropy": 5.278888988494873,
      "epoch": 1.8539865513928915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046591289342195184,
      "loss": 4.8049,
      "mean_token_accuracy": 0.23508245795965194,
      "num_tokens": 44230480.0,
      "step": 19300
    },
    {
      "entropy": 5.274936676025391,
      "epoch": 1.8544668587896254,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004658947486027562,
      "loss": 4.7108,
      "mean_token_accuracy": 0.24444971084594727,
      "num_tokens": 44240907.0,
      "step": 19305
    },
    {
      "entropy": 5.337499189376831,
      "epoch": 1.8549471661863592,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046587659935155124,
      "loss": 4.8946,
      "mean_token_accuracy": 0.2266909435391426,
      "num_tokens": 44251469.0,
      "step": 19310
    },
    {
      "entropy": 5.3846203804016115,
      "epoch": 1.8554274735830933,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046585844566875845,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23318729251623155,
      "num_tokens": 44263875.0,
      "step": 19315
    },
    {
      "entropy": 5.359569358825683,
      "epoch": 1.8559077809798272,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004658402875547993,
      "loss": 4.8776,
      "mean_token_accuracy": 0.232925483584404,
      "num_tokens": 44276163.0,
      "step": 19320
    },
    {
      "entropy": 5.2775561809539795,
      "epoch": 1.856388088376561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046582212501009533,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23144183307886124,
      "num_tokens": 44287985.0,
      "step": 19325
    },
    {
      "entropy": 5.323008966445923,
      "epoch": 1.856868395773295,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046580395803506825,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2346382811665535,
      "num_tokens": 44299360.0,
      "step": 19330
    },
    {
      "entropy": 5.332439994812011,
      "epoch": 1.8573487031700289,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004657857866301399,
      "loss": 4.7764,
      "mean_token_accuracy": 0.24931450337171554,
      "num_tokens": 44310459.0,
      "step": 19335
    },
    {
      "entropy": 5.42651515007019,
      "epoch": 1.8578290105667628,
      "grad_norm": 2.921875,
      "learning_rate": 0.00046576761079573204,
      "loss": 4.9046,
      "mean_token_accuracy": 0.22902050465345383,
      "num_tokens": 44321832.0,
      "step": 19340
    },
    {
      "entropy": 5.325369548797608,
      "epoch": 1.8583093179634966,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004657494305322667,
      "loss": 4.8869,
      "mean_token_accuracy": 0.2290043279528618,
      "num_tokens": 44332829.0,
      "step": 19345
    },
    {
      "entropy": 5.319242191314697,
      "epoch": 1.8587896253602305,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004657312458401661,
      "loss": 4.7722,
      "mean_token_accuracy": 0.23643431663513184,
      "num_tokens": 44344321.0,
      "step": 19350
    },
    {
      "entropy": 5.440406751632691,
      "epoch": 1.8592699327569644,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004657130567198522,
      "loss": 4.8602,
      "mean_token_accuracy": 0.23364004045724868,
      "num_tokens": 44357183.0,
      "step": 19355
    },
    {
      "entropy": 5.372306776046753,
      "epoch": 1.8597502401536983,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046569486317174746,
      "loss": 4.8797,
      "mean_token_accuracy": 0.23053097426891328,
      "num_tokens": 44367118.0,
      "step": 19360
    },
    {
      "entropy": 5.315132141113281,
      "epoch": 1.8602305475504322,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004656766651962742,
      "loss": 4.8711,
      "mean_token_accuracy": 0.23215980231761932,
      "num_tokens": 44377699.0,
      "step": 19365
    },
    {
      "entropy": 5.3213825702667235,
      "epoch": 1.860710854947166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004656584627938551,
      "loss": 4.8166,
      "mean_token_accuracy": 0.22702330201864243,
      "num_tokens": 44389457.0,
      "step": 19370
    },
    {
      "entropy": 5.2871137142181395,
      "epoch": 1.8611911623439001,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046564025596491254,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24169861525297165,
      "num_tokens": 44400140.0,
      "step": 19375
    },
    {
      "entropy": 5.264038133621216,
      "epoch": 1.861671469740634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004656220447098693,
      "loss": 4.7451,
      "mean_token_accuracy": 0.2477394610643387,
      "num_tokens": 44411088.0,
      "step": 19380
    },
    {
      "entropy": 5.306335687637329,
      "epoch": 1.862151777137368,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004656038290291483,
      "loss": 4.7515,
      "mean_token_accuracy": 0.23716589510440828,
      "num_tokens": 44422515.0,
      "step": 19385
    },
    {
      "entropy": 5.232414960861206,
      "epoch": 1.8626320845341018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004655856089231723,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23960181325674057,
      "num_tokens": 44433711.0,
      "step": 19390
    },
    {
      "entropy": 5.305027961730957,
      "epoch": 1.8631123919308359,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004655673843923644,
      "loss": 4.8524,
      "mean_token_accuracy": 0.23571337312459945,
      "num_tokens": 44446157.0,
      "step": 19395
    },
    {
      "entropy": 5.450185060501099,
      "epoch": 1.8635926993275698,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004655491554371477,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22616788893938064,
      "num_tokens": 44457506.0,
      "step": 19400
    },
    {
      "entropy": 5.354125022888184,
      "epoch": 1.8640730067243036,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046553092205794543,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22712238878011703,
      "num_tokens": 44469572.0,
      "step": 19405
    },
    {
      "entropy": 5.305078172683716,
      "epoch": 1.8645533141210375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046551268425518096,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23454457372426987,
      "num_tokens": 44481932.0,
      "step": 19410
    },
    {
      "entropy": 5.284637451171875,
      "epoch": 1.8650336215177714,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004654944420292776,
      "loss": 4.7547,
      "mean_token_accuracy": 0.24073764234781264,
      "num_tokens": 44493081.0,
      "step": 19415
    },
    {
      "entropy": 5.324727344512939,
      "epoch": 1.8655139289145053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000465476195380659,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23778435438871384,
      "num_tokens": 44503919.0,
      "step": 19420
    },
    {
      "entropy": 5.3304877281188965,
      "epoch": 1.8659942363112392,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004654579443097487,
      "loss": 4.7964,
      "mean_token_accuracy": 0.2395562171936035,
      "num_tokens": 44515493.0,
      "step": 19425
    },
    {
      "entropy": 5.385239601135254,
      "epoch": 1.866474543707973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004654396888169705,
      "loss": 4.84,
      "mean_token_accuracy": 0.2301520511507988,
      "num_tokens": 44527432.0,
      "step": 19430
    },
    {
      "entropy": 5.271756553649903,
      "epoch": 1.866954851104707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046542142890274816,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23308956176042556,
      "num_tokens": 44537634.0,
      "step": 19435
    },
    {
      "entropy": 5.2604231357574465,
      "epoch": 1.8674351585014408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004654031645675057,
      "loss": 4.8436,
      "mean_token_accuracy": 0.23136336207389832,
      "num_tokens": 44549856.0,
      "step": 19440
    },
    {
      "entropy": 5.238847303390503,
      "epoch": 1.8679154658981747,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004653848958116672,
      "loss": 4.7595,
      "mean_token_accuracy": 0.2362754002213478,
      "num_tokens": 44563214.0,
      "step": 19445
    },
    {
      "entropy": 5.277080726623535,
      "epoch": 1.8683957732949086,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046536662263565667,
      "loss": 4.7783,
      "mean_token_accuracy": 0.23123113214969634,
      "num_tokens": 44572982.0,
      "step": 19450
    },
    {
      "entropy": 5.2860595226287845,
      "epoch": 1.8688760806916427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653483450398985,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24225889444351195,
      "num_tokens": 44583664.0,
      "step": 19455
    },
    {
      "entropy": 5.25996413230896,
      "epoch": 1.8693563880883766,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046533006302481694,
      "loss": 4.6638,
      "mean_token_accuracy": 0.24769596606492997,
      "num_tokens": 44594987.0,
      "step": 19460
    },
    {
      "entropy": 5.297488260269165,
      "epoch": 1.8698366954851104,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004653117765908365,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23929226994514466,
      "num_tokens": 44607123.0,
      "step": 19465
    },
    {
      "entropy": 5.276364135742187,
      "epoch": 1.8703170028818443,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004652934857383816,
      "loss": 4.7464,
      "mean_token_accuracy": 0.2350808396935463,
      "num_tokens": 44618736.0,
      "step": 19470
    },
    {
      "entropy": 5.2509440898895265,
      "epoch": 1.8707973102785784,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004652751904678772,
      "loss": 4.7236,
      "mean_token_accuracy": 0.25003165006637573,
      "num_tokens": 44630681.0,
      "step": 19475
    },
    {
      "entropy": 5.291375398635864,
      "epoch": 1.8712776176753123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046525689077974775,
      "loss": 4.7558,
      "mean_token_accuracy": 0.24440109133720397,
      "num_tokens": 44641641.0,
      "step": 19480
    },
    {
      "entropy": 5.257997226715088,
      "epoch": 1.8717579250720462,
      "grad_norm": 1.0,
      "learning_rate": 0.00046523858667441834,
      "loss": 4.7586,
      "mean_token_accuracy": 0.2400527849793434,
      "num_tokens": 44652489.0,
      "step": 19485
    },
    {
      "entropy": 5.324246501922607,
      "epoch": 1.87223823246878,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004652202781523138,
      "loss": 4.8324,
      "mean_token_accuracy": 0.23313162922859193,
      "num_tokens": 44665659.0,
      "step": 19490
    },
    {
      "entropy": 5.307474994659424,
      "epoch": 1.872718539865514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004652019652138592,
      "loss": 4.7461,
      "mean_token_accuracy": 0.2459231972694397,
      "num_tokens": 44677300.0,
      "step": 19495
    },
    {
      "entropy": 5.295508575439453,
      "epoch": 1.8731988472622478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004651836478594798,
      "loss": 4.8046,
      "mean_token_accuracy": 0.2332077667117119,
      "num_tokens": 44690084.0,
      "step": 19500
    },
    {
      "entropy": 5.288603639602661,
      "epoch": 1.8736791546589817,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651653260896008,
      "loss": 4.8051,
      "mean_token_accuracy": 0.2422592431306839,
      "num_tokens": 44701208.0,
      "step": 19505
    },
    {
      "entropy": 5.398444986343383,
      "epoch": 1.8741594620557156,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046514699990464763,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22635384052991867,
      "num_tokens": 44713546.0,
      "step": 19510
    },
    {
      "entropy": 5.50647120475769,
      "epoch": 1.8746397694524495,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004651286693050458,
      "loss": 5.0157,
      "mean_token_accuracy": 0.2163140058517456,
      "num_tokens": 44724802.0,
      "step": 19515
    },
    {
      "entropy": 5.313358402252197,
      "epoch": 1.8751200768491834,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004651103342912207,
      "loss": 4.7991,
      "mean_token_accuracy": 0.24109958857297897,
      "num_tokens": 44736625.0,
      "step": 19520
    },
    {
      "entropy": 5.280191802978516,
      "epoch": 1.8756003842459172,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046509199486359824,
      "loss": 4.753,
      "mean_token_accuracy": 0.22784009128808974,
      "num_tokens": 44750451.0,
      "step": 19525
    },
    {
      "entropy": 5.245674562454224,
      "epoch": 1.8760806916426513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046507365102260403,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23763348013162613,
      "num_tokens": 44762043.0,
      "step": 19530
    },
    {
      "entropy": 5.328298759460449,
      "epoch": 1.8765609990393852,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046505530276866417,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23458503186702728,
      "num_tokens": 44774329.0,
      "step": 19535
    },
    {
      "entropy": 5.366810607910156,
      "epoch": 1.877041306436119,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046503695010220443,
      "loss": 4.837,
      "mean_token_accuracy": 0.2376820057630539,
      "num_tokens": 44786352.0,
      "step": 19540
    },
    {
      "entropy": 5.228633451461792,
      "epoch": 1.877521613832853,
      "grad_norm": 1.078125,
      "learning_rate": 0.000465018593023651,
      "loss": 4.6641,
      "mean_token_accuracy": 0.24474628120660782,
      "num_tokens": 44796829.0,
      "step": 19545
    },
    {
      "entropy": 5.34697904586792,
      "epoch": 1.878001921229587,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004650002315334302,
      "loss": 4.8567,
      "mean_token_accuracy": 0.2327722728252411,
      "num_tokens": 44808900.0,
      "step": 19550
    },
    {
      "entropy": 5.252507877349854,
      "epoch": 1.878482228626321,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000464981865631968,
      "loss": 4.7998,
      "mean_token_accuracy": 0.24095992594957352,
      "num_tokens": 44820638.0,
      "step": 19555
    },
    {
      "entropy": 5.27784481048584,
      "epoch": 1.8789625360230549,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004649634953196912,
      "loss": 4.7732,
      "mean_token_accuracy": 0.2394363611936569,
      "num_tokens": 44833370.0,
      "step": 19560
    },
    {
      "entropy": 5.30045223236084,
      "epoch": 1.8794428434197887,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046494512059702605,
      "loss": 4.8189,
      "mean_token_accuracy": 0.2402859479188919,
      "num_tokens": 44844291.0,
      "step": 19565
    },
    {
      "entropy": 5.2842058658599855,
      "epoch": 1.8799231508165226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004649267414643992,
      "loss": 4.7456,
      "mean_token_accuracy": 0.24366634637117385,
      "num_tokens": 44855303.0,
      "step": 19570
    },
    {
      "entropy": 5.23927936553955,
      "epoch": 1.8804034582132565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004649083579222374,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24013349562883377,
      "num_tokens": 44866593.0,
      "step": 19575
    },
    {
      "entropy": 5.27313346862793,
      "epoch": 1.8808837656099904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046488996997096744,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23745992481708528,
      "num_tokens": 44879124.0,
      "step": 19580
    },
    {
      "entropy": 5.263116693496704,
      "epoch": 1.8813640730067243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004648715776110162,
      "loss": 4.758,
      "mean_token_accuracy": 0.24935708045959473,
      "num_tokens": 44889942.0,
      "step": 19585
    },
    {
      "entropy": 5.2413952350616455,
      "epoch": 1.8818443804034581,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004648531808428108,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2464038461446762,
      "num_tokens": 44901002.0,
      "step": 19590
    },
    {
      "entropy": 5.303162956237793,
      "epoch": 1.882324687800192,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004648347796667782,
      "loss": 4.873,
      "mean_token_accuracy": 0.22743603885173796,
      "num_tokens": 44913495.0,
      "step": 19595
    },
    {
      "entropy": 5.314231252670288,
      "epoch": 1.882804995196926,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004648163740833458,
      "loss": 4.7737,
      "mean_token_accuracy": 0.23102711141109467,
      "num_tokens": 44924624.0,
      "step": 19600
    },
    {
      "entropy": 5.318283176422119,
      "epoch": 1.8832853025936598,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046479796409294076,
      "loss": 4.8445,
      "mean_token_accuracy": 0.2385002300143242,
      "num_tokens": 44935539.0,
      "step": 19605
    },
    {
      "entropy": 5.265105533599853,
      "epoch": 1.8837656099903939,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004647795496959907,
      "loss": 4.7731,
      "mean_token_accuracy": 0.2396216318011284,
      "num_tokens": 44945979.0,
      "step": 19610
    },
    {
      "entropy": 5.3072856903076175,
      "epoch": 1.8842459173871278,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046476113089292286,
      "loss": 4.8259,
      "mean_token_accuracy": 0.22717552334070207,
      "num_tokens": 44958294.0,
      "step": 19615
    },
    {
      "entropy": 5.381676387786865,
      "epoch": 1.8847262247838616,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004647427076841651,
      "loss": 4.905,
      "mean_token_accuracy": 0.23304453045129775,
      "num_tokens": 44970261.0,
      "step": 19620
    },
    {
      "entropy": 5.314013671875,
      "epoch": 1.8852065321805955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046472428007014515,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23451007753610612,
      "num_tokens": 44981515.0,
      "step": 19625
    },
    {
      "entropy": 5.26247730255127,
      "epoch": 1.8856868395773296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004647058480512907,
      "loss": 4.7118,
      "mean_token_accuracy": 0.24289604127407075,
      "num_tokens": 44992473.0,
      "step": 19630
    },
    {
      "entropy": 5.370556879043579,
      "epoch": 1.8861671469740635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046468741162802987,
      "loss": 4.8377,
      "mean_token_accuracy": 0.23767558932304383,
      "num_tokens": 45004264.0,
      "step": 19635
    },
    {
      "entropy": 5.285222148895263,
      "epoch": 1.8866474543707974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004646689708007905,
      "loss": 4.7267,
      "mean_token_accuracy": 0.2475736364722252,
      "num_tokens": 45015601.0,
      "step": 19640
    },
    {
      "entropy": 5.357336044311523,
      "epoch": 1.8871277617675313,
      "grad_norm": 1.0,
      "learning_rate": 0.00046465052557000087,
      "loss": 4.888,
      "mean_token_accuracy": 0.23986945748329164,
      "num_tokens": 45028044.0,
      "step": 19645
    },
    {
      "entropy": 5.367936372756958,
      "epoch": 1.8876080691642652,
      "grad_norm": 1.78125,
      "learning_rate": 0.00046463207593608916,
      "loss": 4.8728,
      "mean_token_accuracy": 0.23946283459663392,
      "num_tokens": 45040144.0,
      "step": 19650
    },
    {
      "entropy": 5.416708087921142,
      "epoch": 1.888088376560999,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004646136218994837,
      "loss": 4.9204,
      "mean_token_accuracy": 0.22501070350408553,
      "num_tokens": 45051844.0,
      "step": 19655
    },
    {
      "entropy": 5.3446502685546875,
      "epoch": 1.888568683957733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046459516346061304,
      "loss": 4.8651,
      "mean_token_accuracy": 0.23263224065303803,
      "num_tokens": 45062684.0,
      "step": 19660
    },
    {
      "entropy": 5.289200925827027,
      "epoch": 1.8890489913544668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046457670061990564,
      "loss": 4.8037,
      "mean_token_accuracy": 0.23918602168560027,
      "num_tokens": 45073327.0,
      "step": 19665
    },
    {
      "entropy": 5.319364786148071,
      "epoch": 1.8895292987512007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046455823337779024,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23526244908571242,
      "num_tokens": 45085309.0,
      "step": 19670
    },
    {
      "entropy": 5.380319881439209,
      "epoch": 1.8900096061479346,
      "grad_norm": 1.0,
      "learning_rate": 0.0004645397617346954,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23334037214517594,
      "num_tokens": 45095934.0,
      "step": 19675
    },
    {
      "entropy": 5.336351442337036,
      "epoch": 1.8904899135446684,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645212856910502,
      "loss": 4.796,
      "mean_token_accuracy": 0.24278501123189927,
      "num_tokens": 45106220.0,
      "step": 19680
    },
    {
      "entropy": 5.2774876117706295,
      "epoch": 1.8909702209414025,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004645028052472835,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23957519680261613,
      "num_tokens": 45117861.0,
      "step": 19685
    },
    {
      "entropy": 5.265826892852783,
      "epoch": 1.8914505283381364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046448432040382444,
      "loss": 4.7599,
      "mean_token_accuracy": 0.24781273305416107,
      "num_tokens": 45129248.0,
      "step": 19690
    },
    {
      "entropy": 5.364680624008178,
      "epoch": 1.8919308357348703,
      "grad_norm": 1.078125,
      "learning_rate": 0.000464465831161102,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22990724444389343,
      "num_tokens": 45140610.0,
      "step": 19695
    },
    {
      "entropy": 5.284329128265381,
      "epoch": 1.8924111431316042,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004644473375195456,
      "loss": 4.7489,
      "mean_token_accuracy": 0.24478370398283006,
      "num_tokens": 45150697.0,
      "step": 19700
    },
    {
      "entropy": 5.328828620910644,
      "epoch": 1.8928914505283383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046442883947958466,
      "loss": 4.8114,
      "mean_token_accuracy": 0.23548468202352524,
      "num_tokens": 45162627.0,
      "step": 19705
    },
    {
      "entropy": 5.216463088989258,
      "epoch": 1.8933717579250722,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046441033704164845,
      "loss": 4.6338,
      "mean_token_accuracy": 0.25243043899536133,
      "num_tokens": 45173030.0,
      "step": 19710
    },
    {
      "entropy": 5.339263486862182,
      "epoch": 1.893852065321806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004643918302061666,
      "loss": 4.8148,
      "mean_token_accuracy": 0.23441403359174728,
      "num_tokens": 45183588.0,
      "step": 19715
    },
    {
      "entropy": 5.367744779586792,
      "epoch": 1.89433237271854,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004643733189735689,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23680482804775238,
      "num_tokens": 45195289.0,
      "step": 19720
    },
    {
      "entropy": 5.336230945587158,
      "epoch": 1.8948126801152738,
      "grad_norm": 1.046875,
      "learning_rate": 0.000464354803344285,
      "loss": 4.7798,
      "mean_token_accuracy": 0.23905843794345855,
      "num_tokens": 45207001.0,
      "step": 19725
    },
    {
      "entropy": 5.267214679718018,
      "epoch": 1.8952929875120077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046433628331874496,
      "loss": 4.7959,
      "mean_token_accuracy": 0.2403460755944252,
      "num_tokens": 45217447.0,
      "step": 19730
    },
    {
      "entropy": 5.300955057144165,
      "epoch": 1.8957732949087416,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004643177588973785,
      "loss": 4.776,
      "mean_token_accuracy": 0.23652008175849915,
      "num_tokens": 45227588.0,
      "step": 19735
    },
    {
      "entropy": 5.332609701156616,
      "epoch": 1.8962536023054755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004642992300806159,
      "loss": 4.7852,
      "mean_token_accuracy": 0.24063927829265594,
      "num_tokens": 45239969.0,
      "step": 19740
    },
    {
      "entropy": 5.324943923950196,
      "epoch": 1.8967339097022093,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004642806968688873,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23651886880397796,
      "num_tokens": 45250877.0,
      "step": 19745
    },
    {
      "entropy": 5.346182632446289,
      "epoch": 1.8972142170989432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046426215926262295,
      "loss": 4.8459,
      "mean_token_accuracy": 0.23952964544296265,
      "num_tokens": 45262846.0,
      "step": 19750
    },
    {
      "entropy": 5.255933856964111,
      "epoch": 1.897694524495677,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004642436172622532,
      "loss": 4.7882,
      "mean_token_accuracy": 0.23984862267971038,
      "num_tokens": 45274198.0,
      "step": 19755
    },
    {
      "entropy": 5.21840534210205,
      "epoch": 1.898174831892411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004642250708682086,
      "loss": 4.7844,
      "mean_token_accuracy": 0.2443023666739464,
      "num_tokens": 45285803.0,
      "step": 19760
    },
    {
      "entropy": 5.32810263633728,
      "epoch": 1.898655139289145,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046420652008091984,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23166433423757554,
      "num_tokens": 45298242.0,
      "step": 19765
    },
    {
      "entropy": 5.287969064712525,
      "epoch": 1.899135446685879,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004641879649008174,
      "loss": 4.7799,
      "mean_token_accuracy": 0.24073860049247742,
      "num_tokens": 45309567.0,
      "step": 19770
    },
    {
      "entropy": 5.367614650726319,
      "epoch": 1.8996157540826129,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004641694053283323,
      "loss": 4.8748,
      "mean_token_accuracy": 0.23722658008337022,
      "num_tokens": 45321693.0,
      "step": 19775
    },
    {
      "entropy": 5.385435104370117,
      "epoch": 1.9000960614793467,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046415084136389525,
      "loss": 4.8421,
      "mean_token_accuracy": 0.2331385627388954,
      "num_tokens": 45333094.0,
      "step": 19780
    },
    {
      "entropy": 5.306300783157349,
      "epoch": 1.9005763688760808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004641322730079374,
      "loss": 4.7824,
      "mean_token_accuracy": 0.24160833060741424,
      "num_tokens": 45344549.0,
      "step": 19785
    },
    {
      "entropy": 5.343528175354004,
      "epoch": 1.9010566762728147,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004641137002608897,
      "loss": 4.849,
      "mean_token_accuracy": 0.22804128080606462,
      "num_tokens": 45356433.0,
      "step": 19790
    },
    {
      "entropy": 5.313150882720947,
      "epoch": 1.9015369836695486,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046409512312318345,
      "loss": 4.783,
      "mean_token_accuracy": 0.23705871105194093,
      "num_tokens": 45368224.0,
      "step": 19795
    },
    {
      "entropy": 5.213432550430298,
      "epoch": 1.9020172910662825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046407654159524994,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24115439206361772,
      "num_tokens": 45379673.0,
      "step": 19800
    },
    {
      "entropy": 5.348945808410645,
      "epoch": 1.9024975984630164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046405795567752055,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23408652544021608,
      "num_tokens": 45391613.0,
      "step": 19805
    },
    {
      "entropy": 5.350124073028565,
      "epoch": 1.9029779058597502,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046403936537042686,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23488420099020005,
      "num_tokens": 45403693.0,
      "step": 19810
    },
    {
      "entropy": 5.319928789138794,
      "epoch": 1.9034582132564841,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046402077067440043,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23544506281614302,
      "num_tokens": 45416151.0,
      "step": 19815
    },
    {
      "entropy": 5.316835355758667,
      "epoch": 1.903938520653218,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046400217158987293,
      "loss": 4.8544,
      "mean_token_accuracy": 0.2322601556777954,
      "num_tokens": 45428276.0,
      "step": 19820
    },
    {
      "entropy": 5.2537205696105955,
      "epoch": 1.9044188280499519,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046398356811727626,
      "loss": 4.7276,
      "mean_token_accuracy": 0.24548156857490538,
      "num_tokens": 45439877.0,
      "step": 19825
    },
    {
      "entropy": 5.311606168746948,
      "epoch": 1.9048991354466858,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004639649602570423,
      "loss": 4.8432,
      "mean_token_accuracy": 0.24545134156942366,
      "num_tokens": 45451459.0,
      "step": 19830
    },
    {
      "entropy": 5.20669994354248,
      "epoch": 1.9053794428434196,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046394634800960314,
      "loss": 4.6999,
      "mean_token_accuracy": 0.24281982332468033,
      "num_tokens": 45463341.0,
      "step": 19835
    },
    {
      "entropy": 5.358258867263794,
      "epoch": 1.9058597502401537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046392773137539074,
      "loss": 4.8093,
      "mean_token_accuracy": 0.242098768055439,
      "num_tokens": 45475585.0,
      "step": 19840
    },
    {
      "entropy": 5.3434733867645265,
      "epoch": 1.9063400576368876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046390911035483744,
      "loss": 4.7796,
      "mean_token_accuracy": 0.2384744554758072,
      "num_tokens": 45487178.0,
      "step": 19845
    },
    {
      "entropy": 5.338760423660278,
      "epoch": 1.9068203650336215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004638904849483756,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23683205842971802,
      "num_tokens": 45498832.0,
      "step": 19850
    },
    {
      "entropy": 5.319513559341431,
      "epoch": 1.9073006724303554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046387185515643756,
      "loss": 4.8464,
      "mean_token_accuracy": 0.23450331687927245,
      "num_tokens": 45511549.0,
      "step": 19855
    },
    {
      "entropy": 5.328990316390991,
      "epoch": 1.9077809798270895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004638532209794559,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23245540708303453,
      "num_tokens": 45523421.0,
      "step": 19860
    },
    {
      "entropy": 5.368109464645386,
      "epoch": 1.9082612872238234,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004638345824178631,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23810838907957077,
      "num_tokens": 45535247.0,
      "step": 19865
    },
    {
      "entropy": 5.258771800994873,
      "epoch": 1.9087415946205573,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046381593947209215,
      "loss": 4.688,
      "mean_token_accuracy": 0.2461639642715454,
      "num_tokens": 45546296.0,
      "step": 19870
    },
    {
      "entropy": 5.31396975517273,
      "epoch": 1.9092219020172911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004637972921425757,
      "loss": 4.8558,
      "mean_token_accuracy": 0.2320055216550827,
      "num_tokens": 45557360.0,
      "step": 19875
    },
    {
      "entropy": 5.316639375686646,
      "epoch": 1.909702209414025,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046377864042974675,
      "loss": 4.7861,
      "mean_token_accuracy": 0.24341519325971603,
      "num_tokens": 45568500.0,
      "step": 19880
    },
    {
      "entropy": 5.341763305664062,
      "epoch": 1.910182516810759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004637599843340384,
      "loss": 4.8646,
      "mean_token_accuracy": 0.22875684201717378,
      "num_tokens": 45579513.0,
      "step": 19885
    },
    {
      "entropy": 5.316559839248657,
      "epoch": 1.9106628242074928,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046374132385588356,
      "loss": 4.7692,
      "mean_token_accuracy": 0.24174658209085464,
      "num_tokens": 45591144.0,
      "step": 19890
    },
    {
      "entropy": 5.327800035476685,
      "epoch": 1.9111431316042267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046372265899571576,
      "loss": 4.8299,
      "mean_token_accuracy": 0.23327631801366805,
      "num_tokens": 45602487.0,
      "step": 19895
    },
    {
      "entropy": 5.40095067024231,
      "epoch": 1.9116234390009605,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046370398975396817,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22882772982120514,
      "num_tokens": 45614448.0,
      "step": 19900
    },
    {
      "entropy": 5.345189428329467,
      "epoch": 1.9121037463976944,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004636853161310743,
      "loss": 4.8172,
      "mean_token_accuracy": 0.23033926486968995,
      "num_tokens": 45625660.0,
      "step": 19905
    },
    {
      "entropy": 5.246729373931885,
      "epoch": 1.9125840537944283,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046366663812746764,
      "loss": 4.7048,
      "mean_token_accuracy": 0.24989546835422516,
      "num_tokens": 45638602.0,
      "step": 19910
    },
    {
      "entropy": 5.269807243347168,
      "epoch": 1.9130643611911622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004636479557435818,
      "loss": 4.7065,
      "mean_token_accuracy": 0.2527043014764786,
      "num_tokens": 45649460.0,
      "step": 19915
    },
    {
      "entropy": 5.2054048538208,
      "epoch": 1.9135446685878963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046362926897985067,
      "loss": 4.6834,
      "mean_token_accuracy": 0.2499219536781311,
      "num_tokens": 45659995.0,
      "step": 19920
    },
    {
      "entropy": 5.245572853088379,
      "epoch": 1.9140249759846302,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000463610577836708,
      "loss": 4.7655,
      "mean_token_accuracy": 0.23687924295663834,
      "num_tokens": 45672262.0,
      "step": 19925
    },
    {
      "entropy": 5.318477821350098,
      "epoch": 1.914505283381364,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046359188231458783,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23806800544261933,
      "num_tokens": 45685210.0,
      "step": 19930
    },
    {
      "entropy": 5.310871410369873,
      "epoch": 1.914985590778098,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046357318241392414,
      "loss": 4.8541,
      "mean_token_accuracy": 0.2354329466819763,
      "num_tokens": 45696418.0,
      "step": 19935
    },
    {
      "entropy": 5.330469942092895,
      "epoch": 1.915465898174832,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000463554478135151,
      "loss": 4.8581,
      "mean_token_accuracy": 0.23355614244937897,
      "num_tokens": 45708393.0,
      "step": 19940
    },
    {
      "entropy": 5.346993827819825,
      "epoch": 1.915946205571566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004635357694787029,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2283647358417511,
      "num_tokens": 45721287.0,
      "step": 19945
    },
    {
      "entropy": 5.2598730564117435,
      "epoch": 1.9164265129682998,
      "grad_norm": 1.015625,
      "learning_rate": 0.000463517056445014,
      "loss": 4.6902,
      "mean_token_accuracy": 0.24743132144212723,
      "num_tokens": 45732846.0,
      "step": 19950
    },
    {
      "entropy": 5.334761619567871,
      "epoch": 1.9169068203650337,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046349833903451884,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22585740834474563,
      "num_tokens": 45743931.0,
      "step": 19955
    },
    {
      "entropy": 5.319524812698364,
      "epoch": 1.9173871277617676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046347961724765196,
      "loss": 4.7754,
      "mean_token_accuracy": 0.24143829941749573,
      "num_tokens": 45755512.0,
      "step": 19960
    },
    {
      "entropy": 5.338935041427613,
      "epoch": 1.9178674351585014,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046346089108484806,
      "loss": 4.7692,
      "mean_token_accuracy": 0.2408650740981102,
      "num_tokens": 45766793.0,
      "step": 19965
    },
    {
      "entropy": 5.32054295539856,
      "epoch": 1.9183477425552353,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046344216054654193,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23398657888174057,
      "num_tokens": 45778214.0,
      "step": 19970
    },
    {
      "entropy": 5.420439624786377,
      "epoch": 1.9188280499519692,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046342342563316833,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2190099239349365,
      "num_tokens": 45791255.0,
      "step": 19975
    },
    {
      "entropy": 5.3943469524383545,
      "epoch": 1.919308357348703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046340468634516223,
      "loss": 4.8334,
      "mean_token_accuracy": 0.22953826040029526,
      "num_tokens": 45803351.0,
      "step": 19980
    },
    {
      "entropy": 5.287267541885376,
      "epoch": 1.919788664745437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046338594268295884,
      "loss": 4.7899,
      "mean_token_accuracy": 0.2422609582543373,
      "num_tokens": 45814986.0,
      "step": 19985
    },
    {
      "entropy": 5.222335529327393,
      "epoch": 1.9202689721421708,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004633671946469933,
      "loss": 4.7635,
      "mean_token_accuracy": 0.24133737832307817,
      "num_tokens": 45827073.0,
      "step": 19990
    },
    {
      "entropy": 5.321532773971557,
      "epoch": 1.920749279538905,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046334844223770076,
      "loss": 4.82,
      "mean_token_accuracy": 0.24343963712453842,
      "num_tokens": 45837808.0,
      "step": 19995
    },
    {
      "entropy": 5.3151586055755615,
      "epoch": 1.9212295869356388,
      "grad_norm": 1.0,
      "learning_rate": 0.00046332968545551674,
      "loss": 4.829,
      "mean_token_accuracy": 0.24208138436079024,
      "num_tokens": 45849932.0,
      "step": 20000
    },
    {
      "entropy": 5.316223192214966,
      "epoch": 1.9217098943323727,
      "grad_norm": 1.0,
      "learning_rate": 0.0004633109243008765,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23708903789520264,
      "num_tokens": 45860961.0,
      "step": 20005
    },
    {
      "entropy": 5.266116380691528,
      "epoch": 1.9221902017291066,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004632921587742159,
      "loss": 4.749,
      "mean_token_accuracy": 0.24234411120414734,
      "num_tokens": 45872251.0,
      "step": 20010
    },
    {
      "entropy": 5.424469137191773,
      "epoch": 1.9226705091258407,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046327338887597043,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23287390917539597,
      "num_tokens": 45883234.0,
      "step": 20015
    },
    {
      "entropy": 5.3105216979980465,
      "epoch": 1.9231508165225746,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463254614606576,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23972945660352707,
      "num_tokens": 45894003.0,
      "step": 20020
    },
    {
      "entropy": 5.221765613555908,
      "epoch": 1.9236311239193085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004632358359664683,
      "loss": 4.7306,
      "mean_token_accuracy": 0.24168601930141448,
      "num_tokens": 45906109.0,
      "step": 20025
    },
    {
      "entropy": 5.30587477684021,
      "epoch": 1.9241114313160423,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046321705295608356,
      "loss": 4.863,
      "mean_token_accuracy": 0.23678205162286758,
      "num_tokens": 45918372.0,
      "step": 20030
    },
    {
      "entropy": 5.374008989334106,
      "epoch": 1.9245917387127762,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046319826557585764,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22779334038496019,
      "num_tokens": 45928750.0,
      "step": 20035
    },
    {
      "entropy": 5.375155448913574,
      "epoch": 1.92507204610951,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004631794738262269,
      "loss": 4.8573,
      "mean_token_accuracy": 0.24168580025434494,
      "num_tokens": 45939297.0,
      "step": 20040
    },
    {
      "entropy": 5.327069187164307,
      "epoch": 1.925552353506244,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004631606777076275,
      "loss": 4.7885,
      "mean_token_accuracy": 0.2369877114892006,
      "num_tokens": 45949986.0,
      "step": 20045
    },
    {
      "entropy": 5.314167594909668,
      "epoch": 1.9260326609029779,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046314187722049587,
      "loss": 4.8187,
      "mean_token_accuracy": 0.23713724315166473,
      "num_tokens": 45960747.0,
      "step": 20050
    },
    {
      "entropy": 5.327861547470093,
      "epoch": 1.9265129682997117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046312307236526863,
      "loss": 4.8147,
      "mean_token_accuracy": 0.2420434758067131,
      "num_tokens": 45972260.0,
      "step": 20055
    },
    {
      "entropy": 5.289991855621338,
      "epoch": 1.9269932756964456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046310426314238217,
      "loss": 4.744,
      "mean_token_accuracy": 0.24283111989498138,
      "num_tokens": 45983589.0,
      "step": 20060
    },
    {
      "entropy": 5.251757669448852,
      "epoch": 1.9274735830931795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004630854495522733,
      "loss": 4.785,
      "mean_token_accuracy": 0.2428109362721443,
      "num_tokens": 45994776.0,
      "step": 20065
    },
    {
      "entropy": 5.380179405212402,
      "epoch": 1.9279538904899134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046306663159537874,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23233729153871535,
      "num_tokens": 46005530.0,
      "step": 20070
    },
    {
      "entropy": 5.297795629501342,
      "epoch": 1.9284341978866475,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046304780927213554,
      "loss": 4.7605,
      "mean_token_accuracy": 0.24644196927547454,
      "num_tokens": 46017538.0,
      "step": 20075
    },
    {
      "entropy": 5.2931435108184814,
      "epoch": 1.9289145052833814,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046302898258298046,
      "loss": 4.748,
      "mean_token_accuracy": 0.23416173160076142,
      "num_tokens": 46029011.0,
      "step": 20080
    },
    {
      "entropy": 5.255220937728882,
      "epoch": 1.9293948126801153,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004630101515283509,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24356022328138352,
      "num_tokens": 46040354.0,
      "step": 20085
    },
    {
      "entropy": 5.28432674407959,
      "epoch": 1.9298751200768491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046299131610868377,
      "loss": 4.7749,
      "mean_token_accuracy": 0.24315630048513412,
      "num_tokens": 46051127.0,
      "step": 20090
    },
    {
      "entropy": 5.32443413734436,
      "epoch": 1.9303554274735832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004629724763244165,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2427551954984665,
      "num_tokens": 46062343.0,
      "step": 20095
    },
    {
      "entropy": 5.233885478973389,
      "epoch": 1.9308357348703171,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004629536321759866,
      "loss": 4.7572,
      "mean_token_accuracy": 0.2369578167796135,
      "num_tokens": 46074193.0,
      "step": 20100
    },
    {
      "entropy": 5.31677770614624,
      "epoch": 1.931316042267051,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046293478366383133,
      "loss": 4.7871,
      "mean_token_accuracy": 0.24570767730474471,
      "num_tokens": 46085930.0,
      "step": 20105
    },
    {
      "entropy": 5.369736623764038,
      "epoch": 1.9317963496637849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004629159307883885,
      "loss": 4.8789,
      "mean_token_accuracy": 0.22664726227521897,
      "num_tokens": 46099759.0,
      "step": 20110
    },
    {
      "entropy": 5.453097629547119,
      "epoch": 1.9322766570605188,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004628970735500958,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22117538005113602,
      "num_tokens": 46110907.0,
      "step": 20115
    },
    {
      "entropy": 5.343665599822998,
      "epoch": 1.9327569644572526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046287821194939094,
      "loss": 4.8282,
      "mean_token_accuracy": 0.23513922840356827,
      "num_tokens": 46122167.0,
      "step": 20120
    },
    {
      "entropy": 5.362386083602905,
      "epoch": 1.9332372718539865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046285934598671186,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22885568290948868,
      "num_tokens": 46132948.0,
      "step": 20125
    },
    {
      "entropy": 5.34357180595398,
      "epoch": 1.9337175792507204,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046284047566249665,
      "loss": 4.8693,
      "mean_token_accuracy": 0.23474375158548355,
      "num_tokens": 46144275.0,
      "step": 20130
    },
    {
      "entropy": 5.315613460540772,
      "epoch": 1.9341978866474543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046282160097718336,
      "loss": 4.8347,
      "mean_token_accuracy": 0.24117353558540344,
      "num_tokens": 46156138.0,
      "step": 20135
    },
    {
      "entropy": 5.332522201538086,
      "epoch": 1.9346781940441882,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004628027219312102,
      "loss": 4.8368,
      "mean_token_accuracy": 0.2380861833691597,
      "num_tokens": 46169437.0,
      "step": 20140
    },
    {
      "entropy": 5.306038045883179,
      "epoch": 1.935158501440922,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004627838385250155,
      "loss": 4.7607,
      "mean_token_accuracy": 0.2387181892991066,
      "num_tokens": 46182150.0,
      "step": 20145
    },
    {
      "entropy": 5.353881740570069,
      "epoch": 1.9356388088376562,
      "grad_norm": 1.0,
      "learning_rate": 0.00046276495075903764,
      "loss": 4.8105,
      "mean_token_accuracy": 0.24071639329195021,
      "num_tokens": 46193945.0,
      "step": 20150
    },
    {
      "entropy": 5.342187404632568,
      "epoch": 1.93611911623439,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046274605863371517,
      "loss": 4.9064,
      "mean_token_accuracy": 0.23480011969804765,
      "num_tokens": 46206130.0,
      "step": 20155
    },
    {
      "entropy": 5.262924528121948,
      "epoch": 1.936599423631124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004627271621494868,
      "loss": 4.7382,
      "mean_token_accuracy": 0.24468212574720383,
      "num_tokens": 46217284.0,
      "step": 20160
    },
    {
      "entropy": 5.29738450050354,
      "epoch": 1.9370797310278578,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000462708261306791,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23973924666643143,
      "num_tokens": 46227949.0,
      "step": 20165
    },
    {
      "entropy": 5.290896987915039,
      "epoch": 1.937560038424592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004626893561060669,
      "loss": 4.7552,
      "mean_token_accuracy": 0.23574090003967285,
      "num_tokens": 46239101.0,
      "step": 20170
    },
    {
      "entropy": 5.359926700592041,
      "epoch": 1.9380403458213258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046267044654775324,
      "loss": 4.9571,
      "mean_token_accuracy": 0.23416971266269684,
      "num_tokens": 46251272.0,
      "step": 20175
    },
    {
      "entropy": 5.327955436706543,
      "epoch": 1.9385206532180597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000462651532632289,
      "loss": 4.8868,
      "mean_token_accuracy": 0.23447102606296538,
      "num_tokens": 46262603.0,
      "step": 20180
    },
    {
      "entropy": 5.333182048797608,
      "epoch": 1.9390009606147935,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046263261436011344,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23636692762374878,
      "num_tokens": 46275142.0,
      "step": 20185
    },
    {
      "entropy": 5.310083293914795,
      "epoch": 1.9394812680115274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004626136917316657,
      "loss": 4.8155,
      "mean_token_accuracy": 0.2348065659403801,
      "num_tokens": 46286339.0,
      "step": 20190
    },
    {
      "entropy": 5.291995906829834,
      "epoch": 1.9399615754082613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046259476474738514,
      "loss": 4.7724,
      "mean_token_accuracy": 0.23519996851682662,
      "num_tokens": 46297514.0,
      "step": 20195
    },
    {
      "entropy": 5.2329161167144775,
      "epoch": 1.9404418828049952,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046257583340771123,
      "loss": 4.6885,
      "mean_token_accuracy": 0.24532767832279206,
      "num_tokens": 46309249.0,
      "step": 20200
    },
    {
      "entropy": 5.341589832305909,
      "epoch": 1.940922190201729,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004625568977130835,
      "loss": 4.8546,
      "mean_token_accuracy": 0.23121865093708038,
      "num_tokens": 46320759.0,
      "step": 20205
    },
    {
      "entropy": 5.392199611663818,
      "epoch": 1.941402497598463,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004625379576639414,
      "loss": 4.8612,
      "mean_token_accuracy": 0.23021113723516465,
      "num_tokens": 46332113.0,
      "step": 20210
    },
    {
      "entropy": 5.3305047988891605,
      "epoch": 1.9418828049951968,
      "grad_norm": 1.125,
      "learning_rate": 0.00046251901326072487,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23936256170272827,
      "num_tokens": 46342648.0,
      "step": 20215
    },
    {
      "entropy": 5.269152450561523,
      "epoch": 1.9423631123919307,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046250006450387367,
      "loss": 4.7436,
      "mean_token_accuracy": 0.242187736928463,
      "num_tokens": 46354066.0,
      "step": 20220
    },
    {
      "entropy": 5.377374744415283,
      "epoch": 1.9428434197886646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004624811113938277,
      "loss": 4.8578,
      "mean_token_accuracy": 0.2307800382375717,
      "num_tokens": 46366586.0,
      "step": 20225
    },
    {
      "entropy": 5.349365854263306,
      "epoch": 1.9433237271853987,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004624621539310271,
      "loss": 4.7806,
      "mean_token_accuracy": 0.24163072258234025,
      "num_tokens": 46378155.0,
      "step": 20230
    },
    {
      "entropy": 5.324305200576783,
      "epoch": 1.9438040345821326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046244319211591193,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23240256905555726,
      "num_tokens": 46389631.0,
      "step": 20235
    },
    {
      "entropy": 5.338899898529053,
      "epoch": 1.9442843419788665,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004624242259489223,
      "loss": 4.8863,
      "mean_token_accuracy": 0.22740271240472792,
      "num_tokens": 46401915.0,
      "step": 20240
    },
    {
      "entropy": 5.3762535572052,
      "epoch": 1.9447646493756003,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046240525543049884,
      "loss": 4.8253,
      "mean_token_accuracy": 0.2363262355327606,
      "num_tokens": 46412396.0,
      "step": 20245
    },
    {
      "entropy": 5.251129913330078,
      "epoch": 1.9452449567723344,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046238628056108176,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24211326539516448,
      "num_tokens": 46423089.0,
      "step": 20250
    },
    {
      "entropy": 5.374421834945679,
      "epoch": 1.9457252641690683,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046236730134111166,
      "loss": 4.8419,
      "mean_token_accuracy": 0.22730725705623628,
      "num_tokens": 46434858.0,
      "step": 20255
    },
    {
      "entropy": 5.264446258544922,
      "epoch": 1.9462055715658022,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004623483177710291,
      "loss": 4.7715,
      "mean_token_accuracy": 0.24954359382390975,
      "num_tokens": 46445628.0,
      "step": 20260
    },
    {
      "entropy": 5.294203901290894,
      "epoch": 1.946685878962536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004623293298512751,
      "loss": 4.7811,
      "mean_token_accuracy": 0.23966595232486726,
      "num_tokens": 46457382.0,
      "step": 20265
    },
    {
      "entropy": 5.302719259262085,
      "epoch": 1.94716618635927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046231033758229026,
      "loss": 4.8155,
      "mean_token_accuracy": 0.2294058158993721,
      "num_tokens": 46469765.0,
      "step": 20270
    },
    {
      "entropy": 5.432910490036011,
      "epoch": 1.9476464937560038,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004622913409645154,
      "loss": 4.9259,
      "mean_token_accuracy": 0.22848036587238313,
      "num_tokens": 46481451.0,
      "step": 20275
    },
    {
      "entropy": 5.340350723266601,
      "epoch": 1.9481268011527377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004622723399983919,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24436736553907396,
      "num_tokens": 46491585.0,
      "step": 20280
    },
    {
      "entropy": 5.3414952754974365,
      "epoch": 1.9486071085494716,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046225333468436077,
      "loss": 4.9211,
      "mean_token_accuracy": 0.224766705930233,
      "num_tokens": 46503789.0,
      "step": 20285
    },
    {
      "entropy": 5.390402507781983,
      "epoch": 1.9490874159462055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046223432502286323,
      "loss": 4.8883,
      "mean_token_accuracy": 0.230931556224823,
      "num_tokens": 46516647.0,
      "step": 20290
    },
    {
      "entropy": 5.361443758010864,
      "epoch": 1.9495677233429394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046221531101434056,
      "loss": 4.8258,
      "mean_token_accuracy": 0.22949367612600327,
      "num_tokens": 46527432.0,
      "step": 20295
    },
    {
      "entropy": 5.3544440269470215,
      "epoch": 1.9500480307396733,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004621962926592343,
      "loss": 4.8484,
      "mean_token_accuracy": 0.2350536286830902,
      "num_tokens": 46538990.0,
      "step": 20300
    },
    {
      "entropy": 5.312886333465576,
      "epoch": 1.9505283381364071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000462177269957986,
      "loss": 4.8347,
      "mean_token_accuracy": 0.23235955089330673,
      "num_tokens": 46551291.0,
      "step": 20305
    },
    {
      "entropy": 5.309301328659058,
      "epoch": 1.9510086455331412,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004621582429110373,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23936838656663895,
      "num_tokens": 46562415.0,
      "step": 20310
    },
    {
      "entropy": 5.387483310699463,
      "epoch": 1.9514889529298751,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046213921151883,
      "loss": 4.878,
      "mean_token_accuracy": 0.24066894948482515,
      "num_tokens": 46572544.0,
      "step": 20315
    },
    {
      "entropy": 5.281074666976929,
      "epoch": 1.951969260326609,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004621201757818059,
      "loss": 4.7713,
      "mean_token_accuracy": 0.2395360916852951,
      "num_tokens": 46584694.0,
      "step": 20320
    },
    {
      "entropy": 5.350608491897583,
      "epoch": 1.952449567723343,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046210113570040683,
      "loss": 4.8141,
      "mean_token_accuracy": 0.23648386150598527,
      "num_tokens": 46595086.0,
      "step": 20325
    },
    {
      "entropy": 5.387836360931397,
      "epoch": 1.952929875120077,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000462082091275075,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23762939870357513,
      "num_tokens": 46606627.0,
      "step": 20330
    },
    {
      "entropy": 5.349586868286133,
      "epoch": 1.9534101825168109,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004620630425062526,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23325287103652953,
      "num_tokens": 46618560.0,
      "step": 20335
    },
    {
      "entropy": 5.336157941818238,
      "epoch": 1.9538904899135447,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004620439893943817,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23103161305189132,
      "num_tokens": 46630109.0,
      "step": 20340
    },
    {
      "entropy": 5.285204076766968,
      "epoch": 1.9543707973102786,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004620249319399049,
      "loss": 4.7898,
      "mean_token_accuracy": 0.24304262101650237,
      "num_tokens": 46642247.0,
      "step": 20345
    },
    {
      "entropy": 5.386840438842773,
      "epoch": 1.9548511047070125,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046200587014326455,
      "loss": 4.8538,
      "mean_token_accuracy": 0.23417021781206132,
      "num_tokens": 46653119.0,
      "step": 20350
    },
    {
      "entropy": 5.284787988662719,
      "epoch": 1.9553314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004619868040049031,
      "loss": 4.7746,
      "mean_token_accuracy": 0.24032128751277923,
      "num_tokens": 46664392.0,
      "step": 20355
    },
    {
      "entropy": 5.295685148239135,
      "epoch": 1.9558117195004803,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004619677335252633,
      "loss": 4.7719,
      "mean_token_accuracy": 0.24190528839826583,
      "num_tokens": 46674981.0,
      "step": 20360
    },
    {
      "entropy": 5.301712608337402,
      "epoch": 1.9562920268972142,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046194865870478793,
      "loss": 4.8646,
      "mean_token_accuracy": 0.23131226301193236,
      "num_tokens": 46688692.0,
      "step": 20365
    },
    {
      "entropy": 5.226872491836548,
      "epoch": 1.956772334293948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046192957954391983,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24711971282958983,
      "num_tokens": 46699799.0,
      "step": 20370
    },
    {
      "entropy": 5.304947137832642,
      "epoch": 1.957252641690682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000461910496043102,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2389782816171646,
      "num_tokens": 46711994.0,
      "step": 20375
    },
    {
      "entropy": 5.343099546432495,
      "epoch": 1.9577329490874158,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004618914082027773,
      "loss": 4.8457,
      "mean_token_accuracy": 0.23591487109661102,
      "num_tokens": 46723247.0,
      "step": 20380
    },
    {
      "entropy": 5.401315593719483,
      "epoch": 1.95821325648415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046187231602338926,
      "loss": 4.8747,
      "mean_token_accuracy": 0.2305608794093132,
      "num_tokens": 46732824.0,
      "step": 20385
    },
    {
      "entropy": 5.288520383834839,
      "epoch": 1.9586935638808838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046185321950538086,
      "loss": 4.793,
      "mean_token_accuracy": 0.23967974185943602,
      "num_tokens": 46744640.0,
      "step": 20390
    },
    {
      "entropy": 5.296942710876465,
      "epoch": 1.9591738712776177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004618341186491955,
      "loss": 4.8141,
      "mean_token_accuracy": 0.23651195466518402,
      "num_tokens": 46757513.0,
      "step": 20395
    },
    {
      "entropy": 5.2820416450500485,
      "epoch": 1.9596541786743515,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004618150134552768,
      "loss": 4.7612,
      "mean_token_accuracy": 0.24186078011989592,
      "num_tokens": 46769021.0,
      "step": 20400
    },
    {
      "entropy": 5.277998924255371,
      "epoch": 1.9601344860710856,
      "grad_norm": 1.0,
      "learning_rate": 0.000461795903924068,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2422717973589897,
      "num_tokens": 46781175.0,
      "step": 20405
    },
    {
      "entropy": 5.3039796352386475,
      "epoch": 1.9606147934678195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046177679005601313,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24465503990650178,
      "num_tokens": 46792256.0,
      "step": 20410
    },
    {
      "entropy": 5.255706787109375,
      "epoch": 1.9610951008645534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004617576718515558,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24627446085214616,
      "num_tokens": 46803215.0,
      "step": 20415
    },
    {
      "entropy": 5.364198350906372,
      "epoch": 1.9615754082612873,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004617385493111399,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23957298547029496,
      "num_tokens": 46813770.0,
      "step": 20420
    },
    {
      "entropy": 5.335280323028565,
      "epoch": 1.9620557156580212,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004617194224352093,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23685489147901534,
      "num_tokens": 46826923.0,
      "step": 20425
    },
    {
      "entropy": 5.278520727157593,
      "epoch": 1.962536023054755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004617002912242083,
      "loss": 4.7413,
      "mean_token_accuracy": 0.2448288232088089,
      "num_tokens": 46837761.0,
      "step": 20430
    },
    {
      "entropy": 5.310571908950806,
      "epoch": 1.963016330451489,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046168115567858084,
      "loss": 4.8137,
      "mean_token_accuracy": 0.2335599273443222,
      "num_tokens": 46848190.0,
      "step": 20435
    },
    {
      "entropy": 5.264204025268555,
      "epoch": 1.9634966378482228,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046166201579877125,
      "loss": 4.7083,
      "mean_token_accuracy": 0.24483490288257598,
      "num_tokens": 46858643.0,
      "step": 20440
    },
    {
      "entropy": 5.327106142044068,
      "epoch": 1.9639769452449567,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616428715852239,
      "loss": 4.7962,
      "mean_token_accuracy": 0.2404761865735054,
      "num_tokens": 46868711.0,
      "step": 20445
    },
    {
      "entropy": 5.323641538619995,
      "epoch": 1.9644572526416906,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004616237230383833,
      "loss": 4.8812,
      "mean_token_accuracy": 0.2326974406838417,
      "num_tokens": 46880045.0,
      "step": 20450
    },
    {
      "entropy": 5.293602752685547,
      "epoch": 1.9649375600384245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046160457015869414,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2329004481434822,
      "num_tokens": 46890680.0,
      "step": 20455
    },
    {
      "entropy": 5.283541107177735,
      "epoch": 1.9654178674351583,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046158541294660083,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23503431230783461,
      "num_tokens": 46904807.0,
      "step": 20460
    },
    {
      "entropy": 5.369705677032471,
      "epoch": 1.9658981748318924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004615662514025483,
      "loss": 4.7666,
      "mean_token_accuracy": 0.23994292616844176,
      "num_tokens": 46916039.0,
      "step": 20465
    },
    {
      "entropy": 5.333849239349365,
      "epoch": 1.9663784822286263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046154708552698147,
      "loss": 4.7808,
      "mean_token_accuracy": 0.23484014719724655,
      "num_tokens": 46927175.0,
      "step": 20470
    },
    {
      "entropy": 5.2499795913696286,
      "epoch": 1.9668587896253602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046152791532034517,
      "loss": 4.6911,
      "mean_token_accuracy": 0.24924637228250504,
      "num_tokens": 46938711.0,
      "step": 20475
    },
    {
      "entropy": 5.28057541847229,
      "epoch": 1.967339097022094,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046150874078308463,
      "loss": 4.7953,
      "mean_token_accuracy": 0.2373875617980957,
      "num_tokens": 46951002.0,
      "step": 20480
    },
    {
      "entropy": 5.378773736953735,
      "epoch": 1.9678194044188282,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004614895619156449,
      "loss": 4.8526,
      "mean_token_accuracy": 0.23698788285255432,
      "num_tokens": 46962673.0,
      "step": 20485
    },
    {
      "entropy": 5.328410863876343,
      "epoch": 1.968299711815562,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046147037871847125,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23796840012073517,
      "num_tokens": 46973955.0,
      "step": 20490
    },
    {
      "entropy": 5.346768140792847,
      "epoch": 1.968780019212296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004614511911920092,
      "loss": 4.9088,
      "mean_token_accuracy": 0.2299971491098404,
      "num_tokens": 46985791.0,
      "step": 20495
    },
    {
      "entropy": 5.45083703994751,
      "epoch": 1.9692603266090298,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046143199933670404,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22230999320745468,
      "num_tokens": 46998437.0,
      "step": 20500
    },
    {
      "entropy": 5.344652891159058,
      "epoch": 1.9697406340057637,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004614128031530015,
      "loss": 4.7549,
      "mean_token_accuracy": 0.23604216426610947,
      "num_tokens": 47009666.0,
      "step": 20505
    },
    {
      "entropy": 5.346072626113892,
      "epoch": 1.9702209414024976,
      "grad_norm": 1.0,
      "learning_rate": 0.00046139360264134724,
      "loss": 4.846,
      "mean_token_accuracy": 0.23474851250648499,
      "num_tokens": 47021170.0,
      "step": 20510
    },
    {
      "entropy": 5.341988182067871,
      "epoch": 1.9707012487992315,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004613743978021869,
      "loss": 4.7893,
      "mean_token_accuracy": 0.24633179754018783,
      "num_tokens": 47033077.0,
      "step": 20515
    },
    {
      "entropy": 5.339973974227905,
      "epoch": 1.9711815561959654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046135518863596654,
      "loss": 4.8367,
      "mean_token_accuracy": 0.23571707606315612,
      "num_tokens": 47044858.0,
      "step": 20520
    },
    {
      "entropy": 5.3737787246704105,
      "epoch": 1.9716618635926992,
      "grad_norm": 1.0,
      "learning_rate": 0.00046133597514313204,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2261406436562538,
      "num_tokens": 47057587.0,
      "step": 20525
    },
    {
      "entropy": 5.412638711929321,
      "epoch": 1.9721421709894331,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004613167573241295,
      "loss": 4.8477,
      "mean_token_accuracy": 0.22965983003377916,
      "num_tokens": 47068690.0,
      "step": 20530
    },
    {
      "entropy": 5.321409034729004,
      "epoch": 1.972622478386167,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004612975351794051,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23973421156406402,
      "num_tokens": 47081279.0,
      "step": 20535
    },
    {
      "entropy": 5.347462892532349,
      "epoch": 1.973102785782901,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004612783087094051,
      "loss": 4.8607,
      "mean_token_accuracy": 0.23635558933019638,
      "num_tokens": 47093372.0,
      "step": 20540
    },
    {
      "entropy": 5.428307628631591,
      "epoch": 1.973583093179635,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046125907791457594,
      "loss": 4.935,
      "mean_token_accuracy": 0.23006318807601928,
      "num_tokens": 47104332.0,
      "step": 20545
    },
    {
      "entropy": 5.3654345035552975,
      "epoch": 1.9740634005763689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046123984279536405,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23310890346765517,
      "num_tokens": 47116941.0,
      "step": 20550
    },
    {
      "entropy": 5.401148080825806,
      "epoch": 1.9745437079731027,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046122060335221604,
      "loss": 4.8576,
      "mean_token_accuracy": 0.2324142798781395,
      "num_tokens": 47128473.0,
      "step": 20555
    },
    {
      "entropy": 5.3331139087677,
      "epoch": 1.9750240153698368,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046120135958557855,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23128969669342042,
      "num_tokens": 47140586.0,
      "step": 20560
    },
    {
      "entropy": 5.277958583831787,
      "epoch": 1.9755043227665707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046118211149589843,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2451483801007271,
      "num_tokens": 47150875.0,
      "step": 20565
    },
    {
      "entropy": 5.291774654388428,
      "epoch": 1.9759846301633046,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004611628590836225,
      "loss": 4.7199,
      "mean_token_accuracy": 0.2494172841310501,
      "num_tokens": 47162074.0,
      "step": 20570
    },
    {
      "entropy": 5.26590781211853,
      "epoch": 1.9764649375600385,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004611436023491979,
      "loss": 4.7585,
      "mean_token_accuracy": 0.2426767259836197,
      "num_tokens": 47174544.0,
      "step": 20575
    },
    {
      "entropy": 5.361579322814942,
      "epoch": 1.9769452449567724,
      "grad_norm": 1.0,
      "learning_rate": 0.0004611243412930714,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23082381784915923,
      "num_tokens": 47186260.0,
      "step": 20580
    },
    {
      "entropy": 5.336852169036865,
      "epoch": 1.9774255523535063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046110507591569047,
      "loss": 4.7784,
      "mean_token_accuracy": 0.2373894199728966,
      "num_tokens": 47198059.0,
      "step": 20585
    },
    {
      "entropy": 5.3120215892791744,
      "epoch": 1.9779058597502401,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004610858062175023,
      "loss": 4.7287,
      "mean_token_accuracy": 0.23927282243967057,
      "num_tokens": 47208300.0,
      "step": 20590
    },
    {
      "entropy": 5.328384590148926,
      "epoch": 1.978386167146974,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046106653219895417,
      "loss": 4.7844,
      "mean_token_accuracy": 0.24204044789075851,
      "num_tokens": 47220135.0,
      "step": 20595
    },
    {
      "entropy": 5.427314424514771,
      "epoch": 1.978866474543708,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004610472538604938,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2317245453596115,
      "num_tokens": 47230385.0,
      "step": 20600
    },
    {
      "entropy": 5.268126106262207,
      "epoch": 1.9793467819404418,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046102797120256854,
      "loss": 4.7175,
      "mean_token_accuracy": 0.23851021826267244,
      "num_tokens": 47242593.0,
      "step": 20605
    },
    {
      "entropy": 5.3398229598999025,
      "epoch": 1.9798270893371757,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004610086842256262,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23718530237674712,
      "num_tokens": 47254223.0,
      "step": 20610
    },
    {
      "entropy": 5.32624454498291,
      "epoch": 1.9803073967339095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004609893929301146,
      "loss": 4.8604,
      "mean_token_accuracy": 0.23888598382472992,
      "num_tokens": 47266559.0,
      "step": 20615
    },
    {
      "entropy": 5.407634639739991,
      "epoch": 1.9807877041306436,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004609700973164816,
      "loss": 4.842,
      "mean_token_accuracy": 0.23708957731723784,
      "num_tokens": 47276438.0,
      "step": 20620
    },
    {
      "entropy": 5.328380584716797,
      "epoch": 1.9812680115273775,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004609507973851751,
      "loss": 4.8464,
      "mean_token_accuracy": 0.2324322819709778,
      "num_tokens": 47287791.0,
      "step": 20625
    },
    {
      "entropy": 5.269200181961059,
      "epoch": 1.9817483189241114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046093149313664316,
      "loss": 4.7988,
      "mean_token_accuracy": 0.23773234933614731,
      "num_tokens": 47300282.0,
      "step": 20630
    },
    {
      "entropy": 5.426207876205444,
      "epoch": 1.9822286263208453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004609121845713342,
      "loss": 4.8753,
      "mean_token_accuracy": 0.23744504600763322,
      "num_tokens": 47310473.0,
      "step": 20635
    },
    {
      "entropy": 5.315967035293579,
      "epoch": 1.9827089337175794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004608928716896963,
      "loss": 4.7758,
      "mean_token_accuracy": 0.2403537690639496,
      "num_tokens": 47322245.0,
      "step": 20640
    },
    {
      "entropy": 5.371425151824951,
      "epoch": 1.9831892411143133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004608735544921778,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2423245519399643,
      "num_tokens": 47334316.0,
      "step": 20645
    },
    {
      "entropy": 5.402907943725586,
      "epoch": 1.9836695485110472,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046085423297922745,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2362591192126274,
      "num_tokens": 47344909.0,
      "step": 20650
    },
    {
      "entropy": 5.2584466457366945,
      "epoch": 1.984149855907781,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046083490715129367,
      "loss": 4.7129,
      "mean_token_accuracy": 0.24493313282728196,
      "num_tokens": 47356682.0,
      "step": 20655
    },
    {
      "entropy": 5.288189220428467,
      "epoch": 1.984630163304515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004608155770088251,
      "loss": 4.7075,
      "mean_token_accuracy": 0.24786777049303055,
      "num_tokens": 47368362.0,
      "step": 20660
    },
    {
      "entropy": 5.375208616256714,
      "epoch": 1.9851104707012488,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046079624255227066,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23772752285003662,
      "num_tokens": 47379932.0,
      "step": 20665
    },
    {
      "entropy": 5.368642091751099,
      "epoch": 1.9855907780979827,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004607769037820791,
      "loss": 4.9159,
      "mean_token_accuracy": 0.22869734168052674,
      "num_tokens": 47391244.0,
      "step": 20670
    },
    {
      "entropy": 5.352165699005127,
      "epoch": 1.9860710854947166,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004607575606986995,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23427304476499558,
      "num_tokens": 47404005.0,
      "step": 20675
    },
    {
      "entropy": 5.264745044708252,
      "epoch": 1.9865513928914504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004607382133025809,
      "loss": 4.7186,
      "mean_token_accuracy": 0.24213519990444182,
      "num_tokens": 47414555.0,
      "step": 20680
    },
    {
      "entropy": 5.324719476699829,
      "epoch": 1.9870317002881843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046071886159417257,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2327147275209427,
      "num_tokens": 47425902.0,
      "step": 20685
    },
    {
      "entropy": 5.253840637207031,
      "epoch": 1.9875120076849182,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004606995055739238,
      "loss": 4.7269,
      "mean_token_accuracy": 0.243233759701252,
      "num_tokens": 47438068.0,
      "step": 20690
    },
    {
      "entropy": 5.402299261093139,
      "epoch": 1.9879923150816523,
      "grad_norm": 1.0,
      "learning_rate": 0.00046068014524228374,
      "loss": 4.8674,
      "mean_token_accuracy": 0.23186666518449783,
      "num_tokens": 47450322.0,
      "step": 20695
    },
    {
      "entropy": 5.3117505550384525,
      "epoch": 1.9884726224783862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046066078059970217,
      "loss": 4.7994,
      "mean_token_accuracy": 0.24054647982120514,
      "num_tokens": 47462615.0,
      "step": 20700
    },
    {
      "entropy": 5.378345108032226,
      "epoch": 1.98895292987512,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004606414116466286,
      "loss": 4.7506,
      "mean_token_accuracy": 0.23532755970954894,
      "num_tokens": 47472530.0,
      "step": 20705
    },
    {
      "entropy": 5.259029865264893,
      "epoch": 1.989433237271854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046062203838351267,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2455167979001999,
      "num_tokens": 47484217.0,
      "step": 20710
    },
    {
      "entropy": 5.346635341644287,
      "epoch": 1.989913544668588,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046060266081080414,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23379007428884507,
      "num_tokens": 47496365.0,
      "step": 20715
    },
    {
      "entropy": 5.3326152801513675,
      "epoch": 1.990393852065322,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004605832789289531,
      "loss": 4.7986,
      "mean_token_accuracy": 0.24910591095685958,
      "num_tokens": 47509437.0,
      "step": 20720
    },
    {
      "entropy": 5.363722944259644,
      "epoch": 1.9908741594620558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004605638927384093,
      "loss": 4.8603,
      "mean_token_accuracy": 0.2328261688351631,
      "num_tokens": 47520913.0,
      "step": 20725
    },
    {
      "entropy": 5.390057039260864,
      "epoch": 1.9913544668587897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046054450223962284,
      "loss": 4.846,
      "mean_token_accuracy": 0.2308248519897461,
      "num_tokens": 47532015.0,
      "step": 20730
    },
    {
      "entropy": 5.429323673248291,
      "epoch": 1.9918347742555236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046052510743304405,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22847892343997955,
      "num_tokens": 47543675.0,
      "step": 20735
    },
    {
      "entropy": 5.349146938323974,
      "epoch": 1.9923150816522575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004605057083191232,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2330466017127037,
      "num_tokens": 47555757.0,
      "step": 20740
    },
    {
      "entropy": 5.398453187942505,
      "epoch": 1.9927953890489913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004604863048983106,
      "loss": 4.9267,
      "mean_token_accuracy": 0.23002221286296845,
      "num_tokens": 47568043.0,
      "step": 20745
    },
    {
      "entropy": 5.364987564086914,
      "epoch": 1.9932756964457252,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004604668971710568,
      "loss": 4.7863,
      "mean_token_accuracy": 0.24078710824251176,
      "num_tokens": 47577811.0,
      "step": 20750
    },
    {
      "entropy": 5.436523818969727,
      "epoch": 1.993756003842459,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004604474851378124,
      "loss": 4.8873,
      "mean_token_accuracy": 0.23057116717100143,
      "num_tokens": 47588309.0,
      "step": 20755
    },
    {
      "entropy": 5.45635347366333,
      "epoch": 1.994236311239193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046042806879902803,
      "loss": 4.9671,
      "mean_token_accuracy": 0.22272566556930543,
      "num_tokens": 47599814.0,
      "step": 20760
    },
    {
      "entropy": 5.425264835357666,
      "epoch": 1.9947166186359269,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004604086481551546,
      "loss": 4.8654,
      "mean_token_accuracy": 0.23384713530540466,
      "num_tokens": 47611851.0,
      "step": 20765
    },
    {
      "entropy": 5.338033533096313,
      "epoch": 1.9951969260326607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004603892232066428,
      "loss": 4.7192,
      "mean_token_accuracy": 0.24363452643156053,
      "num_tokens": 47624056.0,
      "step": 20770
    },
    {
      "entropy": 5.306674337387085,
      "epoch": 1.9956772334293948,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046036979395394374,
      "loss": 4.8233,
      "mean_token_accuracy": 0.23897064924240113,
      "num_tokens": 47637214.0,
      "step": 20775
    },
    {
      "entropy": 5.345308780670166,
      "epoch": 1.9961575408261287,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004603503603975085,
      "loss": 4.7809,
      "mean_token_accuracy": 0.2384583592414856,
      "num_tokens": 47647498.0,
      "step": 20780
    },
    {
      "entropy": 5.338013172149658,
      "epoch": 1.9966378482228626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004603309225377883,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23929886221885682,
      "num_tokens": 47659576.0,
      "step": 20785
    },
    {
      "entropy": 5.370355176925659,
      "epoch": 1.9971181556195965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004603114803752344,
      "loss": 4.7887,
      "mean_token_accuracy": 0.23635707050561905,
      "num_tokens": 47670631.0,
      "step": 20790
    },
    {
      "entropy": 5.321663808822632,
      "epoch": 1.9975984630163306,
      "grad_norm": 1.75,
      "learning_rate": 0.00046029203391029813,
      "loss": 4.728,
      "mean_token_accuracy": 0.23331705629825591,
      "num_tokens": 47681647.0,
      "step": 20795
    },
    {
      "entropy": 5.303390264511108,
      "epoch": 1.9980787704130645,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046027258314343107,
      "loss": 4.7603,
      "mean_token_accuracy": 0.242738176882267,
      "num_tokens": 47694386.0,
      "step": 20800
    },
    {
      "entropy": 5.354677295684814,
      "epoch": 1.9985590778097984,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046025312807508487,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23870779424905778,
      "num_tokens": 47704897.0,
      "step": 20805
    },
    {
      "entropy": 5.338613080978393,
      "epoch": 1.9990393852065322,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046023366870571097,
      "loss": 4.8452,
      "mean_token_accuracy": 0.23530489951372147,
      "num_tokens": 47715771.0,
      "step": 20810
    },
    {
      "entropy": 5.365951156616211,
      "epoch": 1.9995196926032661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046021420503576145,
      "loss": 4.7449,
      "mean_token_accuracy": 0.24496710300445557,
      "num_tokens": 47726295.0,
      "step": 20815
    },
    {
      "entropy": 5.290918588638306,
      "epoch": 2.0,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004601947370656879,
      "loss": 4.8274,
      "mean_token_accuracy": 0.23412739783525466,
      "num_tokens": 47737072.0,
      "step": 20820
    },
    {
      "entropy": 5.361708354949951,
      "epoch": 2.000480307396734,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004601752647959426,
      "loss": 4.7891,
      "mean_token_accuracy": 0.23413755297660827,
      "num_tokens": 47750379.0,
      "step": 20825
    },
    {
      "entropy": 5.41068058013916,
      "epoch": 2.0009606147934678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004601557882269775,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23524304032325744,
      "num_tokens": 47763059.0,
      "step": 20830
    },
    {
      "entropy": 5.335291719436645,
      "epoch": 2.0014409221902016,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004601363073592447,
      "loss": 4.718,
      "mean_token_accuracy": 0.24371265918016433,
      "num_tokens": 47775513.0,
      "step": 20835
    },
    {
      "entropy": 5.264129638671875,
      "epoch": 2.0019212295869355,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004601168221931967,
      "loss": 4.6663,
      "mean_token_accuracy": 0.24482130855321885,
      "num_tokens": 47786215.0,
      "step": 20840
    },
    {
      "entropy": 5.3079320907592775,
      "epoch": 2.0024015369836694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004600973327292857,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24480638951063155,
      "num_tokens": 47796894.0,
      "step": 20845
    },
    {
      "entropy": 5.318791818618775,
      "epoch": 2.0028818443804033,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046007783896796436,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2389284536242485,
      "num_tokens": 47808505.0,
      "step": 20850
    },
    {
      "entropy": 5.328690004348755,
      "epoch": 2.003362151777137,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004600583409096851,
      "loss": 4.6596,
      "mean_token_accuracy": 0.24937456101179123,
      "num_tokens": 47819617.0,
      "step": 20855
    },
    {
      "entropy": 5.279451465606689,
      "epoch": 2.0038424591738715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046003883855490066,
      "loss": 4.6926,
      "mean_token_accuracy": 0.2490469202399254,
      "num_tokens": 47830655.0,
      "step": 20860
    },
    {
      "entropy": 5.254589939117432,
      "epoch": 2.0043227665706054,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004600193319040638,
      "loss": 4.6383,
      "mean_token_accuracy": 0.24873753637075424,
      "num_tokens": 47842439.0,
      "step": 20865
    },
    {
      "entropy": 5.229692888259888,
      "epoch": 2.0048030739673393,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045999982095762756,
      "loss": 4.6099,
      "mean_token_accuracy": 0.2568904846906662,
      "num_tokens": 47853327.0,
      "step": 20870
    },
    {
      "entropy": 5.347901296615601,
      "epoch": 2.005283381364073,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045998030571604473,
      "loss": 4.7914,
      "mean_token_accuracy": 0.23871316611766816,
      "num_tokens": 47867155.0,
      "step": 20875
    },
    {
      "entropy": 5.329977941513062,
      "epoch": 2.005763688760807,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004599607861797685,
      "loss": 4.7391,
      "mean_token_accuracy": 0.23901071697473525,
      "num_tokens": 47879641.0,
      "step": 20880
    },
    {
      "entropy": 5.342166709899902,
      "epoch": 2.006243996157541,
      "grad_norm": 1.0,
      "learning_rate": 0.00045994126234925203,
      "loss": 4.7618,
      "mean_token_accuracy": 0.23755911588668824,
      "num_tokens": 47890512.0,
      "step": 20885
    },
    {
      "entropy": 5.364368867874146,
      "epoch": 2.006724303554275,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045992173422494865,
      "loss": 4.7765,
      "mean_token_accuracy": 0.23811352998018265,
      "num_tokens": 47902575.0,
      "step": 20890
    },
    {
      "entropy": 5.308557367324829,
      "epoch": 2.0072046109510087,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004599022018073117,
      "loss": 4.6798,
      "mean_token_accuracy": 0.24892965108156204,
      "num_tokens": 47915447.0,
      "step": 20895
    },
    {
      "entropy": 5.313933372497559,
      "epoch": 2.0076849183477425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004598826650967946,
      "loss": 4.7332,
      "mean_token_accuracy": 0.23779838383197785,
      "num_tokens": 47925702.0,
      "step": 20900
    },
    {
      "entropy": 5.3192737102508545,
      "epoch": 2.0081652257444764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045986312409385105,
      "loss": 4.7271,
      "mean_token_accuracy": 0.246273972094059,
      "num_tokens": 47936621.0,
      "step": 20905
    },
    {
      "entropy": 5.271324825286865,
      "epoch": 2.0086455331412103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004598435787989347,
      "loss": 4.6768,
      "mean_token_accuracy": 0.25027011781930925,
      "num_tokens": 47948026.0,
      "step": 20910
    },
    {
      "entropy": 5.371472692489624,
      "epoch": 2.009125840537944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045982402921249934,
      "loss": 4.7728,
      "mean_token_accuracy": 0.24045882672071456,
      "num_tokens": 47959976.0,
      "step": 20915
    },
    {
      "entropy": 5.349729013442993,
      "epoch": 2.009606147934678,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004598044753349988,
      "loss": 4.7721,
      "mean_token_accuracy": 0.23634253740310668,
      "num_tokens": 47972071.0,
      "step": 20920
    },
    {
      "entropy": 5.340642070770263,
      "epoch": 2.010086455331412,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045978491716688706,
      "loss": 4.6974,
      "mean_token_accuracy": 0.24409151524305345,
      "num_tokens": 47984789.0,
      "step": 20925
    },
    {
      "entropy": 5.3061014175415036,
      "epoch": 2.010566762728146,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004597653547086184,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24705082178115845,
      "num_tokens": 47995661.0,
      "step": 20930
    },
    {
      "entropy": 5.360064363479614,
      "epoch": 2.0110470701248797,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004597457879606467,
      "loss": 4.796,
      "mean_token_accuracy": 0.23789698332548143,
      "num_tokens": 48006043.0,
      "step": 20935
    },
    {
      "entropy": 5.357745027542114,
      "epoch": 2.011527377521614,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045972621692342636,
      "loss": 4.8258,
      "mean_token_accuracy": 0.2385847732424736,
      "num_tokens": 48016904.0,
      "step": 20940
    },
    {
      "entropy": 5.369758605957031,
      "epoch": 2.012007684918348,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045970664159741186,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2423244222998619,
      "num_tokens": 48026533.0,
      "step": 20945
    },
    {
      "entropy": 5.2916487693786625,
      "epoch": 2.012487992315082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045968706198305765,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24820927679538726,
      "num_tokens": 48038280.0,
      "step": 20950
    },
    {
      "entropy": 5.22772445678711,
      "epoch": 2.0129682997118157,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045966747808081824,
      "loss": 4.681,
      "mean_token_accuracy": 0.2490616261959076,
      "num_tokens": 48049663.0,
      "step": 20955
    },
    {
      "entropy": 5.312412166595459,
      "epoch": 2.0134486071085496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004596478898911483,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2373359888792038,
      "num_tokens": 48059826.0,
      "step": 20960
    },
    {
      "entropy": 5.338084506988525,
      "epoch": 2.0139289145052834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045962829741450265,
      "loss": 4.651,
      "mean_token_accuracy": 0.2489929124712944,
      "num_tokens": 48071097.0,
      "step": 20965
    },
    {
      "entropy": 5.321284532546997,
      "epoch": 2.0144092219020173,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004596087006513361,
      "loss": 4.697,
      "mean_token_accuracy": 0.24354044497013091,
      "num_tokens": 48082297.0,
      "step": 20970
    },
    {
      "entropy": 5.254949331283569,
      "epoch": 2.014889529298751,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045958909960210385,
      "loss": 4.6581,
      "mean_token_accuracy": 0.24661597162485122,
      "num_tokens": 48092946.0,
      "step": 20975
    },
    {
      "entropy": 5.26236481666565,
      "epoch": 2.015369836695485,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045956949426726075,
      "loss": 4.7204,
      "mean_token_accuracy": 0.24613651782274246,
      "num_tokens": 48105000.0,
      "step": 20980
    },
    {
      "entropy": 5.2942619800567625,
      "epoch": 2.015850144092219,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045954988464726203,
      "loss": 4.673,
      "mean_token_accuracy": 0.24496925622224808,
      "num_tokens": 48117513.0,
      "step": 20985
    },
    {
      "entropy": 5.415054702758789,
      "epoch": 2.016330451488953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000459530270742563,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22550372481346131,
      "num_tokens": 48130895.0,
      "step": 20990
    },
    {
      "entropy": 5.313234138488769,
      "epoch": 2.0168107588856867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045951065255361905,
      "loss": 4.6681,
      "mean_token_accuracy": 0.2417049989104271,
      "num_tokens": 48143348.0,
      "step": 20995
    },
    {
      "entropy": 5.311700248718262,
      "epoch": 2.0172910662824206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004594910300808856,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24818805158138274,
      "num_tokens": 48154959.0,
      "step": 21000
    },
    {
      "epoch": 2.0172910662824206,
      "eval_entropy": 5.063665089288727,
      "eval_loss": 4.843585968017578,
      "eval_mean_token_accuracy": 0.2452691614102011,
      "eval_num_tokens": 48154959.0,
      "eval_runtime": 20.7071,
      "eval_samples_per_second": 1584.721,
      "eval_steps_per_second": 198.096,
      "step": 21000
    },
    {
      "entropy": 5.231966876983643,
      "epoch": 2.0177713736791545,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004594714033248183,
      "loss": 4.5977,
      "mean_token_accuracy": 0.25526583194732666,
      "num_tokens": 48165227.0,
      "step": 21005
    },
    {
      "entropy": 5.37347936630249,
      "epoch": 2.0182516810758884,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004594517722858728,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23621948659420014,
      "num_tokens": 48175133.0,
      "step": 21010
    },
    {
      "entropy": 5.359442090988159,
      "epoch": 2.0187319884726227,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045943213696450475,
      "loss": 4.7186,
      "mean_token_accuracy": 0.24670082181692124,
      "num_tokens": 48187554.0,
      "step": 21015
    },
    {
      "entropy": 5.354343271255493,
      "epoch": 2.0192122958693566,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045941249736117023,
      "loss": 4.7653,
      "mean_token_accuracy": 0.2394870012998581,
      "num_tokens": 48199410.0,
      "step": 21020
    },
    {
      "entropy": 5.25310869216919,
      "epoch": 2.0196926032660905,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004593928534763251,
      "loss": 4.7054,
      "mean_token_accuracy": 0.2405994191765785,
      "num_tokens": 48211124.0,
      "step": 21025
    },
    {
      "entropy": 5.302400541305542,
      "epoch": 2.0201729106628243,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004593732053104254,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24769226610660552,
      "num_tokens": 48222690.0,
      "step": 21030
    },
    {
      "entropy": 5.3227440357208256,
      "epoch": 2.020653218059558,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045935355286392735,
      "loss": 4.692,
      "mean_token_accuracy": 0.24587682038545608,
      "num_tokens": 48235054.0,
      "step": 21035
    },
    {
      "entropy": 5.227099990844726,
      "epoch": 2.021133525456292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004593338961372873,
      "loss": 4.6481,
      "mean_token_accuracy": 0.2522473454475403,
      "num_tokens": 48246645.0,
      "step": 21040
    },
    {
      "entropy": 5.219429302215576,
      "epoch": 2.021613832853026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004593142351309614,
      "loss": 4.6589,
      "mean_token_accuracy": 0.24848906248807906,
      "num_tokens": 48258149.0,
      "step": 21045
    },
    {
      "entropy": 5.343129301071167,
      "epoch": 2.02209414024976,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004592945698454064,
      "loss": 4.8184,
      "mean_token_accuracy": 0.2378380000591278,
      "num_tokens": 48269892.0,
      "step": 21050
    },
    {
      "entropy": 5.368603992462158,
      "epoch": 2.0225744476464937,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045927490028107866,
      "loss": 4.7893,
      "mean_token_accuracy": 0.2382741764187813,
      "num_tokens": 48281087.0,
      "step": 21055
    },
    {
      "entropy": 5.287789630889892,
      "epoch": 2.0230547550432276,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000459255226438435,
      "loss": 4.6492,
      "mean_token_accuracy": 0.2495877206325531,
      "num_tokens": 48292177.0,
      "step": 21060
    },
    {
      "entropy": 5.2476142883300785,
      "epoch": 2.0235350624399615,
      "grad_norm": 1.03125,
      "learning_rate": 0.000459235548317932,
      "loss": 4.704,
      "mean_token_accuracy": 0.2388739615678787,
      "num_tokens": 48304281.0,
      "step": 21065
    },
    {
      "entropy": 5.2691630840301515,
      "epoch": 2.0240153698366954,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045921586592002667,
      "loss": 4.6901,
      "mean_token_accuracy": 0.24425181448459626,
      "num_tokens": 48316758.0,
      "step": 21070
    },
    {
      "entropy": 5.405640411376953,
      "epoch": 2.0244956772334293,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000459196179245176,
      "loss": 4.8163,
      "mean_token_accuracy": 0.23860495239496232,
      "num_tokens": 48327455.0,
      "step": 21075
    },
    {
      "entropy": 5.311354064941407,
      "epoch": 2.024975984630163,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004591764882938369,
      "loss": 4.7378,
      "mean_token_accuracy": 0.23716120421886444,
      "num_tokens": 48338105.0,
      "step": 21080
    },
    {
      "entropy": 5.349890232086182,
      "epoch": 2.025456292026897,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004591567930664667,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23427077680826186,
      "num_tokens": 48349599.0,
      "step": 21085
    },
    {
      "entropy": 5.3418221950531,
      "epoch": 2.025936599423631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004591370935635226,
      "loss": 4.8205,
      "mean_token_accuracy": 0.234921658039093,
      "num_tokens": 48361488.0,
      "step": 21090
    },
    {
      "entropy": 5.366274738311768,
      "epoch": 2.0264169068203652,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004591173897854619,
      "loss": 4.745,
      "mean_token_accuracy": 0.2376146659255028,
      "num_tokens": 48372049.0,
      "step": 21095
    },
    {
      "entropy": 5.301975584030151,
      "epoch": 2.026897214217099,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004590976817327422,
      "loss": 4.6484,
      "mean_token_accuracy": 0.24913933575153352,
      "num_tokens": 48385599.0,
      "step": 21100
    },
    {
      "entropy": 5.368116092681885,
      "epoch": 2.027377521613833,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004590779694058209,
      "loss": 4.7396,
      "mean_token_accuracy": 0.2433938726782799,
      "num_tokens": 48397390.0,
      "step": 21105
    },
    {
      "entropy": 5.311733675003052,
      "epoch": 2.027857829010567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045905825280515586,
      "loss": 4.7711,
      "mean_token_accuracy": 0.24027154743671417,
      "num_tokens": 48409837.0,
      "step": 21110
    },
    {
      "entropy": 5.255411577224732,
      "epoch": 2.0283381364073008,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045903853193120464,
      "loss": 4.6158,
      "mean_token_accuracy": 0.2580932885408401,
      "num_tokens": 48420765.0,
      "step": 21115
    },
    {
      "entropy": 5.345204782485962,
      "epoch": 2.0288184438040346,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045901880678442524,
      "loss": 4.747,
      "mean_token_accuracy": 0.2469532907009125,
      "num_tokens": 48431939.0,
      "step": 21120
    },
    {
      "entropy": 5.394588327407837,
      "epoch": 2.0292987512007685,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045899907736527556,
      "loss": 4.8114,
      "mean_token_accuracy": 0.2343030571937561,
      "num_tokens": 48444820.0,
      "step": 21125
    },
    {
      "entropy": 5.274328231811523,
      "epoch": 2.0297790585975024,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045897934367421364,
      "loss": 4.739,
      "mean_token_accuracy": 0.24343478977680205,
      "num_tokens": 48457315.0,
      "step": 21130
    },
    {
      "entropy": 5.382942962646484,
      "epoch": 2.0302593659942363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004589596057116977,
      "loss": 4.808,
      "mean_token_accuracy": 0.23130786418914795,
      "num_tokens": 48468612.0,
      "step": 21135
    },
    {
      "entropy": 5.309527015686035,
      "epoch": 2.03073967339097,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045893986347818593,
      "loss": 4.6674,
      "mean_token_accuracy": 0.24793170243501664,
      "num_tokens": 48479453.0,
      "step": 21140
    },
    {
      "entropy": 5.283021926879883,
      "epoch": 2.031219980787704,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004589201169741368,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24922099560499192,
      "num_tokens": 48491656.0,
      "step": 21145
    },
    {
      "entropy": 5.267356491088867,
      "epoch": 2.031700288184438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045890036620000856,
      "loss": 4.6569,
      "mean_token_accuracy": 0.238035349547863,
      "num_tokens": 48503431.0,
      "step": 21150
    },
    {
      "entropy": 5.289393186569214,
      "epoch": 2.032180595581172,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045888061115626,
      "loss": 4.6697,
      "mean_token_accuracy": 0.24742028415203093,
      "num_tokens": 48513922.0,
      "step": 21155
    },
    {
      "entropy": 5.216358947753906,
      "epoch": 2.0326609029779057,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004588608518433496,
      "loss": 4.6607,
      "mean_token_accuracy": 0.2515495553612709,
      "num_tokens": 48524715.0,
      "step": 21160
    },
    {
      "entropy": 5.337769508361816,
      "epoch": 2.0331412103746396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004588410882617362,
      "loss": 4.8121,
      "mean_token_accuracy": 0.23695180416107178,
      "num_tokens": 48536696.0,
      "step": 21165
    },
    {
      "entropy": 5.425889205932617,
      "epoch": 2.0336215177713735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004588213204118786,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23889956772327423,
      "num_tokens": 48550290.0,
      "step": 21170
    },
    {
      "entropy": 5.364410161972046,
      "epoch": 2.034101825168108,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045880154829423586,
      "loss": 4.7883,
      "mean_token_accuracy": 0.24275561720132827,
      "num_tokens": 48561047.0,
      "step": 21175
    },
    {
      "entropy": 5.175294351577759,
      "epoch": 2.0345821325648417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004587817719092668,
      "loss": 4.5638,
      "mean_token_accuracy": 0.25919976383447646,
      "num_tokens": 48571268.0,
      "step": 21180
    },
    {
      "entropy": 5.2678468227386475,
      "epoch": 2.0350624399615755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045876199125743087,
      "loss": 4.6938,
      "mean_token_accuracy": 0.2400224655866623,
      "num_tokens": 48583158.0,
      "step": 21185
    },
    {
      "entropy": 5.360071182250977,
      "epoch": 2.0355427473583094,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004587422063391871,
      "loss": 4.7585,
      "mean_token_accuracy": 0.24362047910690307,
      "num_tokens": 48593419.0,
      "step": 21190
    },
    {
      "entropy": 5.346278476715088,
      "epoch": 2.0360230547550433,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004587224171549949,
      "loss": 4.7444,
      "mean_token_accuracy": 0.2452643483877182,
      "num_tokens": 48605845.0,
      "step": 21195
    },
    {
      "entropy": 5.306459522247314,
      "epoch": 2.036503362151777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045870262370531376,
      "loss": 4.6872,
      "mean_token_accuracy": 0.24987302720546722,
      "num_tokens": 48616503.0,
      "step": 21200
    },
    {
      "entropy": 5.355581521987915,
      "epoch": 2.036983669548511,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045868282599060314,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23836749643087388,
      "num_tokens": 48628410.0,
      "step": 21205
    },
    {
      "entropy": 5.304331970214844,
      "epoch": 2.037463976945245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004586630240113227,
      "loss": 4.6807,
      "mean_token_accuracy": 0.24470422118902208,
      "num_tokens": 48639457.0,
      "step": 21210
    },
    {
      "entropy": 5.299585056304932,
      "epoch": 2.037944284341979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004586432177679322,
      "loss": 4.7457,
      "mean_token_accuracy": 0.2402294397354126,
      "num_tokens": 48652250.0,
      "step": 21215
    },
    {
      "entropy": 5.301463222503662,
      "epoch": 2.0384245917387127,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045862340726089153,
      "loss": 4.8067,
      "mean_token_accuracy": 0.2422426849603653,
      "num_tokens": 48663386.0,
      "step": 21220
    },
    {
      "entropy": 5.360460376739502,
      "epoch": 2.0389048991354466,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004586035924906606,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23357180804014205,
      "num_tokens": 48675111.0,
      "step": 21225
    },
    {
      "entropy": 5.279926729202271,
      "epoch": 2.0393852065321805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045858377345769946,
      "loss": 4.6853,
      "mean_token_accuracy": 0.24186577498912812,
      "num_tokens": 48686264.0,
      "step": 21230
    },
    {
      "entropy": 5.340615510940552,
      "epoch": 2.0398655139289144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004585639501624682,
      "loss": 4.7182,
      "mean_token_accuracy": 0.2385836884379387,
      "num_tokens": 48697249.0,
      "step": 21235
    },
    {
      "entropy": 5.287196111679077,
      "epoch": 2.0403458213256482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000458544122605427,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2454872265458107,
      "num_tokens": 48708808.0,
      "step": 21240
    },
    {
      "entropy": 5.384735584259033,
      "epoch": 2.040826128722382,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045852429078703646,
      "loss": 4.8074,
      "mean_token_accuracy": 0.23957061767578125,
      "num_tokens": 48719879.0,
      "step": 21245
    },
    {
      "entropy": 5.353318500518799,
      "epoch": 2.0413064361191164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045850445470775673,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23976994007825853,
      "num_tokens": 48731110.0,
      "step": 21250
    },
    {
      "entropy": 5.291004705429077,
      "epoch": 2.0417867435158503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004584846143680485,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24754922986030578,
      "num_tokens": 48742799.0,
      "step": 21255
    },
    {
      "entropy": 5.426052141189575,
      "epoch": 2.042267050912584,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004584647697683723,
      "loss": 4.8631,
      "mean_token_accuracy": 0.23528711646795272,
      "num_tokens": 48753773.0,
      "step": 21260
    },
    {
      "entropy": 5.338893222808838,
      "epoch": 2.042747358309318,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045844492090918904,
      "loss": 4.7205,
      "mean_token_accuracy": 0.24674683511257173,
      "num_tokens": 48764561.0,
      "step": 21265
    },
    {
      "entropy": 5.260461950302124,
      "epoch": 2.043227665706052,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045842506779095936,
      "loss": 4.6103,
      "mean_token_accuracy": 0.2444976270198822,
      "num_tokens": 48775319.0,
      "step": 21270
    },
    {
      "entropy": 5.399407052993775,
      "epoch": 2.043707973102786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004584052104141443,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22709066569805145,
      "num_tokens": 48787595.0,
      "step": 21275
    },
    {
      "entropy": 5.395428991317749,
      "epoch": 2.0441882804995197,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004583853487792048,
      "loss": 4.7194,
      "mean_token_accuracy": 0.24202266037464143,
      "num_tokens": 48800015.0,
      "step": 21280
    },
    {
      "entropy": 5.373798942565918,
      "epoch": 2.0446685878962536,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004583654828866021,
      "loss": 4.765,
      "mean_token_accuracy": 0.23735410571098328,
      "num_tokens": 48811651.0,
      "step": 21285
    },
    {
      "entropy": 5.367016077041626,
      "epoch": 2.0451488952929875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004583456127367973,
      "loss": 4.7655,
      "mean_token_accuracy": 0.2426288455724716,
      "num_tokens": 48823561.0,
      "step": 21290
    },
    {
      "entropy": 5.341028547286987,
      "epoch": 2.0456292026897214,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583257383302519,
      "loss": 4.8002,
      "mean_token_accuracy": 0.2344923883676529,
      "num_tokens": 48835385.0,
      "step": 21295
    },
    {
      "entropy": 5.383321094512939,
      "epoch": 2.0461095100864553,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004583058596674271,
      "loss": 4.802,
      "mean_token_accuracy": 0.23687067180871962,
      "num_tokens": 48846899.0,
      "step": 21300
    },
    {
      "entropy": 5.3580550193786625,
      "epoch": 2.046589817483189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004582859767487846,
      "loss": 4.7315,
      "mean_token_accuracy": 0.24596863090991974,
      "num_tokens": 48857850.0,
      "step": 21305
    },
    {
      "entropy": 5.28949990272522,
      "epoch": 2.047070124879923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045826608957478604,
      "loss": 4.7237,
      "mean_token_accuracy": 0.2422050192952156,
      "num_tokens": 48869192.0,
      "step": 21310
    },
    {
      "entropy": 5.283545064926147,
      "epoch": 2.047550432276657,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045824619814589297,
      "loss": 4.6974,
      "mean_token_accuracy": 0.2484436437487602,
      "num_tokens": 48881314.0,
      "step": 21315
    },
    {
      "entropy": 5.269960355758667,
      "epoch": 2.0480307396733908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004582263024625674,
      "loss": 4.7764,
      "mean_token_accuracy": 0.23788559138774873,
      "num_tokens": 48893292.0,
      "step": 21320
    },
    {
      "entropy": 5.314723777770996,
      "epoch": 2.048511047070125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004582064025252711,
      "loss": 4.8292,
      "mean_token_accuracy": 0.2360629543662071,
      "num_tokens": 48905607.0,
      "step": 21325
    },
    {
      "entropy": 5.360835886001587,
      "epoch": 2.048991354466859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004581864983344661,
      "loss": 4.7031,
      "mean_token_accuracy": 0.24590568840503693,
      "num_tokens": 48916194.0,
      "step": 21330
    },
    {
      "entropy": 5.330028629302978,
      "epoch": 2.049471661863593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004581665898906147,
      "loss": 4.703,
      "mean_token_accuracy": 0.2441118463873863,
      "num_tokens": 48927560.0,
      "step": 21335
    },
    {
      "entropy": 5.265310335159302,
      "epoch": 2.0499519692603267,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045814667719417887,
      "loss": 4.7116,
      "mean_token_accuracy": 0.24616371095180511,
      "num_tokens": 48938284.0,
      "step": 21340
    },
    {
      "entropy": 5.335558319091797,
      "epoch": 2.0504322766570606,
      "grad_norm": 1.0625,
      "learning_rate": 0.000458126760245621,
      "loss": 4.7089,
      "mean_token_accuracy": 0.2501352369785309,
      "num_tokens": 48948325.0,
      "step": 21345
    },
    {
      "entropy": 5.31817398071289,
      "epoch": 2.0509125840537945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004581068390454036,
      "loss": 4.7227,
      "mean_token_accuracy": 0.24087598621845246,
      "num_tokens": 48960020.0,
      "step": 21350
    },
    {
      "entropy": 5.3071671485900875,
      "epoch": 2.0513928914505284,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045808691359398905,
      "loss": 4.6787,
      "mean_token_accuracy": 0.24256464540958406,
      "num_tokens": 48970833.0,
      "step": 21355
    },
    {
      "entropy": 5.311675310134888,
      "epoch": 2.0518731988472623,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004580669838918401,
      "loss": 4.7395,
      "mean_token_accuracy": 0.24758011549711229,
      "num_tokens": 48981821.0,
      "step": 21360
    },
    {
      "entropy": 5.3756999492645265,
      "epoch": 2.052353506243996,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045804704993941935,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24109505712985993,
      "num_tokens": 48992341.0,
      "step": 21365
    },
    {
      "entropy": 5.355353164672851,
      "epoch": 2.05283381364073,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045802711173718966,
      "loss": 4.8292,
      "mean_token_accuracy": 0.23630041778087615,
      "num_tokens": 49003063.0,
      "step": 21370
    },
    {
      "entropy": 5.325448608398437,
      "epoch": 2.053314121037464,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004580071692856138,
      "loss": 4.8062,
      "mean_token_accuracy": 0.2414543867111206,
      "num_tokens": 49015353.0,
      "step": 21375
    },
    {
      "entropy": 5.2637059688568115,
      "epoch": 2.053794428434198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045798722258515504,
      "loss": 4.6886,
      "mean_token_accuracy": 0.24662078469991683,
      "num_tokens": 49026405.0,
      "step": 21380
    },
    {
      "entropy": 5.365761852264404,
      "epoch": 2.0542747358309317,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045796727163627623,
      "loss": 4.7749,
      "mean_token_accuracy": 0.23906690627336502,
      "num_tokens": 49038433.0,
      "step": 21385
    },
    {
      "entropy": 5.283199882507324,
      "epoch": 2.0547550432276656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004579473164394408,
      "loss": 4.6359,
      "mean_token_accuracy": 0.2549751684069633,
      "num_tokens": 49049463.0,
      "step": 21390
    },
    {
      "entropy": 5.208337306976318,
      "epoch": 2.0552353506243994,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045792735699511176,
      "loss": 4.6177,
      "mean_token_accuracy": 0.25169252008199694,
      "num_tokens": 49061607.0,
      "step": 21395
    },
    {
      "entropy": 5.328116464614868,
      "epoch": 2.0557156580211333,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045790739330375276,
      "loss": 4.7727,
      "mean_token_accuracy": 0.2363843634724617,
      "num_tokens": 49074687.0,
      "step": 21400
    },
    {
      "entropy": 5.285823631286621,
      "epoch": 2.0561959654178676,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045788742536582717,
      "loss": 4.6514,
      "mean_token_accuracy": 0.2467433214187622,
      "num_tokens": 49086151.0,
      "step": 21405
    },
    {
      "entropy": 5.316674661636353,
      "epoch": 2.0566762728146015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045786745318179866,
      "loss": 4.7659,
      "mean_token_accuracy": 0.23342252224683763,
      "num_tokens": 49097418.0,
      "step": 21410
    },
    {
      "entropy": 5.241256380081177,
      "epoch": 2.0571565802113354,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004578474767521309,
      "loss": 4.6894,
      "mean_token_accuracy": 0.25138217508792876,
      "num_tokens": 49108570.0,
      "step": 21415
    },
    {
      "entropy": 5.244969081878662,
      "epoch": 2.0576368876080693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045782749607728765,
      "loss": 4.7445,
      "mean_token_accuracy": 0.23753925710916518,
      "num_tokens": 49120906.0,
      "step": 21420
    },
    {
      "entropy": 5.301382112503052,
      "epoch": 2.058117195004803,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045780751115773286,
      "loss": 4.6807,
      "mean_token_accuracy": 0.2453414171934128,
      "num_tokens": 49132040.0,
      "step": 21425
    },
    {
      "entropy": 5.332814264297485,
      "epoch": 2.058597502401537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004577875219939304,
      "loss": 4.7771,
      "mean_token_accuracy": 0.24650634974241256,
      "num_tokens": 49143427.0,
      "step": 21430
    },
    {
      "entropy": 5.3474236011505125,
      "epoch": 2.059077809798271,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004577675285863446,
      "loss": 4.7555,
      "mean_token_accuracy": 0.23756640255451203,
      "num_tokens": 49155351.0,
      "step": 21435
    },
    {
      "entropy": 5.359217023849487,
      "epoch": 2.059558117195005,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045774753093543943,
      "loss": 4.753,
      "mean_token_accuracy": 0.24494047611951827,
      "num_tokens": 49167898.0,
      "step": 21440
    },
    {
      "entropy": 5.29768385887146,
      "epoch": 2.0600384245917387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004577275290416791,
      "loss": 4.7407,
      "mean_token_accuracy": 0.2360730439424515,
      "num_tokens": 49179209.0,
      "step": 21445
    },
    {
      "entropy": 5.282809448242188,
      "epoch": 2.0605187319884726,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004577075229055283,
      "loss": 4.6786,
      "mean_token_accuracy": 0.2484931081533432,
      "num_tokens": 49190382.0,
      "step": 21450
    },
    {
      "entropy": 5.274778604507446,
      "epoch": 2.0609990393852065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045768751252745133,
      "loss": 4.7281,
      "mean_token_accuracy": 0.24674810469150543,
      "num_tokens": 49203511.0,
      "step": 21455
    },
    {
      "entropy": 5.3917989253997805,
      "epoch": 2.0614793467819403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045766749790791274,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2385928988456726,
      "num_tokens": 49214276.0,
      "step": 21460
    },
    {
      "entropy": 5.198894691467285,
      "epoch": 2.061959654178674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004576474790473773,
      "loss": 4.6164,
      "mean_token_accuracy": 0.25231376886367796,
      "num_tokens": 49225803.0,
      "step": 21465
    },
    {
      "entropy": 5.283808183670044,
      "epoch": 2.062439961575408,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045762745594630973,
      "loss": 4.7546,
      "mean_token_accuracy": 0.23765883445739747,
      "num_tokens": 49237187.0,
      "step": 21470
    },
    {
      "entropy": 5.40148777961731,
      "epoch": 2.062920268972142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004576074286051749,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23701718598604202,
      "num_tokens": 49250151.0,
      "step": 21475
    },
    {
      "entropy": 5.279868125915527,
      "epoch": 2.063400576368876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045758739702443787,
      "loss": 4.6445,
      "mean_token_accuracy": 0.2568434551358223,
      "num_tokens": 49261970.0,
      "step": 21480
    },
    {
      "entropy": 5.291755723953247,
      "epoch": 2.06388088376561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004575673612045636,
      "loss": 4.6999,
      "mean_token_accuracy": 0.25308124274015426,
      "num_tokens": 49271427.0,
      "step": 21485
    },
    {
      "entropy": 5.275437450408935,
      "epoch": 2.064361191162344,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004575473211460173,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24215928763151168,
      "num_tokens": 49285102.0,
      "step": 21490
    },
    {
      "entropy": 5.327164888381958,
      "epoch": 2.064841498559078,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575272768492644,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24726642519235612,
      "num_tokens": 49297569.0,
      "step": 21495
    },
    {
      "entropy": 5.306490087509156,
      "epoch": 2.065321805955812,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045750722831476993,
      "loss": 4.6785,
      "mean_token_accuracy": 0.24789609313011168,
      "num_tokens": 49308795.0,
      "step": 21500
    },
    {
      "entropy": 5.3558244705200195,
      "epoch": 2.0658021133525457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045748717554299964,
      "loss": 4.7844,
      "mean_token_accuracy": 0.2416103810071945,
      "num_tokens": 49320971.0,
      "step": 21505
    },
    {
      "entropy": 5.387311840057373,
      "epoch": 2.0662824207492796,
      "grad_norm": 1.0,
      "learning_rate": 0.000457467118534419,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2331244617700577,
      "num_tokens": 49333624.0,
      "step": 21510
    },
    {
      "entropy": 5.415111970901489,
      "epoch": 2.0667627281460135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004574470572894938,
      "loss": 4.8561,
      "mean_token_accuracy": 0.23462607860565185,
      "num_tokens": 49344863.0,
      "step": 21515
    },
    {
      "entropy": 5.336499834060669,
      "epoch": 2.0672430355427474,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004574269918086895,
      "loss": 4.719,
      "mean_token_accuracy": 0.23774611651897432,
      "num_tokens": 49355501.0,
      "step": 21520
    },
    {
      "entropy": 5.317561864852905,
      "epoch": 2.0677233429394812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004574069220924722,
      "loss": 4.7701,
      "mean_token_accuracy": 0.25102677196264267,
      "num_tokens": 49366384.0,
      "step": 21525
    },
    {
      "entropy": 5.399748945236206,
      "epoch": 2.068203650336215,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004573868481413079,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22380245178937913,
      "num_tokens": 49378670.0,
      "step": 21530
    },
    {
      "entropy": 5.318468332290649,
      "epoch": 2.068683957732949,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045736676995566244,
      "loss": 4.7004,
      "mean_token_accuracy": 0.2417692095041275,
      "num_tokens": 49390273.0,
      "step": 21535
    },
    {
      "entropy": 5.37339186668396,
      "epoch": 2.069164265129683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045734668753600217,
      "loss": 4.7705,
      "mean_token_accuracy": 0.24154685735702514,
      "num_tokens": 49402503.0,
      "step": 21540
    },
    {
      "entropy": 5.328623485565186,
      "epoch": 2.0696445725264168,
      "grad_norm": 1.0,
      "learning_rate": 0.00045732660088279326,
      "loss": 4.7273,
      "mean_token_accuracy": 0.23728937208652495,
      "num_tokens": 49413948.0,
      "step": 21545
    },
    {
      "entropy": 5.2827167987823485,
      "epoch": 2.0701248799231506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045730650999650216,
      "loss": 4.6881,
      "mean_token_accuracy": 0.25203169137239456,
      "num_tokens": 49424164.0,
      "step": 21550
    },
    {
      "entropy": 5.394902896881104,
      "epoch": 2.0706051873198845,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045728641487759506,
      "loss": 4.8826,
      "mean_token_accuracy": 0.2293961152434349,
      "num_tokens": 49435102.0,
      "step": 21555
    },
    {
      "entropy": 5.278410530090332,
      "epoch": 2.071085494716619,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004572663155265388,
      "loss": 4.7212,
      "mean_token_accuracy": 0.2475061759352684,
      "num_tokens": 49445445.0,
      "step": 21560
    },
    {
      "entropy": 5.276941013336182,
      "epoch": 2.0715658021133527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004572462119437999,
      "loss": 4.7046,
      "mean_token_accuracy": 0.24475764930248262,
      "num_tokens": 49457578.0,
      "step": 21565
    },
    {
      "entropy": 5.261423826217651,
      "epoch": 2.0720461095100866,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045722610412984513,
      "loss": 4.6898,
      "mean_token_accuracy": 0.24997369050979615,
      "num_tokens": 49470253.0,
      "step": 21570
    },
    {
      "entropy": 5.379653596878052,
      "epoch": 2.0725264169068205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004572059920851412,
      "loss": 4.784,
      "mean_token_accuracy": 0.23964751958847047,
      "num_tokens": 49481854.0,
      "step": 21575
    },
    {
      "entropy": 5.380216646194458,
      "epoch": 2.0730067243035544,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045718587581015534,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23289130628108978,
      "num_tokens": 49494311.0,
      "step": 21580
    },
    {
      "entropy": 5.331872272491455,
      "epoch": 2.0734870317002883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004571657553053544,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2488353118300438,
      "num_tokens": 49506186.0,
      "step": 21585
    },
    {
      "entropy": 5.322967052459717,
      "epoch": 2.073967339097022,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004571456305712055,
      "loss": 4.7305,
      "mean_token_accuracy": 0.23816418945789336,
      "num_tokens": 49517841.0,
      "step": 21590
    },
    {
      "entropy": 5.320898580551147,
      "epoch": 2.074447646493756,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004571255016081759,
      "loss": 4.7039,
      "mean_token_accuracy": 0.24262737184762956,
      "num_tokens": 49529260.0,
      "step": 21595
    },
    {
      "entropy": 5.3038108348846436,
      "epoch": 2.07492795389049,
      "grad_norm": 1.03125,
      "learning_rate": 0.000457105368416733,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2514826342463493,
      "num_tokens": 49540892.0,
      "step": 21600
    },
    {
      "entropy": 5.409717082977295,
      "epoch": 2.0754082612872238,
      "grad_norm": 1.125,
      "learning_rate": 0.00045708523099734417,
      "loss": 4.7532,
      "mean_token_accuracy": 0.24337019324302672,
      "num_tokens": 49550924.0,
      "step": 21605
    },
    {
      "entropy": 5.334637546539307,
      "epoch": 2.0758885686839577,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045706508935047693,
      "loss": 4.7748,
      "mean_token_accuracy": 0.24229359477758408,
      "num_tokens": 49562424.0,
      "step": 21610
    },
    {
      "entropy": 5.271863746643066,
      "epoch": 2.0763688760806915,
      "grad_norm": 1.140625,
      "learning_rate": 0.000457044943476599,
      "loss": 4.7109,
      "mean_token_accuracy": 0.24292296767234803,
      "num_tokens": 49573341.0,
      "step": 21615
    },
    {
      "entropy": 5.342797040939331,
      "epoch": 2.0768491834774254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045702479337617795,
      "loss": 4.6831,
      "mean_token_accuracy": 0.24310685098171234,
      "num_tokens": 49584047.0,
      "step": 21620
    },
    {
      "entropy": 5.272264766693115,
      "epoch": 2.0773294908741593,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004570046390496818,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24452677369117737,
      "num_tokens": 49595898.0,
      "step": 21625
    },
    {
      "entropy": 5.28755292892456,
      "epoch": 2.077809798270893,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004569844804975783,
      "loss": 4.7563,
      "mean_token_accuracy": 0.23644920885562898,
      "num_tokens": 49607865.0,
      "step": 21630
    },
    {
      "entropy": 5.396686744689942,
      "epoch": 2.0782901056676275,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004569643177203356,
      "loss": 4.8078,
      "mean_token_accuracy": 0.2407008543610573,
      "num_tokens": 49619513.0,
      "step": 21635
    },
    {
      "entropy": 5.316512727737427,
      "epoch": 2.0787704130643614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004569441507184218,
      "loss": 4.6967,
      "mean_token_accuracy": 0.25257029980421064,
      "num_tokens": 49631232.0,
      "step": 21640
    },
    {
      "entropy": 5.304959392547607,
      "epoch": 2.0792507204610953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045692397949230495,
      "loss": 4.7175,
      "mean_token_accuracy": 0.242829567193985,
      "num_tokens": 49642653.0,
      "step": 21645
    },
    {
      "entropy": 5.273534059524536,
      "epoch": 2.079731027857829,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045690380404245364,
      "loss": 4.6799,
      "mean_token_accuracy": 0.24741346389055252,
      "num_tokens": 49654479.0,
      "step": 21650
    },
    {
      "entropy": 5.295475196838379,
      "epoch": 2.080211335254563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045688362436933607,
      "loss": 4.7532,
      "mean_token_accuracy": 0.238503135740757,
      "num_tokens": 49666461.0,
      "step": 21655
    },
    {
      "entropy": 5.322782945632935,
      "epoch": 2.080691642651297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004568634404734208,
      "loss": 4.7964,
      "mean_token_accuracy": 0.2434151589870453,
      "num_tokens": 49676602.0,
      "step": 21660
    },
    {
      "entropy": 5.250767612457276,
      "epoch": 2.081171950048031,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004568432523551765,
      "loss": 4.7124,
      "mean_token_accuracy": 0.24574712216854094,
      "num_tokens": 49689225.0,
      "step": 21665
    },
    {
      "entropy": 5.3000946044921875,
      "epoch": 2.0816522574447647,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004568230600150718,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24471505135297775,
      "num_tokens": 49700827.0,
      "step": 21670
    },
    {
      "entropy": 5.355616617202759,
      "epoch": 2.0821325648414986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004568028634535757,
      "loss": 4.7729,
      "mean_token_accuracy": 0.24107464402914047,
      "num_tokens": 49712777.0,
      "step": 21675
    },
    {
      "entropy": 5.312104749679565,
      "epoch": 2.0826128722382324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004567826626711568,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24697034060955048,
      "num_tokens": 49725234.0,
      "step": 21680
    },
    {
      "entropy": 5.303418588638306,
      "epoch": 2.0830931796349663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004567624576682843,
      "loss": 4.6858,
      "mean_token_accuracy": 0.24749669730663298,
      "num_tokens": 49736933.0,
      "step": 21685
    },
    {
      "entropy": 5.414926767349243,
      "epoch": 2.0835734870317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004567422484454273,
      "loss": 4.778,
      "mean_token_accuracy": 0.24357067793607712,
      "num_tokens": 49748279.0,
      "step": 21690
    },
    {
      "entropy": 5.259363031387329,
      "epoch": 2.084053794428434,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045672203500305493,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2552573338150978,
      "num_tokens": 49760874.0,
      "step": 21695
    },
    {
      "entropy": 5.254287004470825,
      "epoch": 2.084534101825168,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045670181734163654,
      "loss": 4.6266,
      "mean_token_accuracy": 0.24979099482297898,
      "num_tokens": 49772520.0,
      "step": 21700
    },
    {
      "entropy": 5.287472820281982,
      "epoch": 2.085014409221902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004566815954616414,
      "loss": 4.698,
      "mean_token_accuracy": 0.24709252268075943,
      "num_tokens": 49783019.0,
      "step": 21705
    },
    {
      "entropy": 5.198334980010986,
      "epoch": 2.0854947166186357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045666136936353913,
      "loss": 4.7004,
      "mean_token_accuracy": 0.24601924270391465,
      "num_tokens": 49795111.0,
      "step": 21710
    },
    {
      "entropy": 5.231745910644531,
      "epoch": 2.08597502401537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004566411390477993,
      "loss": 4.6359,
      "mean_token_accuracy": 0.25601002126932143,
      "num_tokens": 49807022.0,
      "step": 21715
    },
    {
      "entropy": 5.36622462272644,
      "epoch": 2.086455331412104,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045662090451489156,
      "loss": 4.7684,
      "mean_token_accuracy": 0.23436764031648635,
      "num_tokens": 49819225.0,
      "step": 21720
    },
    {
      "entropy": 5.323865365982056,
      "epoch": 2.086935638808838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045660066576528577,
      "loss": 4.7153,
      "mean_token_accuracy": 0.2490009605884552,
      "num_tokens": 49830989.0,
      "step": 21725
    },
    {
      "entropy": 5.321580982208252,
      "epoch": 2.0874159462055717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004565804227994518,
      "loss": 4.699,
      "mean_token_accuracy": 0.24515070617198945,
      "num_tokens": 49841618.0,
      "step": 21730
    },
    {
      "entropy": 5.271538257598877,
      "epoch": 2.0878962536023056,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004565601756178595,
      "loss": 4.5882,
      "mean_token_accuracy": 0.25730434358119963,
      "num_tokens": 49851985.0,
      "step": 21735
    },
    {
      "entropy": 5.311316061019897,
      "epoch": 2.0883765609990395,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565399242209791,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23522545546293258,
      "num_tokens": 49863383.0,
      "step": 21740
    },
    {
      "entropy": 5.3112071514129635,
      "epoch": 2.0888568683957733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004565196686092807,
      "loss": 4.7199,
      "mean_token_accuracy": 0.2454364776611328,
      "num_tokens": 49875757.0,
      "step": 21745
    },
    {
      "entropy": 5.296944713592529,
      "epoch": 2.089337175792507,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004564994087832346,
      "loss": 4.7033,
      "mean_token_accuracy": 0.25299854427576063,
      "num_tokens": 49888419.0,
      "step": 21750
    },
    {
      "entropy": 5.320362281799317,
      "epoch": 2.089817483189241,
      "grad_norm": 1.125,
      "learning_rate": 0.00045647914474331123,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23809603750705718,
      "num_tokens": 49899629.0,
      "step": 21755
    },
    {
      "entropy": 5.428266000747681,
      "epoch": 2.090297790585975,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045645887648998094,
      "loss": 4.7679,
      "mean_token_accuracy": 0.2443296104669571,
      "num_tokens": 49909858.0,
      "step": 21760
    },
    {
      "entropy": 5.297444915771484,
      "epoch": 2.090778097982709,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045643860402371433,
      "loss": 4.7478,
      "mean_token_accuracy": 0.2407670259475708,
      "num_tokens": 49921305.0,
      "step": 21765
    },
    {
      "entropy": 5.296270179748535,
      "epoch": 2.0912584053794427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004564183273449821,
      "loss": 4.7567,
      "mean_token_accuracy": 0.24326994270086288,
      "num_tokens": 49931966.0,
      "step": 21770
    },
    {
      "entropy": 5.273635911941528,
      "epoch": 2.0917387127761766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004563980464542551,
      "loss": 4.6755,
      "mean_token_accuracy": 0.2435280427336693,
      "num_tokens": 49943781.0,
      "step": 21775
    },
    {
      "entropy": 5.3576788902282715,
      "epoch": 2.0922190201729105,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045637776135200406,
      "loss": 4.7329,
      "mean_token_accuracy": 0.2423996701836586,
      "num_tokens": 49954149.0,
      "step": 21780
    },
    {
      "entropy": 5.399283313751221,
      "epoch": 2.0926993275696444,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004563574720386999,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23908154368400575,
      "num_tokens": 49965542.0,
      "step": 21785
    },
    {
      "entropy": 5.329218530654908,
      "epoch": 2.0931796349663783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004563371785148139,
      "loss": 4.7245,
      "mean_token_accuracy": 0.24103086143732072,
      "num_tokens": 49977444.0,
      "step": 21790
    },
    {
      "entropy": 5.271674966812133,
      "epoch": 2.0936599423631126,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045631688078081695,
      "loss": 4.6992,
      "mean_token_accuracy": 0.24547140300273895,
      "num_tokens": 49988409.0,
      "step": 21795
    },
    {
      "entropy": 5.321722984313965,
      "epoch": 2.0941402497598465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004562965788371805,
      "loss": 4.7542,
      "mean_token_accuracy": 0.24057189524173736,
      "num_tokens": 49999639.0,
      "step": 21800
    },
    {
      "entropy": 5.332808113098144,
      "epoch": 2.0946205571565804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004562762726843758,
      "loss": 4.6997,
      "mean_token_accuracy": 0.24382775872945786,
      "num_tokens": 50010964.0,
      "step": 21805
    },
    {
      "entropy": 5.417371082305908,
      "epoch": 2.0951008645533142,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045625596232287436,
      "loss": 4.8303,
      "mean_token_accuracy": 0.2280905544757843,
      "num_tokens": 50024487.0,
      "step": 21810
    },
    {
      "entropy": 5.383966159820557,
      "epoch": 2.095581171950048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004562356477531477,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2298755317926407,
      "num_tokens": 50037202.0,
      "step": 21815
    },
    {
      "entropy": 5.309485816955567,
      "epoch": 2.096061479346782,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004562153289756674,
      "loss": 4.6444,
      "mean_token_accuracy": 0.2515628680586815,
      "num_tokens": 50048520.0,
      "step": 21820
    },
    {
      "entropy": 5.402780151367187,
      "epoch": 2.096541786743516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004561950059909053,
      "loss": 4.7833,
      "mean_token_accuracy": 0.23666412085294725,
      "num_tokens": 50058523.0,
      "step": 21825
    },
    {
      "entropy": 5.353758430480957,
      "epoch": 2.0970220941402498,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004561746787993332,
      "loss": 4.7597,
      "mean_token_accuracy": 0.23837963193655015,
      "num_tokens": 50070241.0,
      "step": 21830
    },
    {
      "entropy": 5.317731142044067,
      "epoch": 2.0975024015369836,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045615434740142307,
      "loss": 4.7632,
      "mean_token_accuracy": 0.24693377017974855,
      "num_tokens": 50081326.0,
      "step": 21835
    },
    {
      "entropy": 5.2774937629699705,
      "epoch": 2.0979827089337175,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045613401179764686,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2458815634250641,
      "num_tokens": 50091767.0,
      "step": 21840
    },
    {
      "entropy": 5.333650684356689,
      "epoch": 2.0984630163304514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045611367198847676,
      "loss": 4.7691,
      "mean_token_accuracy": 0.2388371303677559,
      "num_tokens": 50103435.0,
      "step": 21845
    },
    {
      "entropy": 5.371950006484985,
      "epoch": 2.0989433237271853,
      "grad_norm": 1.109375,
      "learning_rate": 0.000456093327974385,
      "loss": 4.7227,
      "mean_token_accuracy": 0.2457210600376129,
      "num_tokens": 50114127.0,
      "step": 21850
    },
    {
      "entropy": 5.332404756546021,
      "epoch": 2.099423631123919,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004560729797558438,
      "loss": 4.7451,
      "mean_token_accuracy": 0.24546629190444946,
      "num_tokens": 50126328.0,
      "step": 21855
    },
    {
      "entropy": 5.342629432678223,
      "epoch": 2.099903938520653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004560526273333259,
      "loss": 4.7094,
      "mean_token_accuracy": 0.24636364877223968,
      "num_tokens": 50137600.0,
      "step": 21860
    },
    {
      "entropy": 5.361260843276978,
      "epoch": 2.100384245917387,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045603227070730346,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23859059661626816,
      "num_tokens": 50148711.0,
      "step": 21865
    },
    {
      "entropy": 5.304969596862793,
      "epoch": 2.1008645533141213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045601190987824933,
      "loss": 4.7485,
      "mean_token_accuracy": 0.23380789607763292,
      "num_tokens": 50161141.0,
      "step": 21870
    },
    {
      "entropy": 5.366174554824829,
      "epoch": 2.101344860710855,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045599154484663606,
      "loss": 4.7693,
      "mean_token_accuracy": 0.23926853239536286,
      "num_tokens": 50173145.0,
      "step": 21875
    },
    {
      "entropy": 5.323134469985962,
      "epoch": 2.101825168107589,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045597117561293663,
      "loss": 4.7305,
      "mean_token_accuracy": 0.24546944499015808,
      "num_tokens": 50184074.0,
      "step": 21880
    },
    {
      "entropy": 5.494198322296143,
      "epoch": 2.102305475504323,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004559508021776238,
      "loss": 4.9819,
      "mean_token_accuracy": 0.22796723991632462,
      "num_tokens": 50195004.0,
      "step": 21885
    },
    {
      "entropy": 5.3084571838378904,
      "epoch": 2.1027857829010568,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004559304245411707,
      "loss": 4.7062,
      "mean_token_accuracy": 0.24657197743654252,
      "num_tokens": 50205329.0,
      "step": 21890
    },
    {
      "entropy": 5.369451856613159,
      "epoch": 2.1032660902977907,
      "grad_norm": 1.0,
      "learning_rate": 0.00045591004270405044,
      "loss": 4.7543,
      "mean_token_accuracy": 0.2338891878724098,
      "num_tokens": 50218346.0,
      "step": 21895
    },
    {
      "entropy": 5.310058069229126,
      "epoch": 2.1037463976945245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004558896566667361,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24483548253774642,
      "num_tokens": 50229790.0,
      "step": 21900
    },
    {
      "entropy": 5.433704662322998,
      "epoch": 2.1042267050912584,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045586926642970113,
      "loss": 4.8763,
      "mean_token_accuracy": 0.24015165865421295,
      "num_tokens": 50240547.0,
      "step": 21905
    },
    {
      "entropy": 5.329795026779175,
      "epoch": 2.1047070124879923,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004558488719934188,
      "loss": 4.703,
      "mean_token_accuracy": 0.24876710772514343,
      "num_tokens": 50251889.0,
      "step": 21910
    },
    {
      "entropy": 5.280230760574341,
      "epoch": 2.105187319884726,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004558284733583627,
      "loss": 4.7661,
      "mean_token_accuracy": 0.23874947279691697,
      "num_tokens": 50263599.0,
      "step": 21915
    },
    {
      "entropy": 5.316542387008667,
      "epoch": 2.10566762728146,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045580807052500645,
      "loss": 4.8001,
      "mean_token_accuracy": 0.2395392581820488,
      "num_tokens": 50275645.0,
      "step": 21920
    },
    {
      "entropy": 5.281728744506836,
      "epoch": 2.106147934678194,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004557876634938236,
      "loss": 4.6839,
      "mean_token_accuracy": 0.2429250657558441,
      "num_tokens": 50287640.0,
      "step": 21925
    },
    {
      "entropy": 5.341038608551026,
      "epoch": 2.106628242074928,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004557672522652881,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23695984929800035,
      "num_tokens": 50299119.0,
      "step": 21930
    },
    {
      "entropy": 5.3307679176330565,
      "epoch": 2.1071085494716617,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004557468368398738,
      "loss": 4.7168,
      "mean_token_accuracy": 0.24152487218379975,
      "num_tokens": 50311377.0,
      "step": 21935
    },
    {
      "entropy": 5.332127666473388,
      "epoch": 2.1075888568683956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004557264172180546,
      "loss": 4.7671,
      "mean_token_accuracy": 0.24172416478395461,
      "num_tokens": 50323185.0,
      "step": 21940
    },
    {
      "entropy": 5.335532522201538,
      "epoch": 2.1080691642651295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004557059934003046,
      "loss": 4.8298,
      "mean_token_accuracy": 0.24292844533920288,
      "num_tokens": 50333992.0,
      "step": 21945
    },
    {
      "entropy": 5.412720441818237,
      "epoch": 2.108549471661864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004556855653870981,
      "loss": 4.79,
      "mean_token_accuracy": 0.2387074738740921,
      "num_tokens": 50346087.0,
      "step": 21950
    },
    {
      "entropy": 5.393613958358765,
      "epoch": 2.1090297790585977,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004556651331789092,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23978904634714127,
      "num_tokens": 50358274.0,
      "step": 21955
    },
    {
      "entropy": 5.3332517623901365,
      "epoch": 2.1095100864553316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004556446967762125,
      "loss": 4.7211,
      "mean_token_accuracy": 0.247772878408432,
      "num_tokens": 50367784.0,
      "step": 21960
    },
    {
      "entropy": 5.3785426139831545,
      "epoch": 2.1099903938520654,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045562425617948226,
      "loss": 4.8101,
      "mean_token_accuracy": 0.23705829083919525,
      "num_tokens": 50377877.0,
      "step": 21965
    },
    {
      "entropy": 5.352587699890137,
      "epoch": 2.1104707012487993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045560381138919315,
      "loss": 4.7967,
      "mean_token_accuracy": 0.24257465302944184,
      "num_tokens": 50391023.0,
      "step": 21970
    },
    {
      "entropy": 5.358722591400147,
      "epoch": 2.110951008645533,
      "grad_norm": 1.125,
      "learning_rate": 0.00045558336240581984,
      "loss": 4.7575,
      "mean_token_accuracy": 0.23682621270418167,
      "num_tokens": 50404417.0,
      "step": 21975
    },
    {
      "entropy": 5.349806451797486,
      "epoch": 2.111431316042267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045556290922983705,
      "loss": 4.7468,
      "mean_token_accuracy": 0.24059578478336335,
      "num_tokens": 50416153.0,
      "step": 21980
    },
    {
      "entropy": 5.304747581481934,
      "epoch": 2.111911623439001,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004555424518617197,
      "loss": 4.7203,
      "mean_token_accuracy": 0.24766730815172194,
      "num_tokens": 50426860.0,
      "step": 21985
    },
    {
      "entropy": 5.287749242782593,
      "epoch": 2.112391930835735,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045552199030194274,
      "loss": 4.6637,
      "mean_token_accuracy": 0.2518756836652756,
      "num_tokens": 50437262.0,
      "step": 21990
    },
    {
      "entropy": 5.299114561080932,
      "epoch": 2.1128722382324687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045550152455098113,
      "loss": 4.6824,
      "mean_token_accuracy": 0.24632540643215178,
      "num_tokens": 50448564.0,
      "step": 21995
    },
    {
      "entropy": 5.307951736450195,
      "epoch": 2.1133525456292026,
      "grad_norm": 1.125,
      "learning_rate": 0.0004554810546093102,
      "loss": 4.6582,
      "mean_token_accuracy": 0.2571727842092514,
      "num_tokens": 50459416.0,
      "step": 22000
    },
    {
      "entropy": 5.349908638000488,
      "epoch": 2.1138328530259365,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000455460580477405,
      "loss": 4.7374,
      "mean_token_accuracy": 0.2513375625014305,
      "num_tokens": 50470934.0,
      "step": 22005
    },
    {
      "entropy": 5.386178302764892,
      "epoch": 2.1143131604226704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000455440102155741,
      "loss": 4.7807,
      "mean_token_accuracy": 0.23916062712669373,
      "num_tokens": 50483050.0,
      "step": 22010
    },
    {
      "entropy": 5.345665550231933,
      "epoch": 2.1147934678194042,
      "grad_norm": 1.125,
      "learning_rate": 0.0004554196196447937,
      "loss": 4.7731,
      "mean_token_accuracy": 0.2344211757183075,
      "num_tokens": 50494655.0,
      "step": 22015
    },
    {
      "entropy": 5.3684123992919925,
      "epoch": 2.115273775216138,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553991329450385,
      "loss": 4.7202,
      "mean_token_accuracy": 0.23738428056240082,
      "num_tokens": 50506302.0,
      "step": 22020
    },
    {
      "entropy": 5.372774219512939,
      "epoch": 2.115754082612872,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045537864205695116,
      "loss": 4.7492,
      "mean_token_accuracy": 0.24560192674398423,
      "num_tokens": 50517979.0,
      "step": 22025
    },
    {
      "entropy": 5.3706080436706545,
      "epoch": 2.1162343900096063,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004553581469810073,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22892662584781648,
      "num_tokens": 50530054.0,
      "step": 22030
    },
    {
      "entropy": 5.315537643432617,
      "epoch": 2.11671469740634,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045533764771768287,
      "loss": 4.7434,
      "mean_token_accuracy": 0.23585093021392822,
      "num_tokens": 50540664.0,
      "step": 22035
    },
    {
      "entropy": 5.3736084461212155,
      "epoch": 2.117195004803074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045531714426745373,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23979576528072358,
      "num_tokens": 50551444.0,
      "step": 22040
    },
    {
      "entropy": 5.297380590438843,
      "epoch": 2.117675312199808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004552966366307959,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24549348950386046,
      "num_tokens": 50562371.0,
      "step": 22045
    },
    {
      "entropy": 5.283372163772583,
      "epoch": 2.118155619596542,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004552761248081856,
      "loss": 4.6941,
      "mean_token_accuracy": 0.24508881866931914,
      "num_tokens": 50573545.0,
      "step": 22050
    },
    {
      "entropy": 5.417281913757324,
      "epoch": 2.1186359269932757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000455255608800099,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2323501169681549,
      "num_tokens": 50584799.0,
      "step": 22055
    },
    {
      "entropy": 5.364929819107056,
      "epoch": 2.1191162343900096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045523508860701237,
      "loss": 4.7598,
      "mean_token_accuracy": 0.23524027317762375,
      "num_tokens": 50596435.0,
      "step": 22060
    },
    {
      "entropy": 5.345365428924561,
      "epoch": 2.1195965417867435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004552145642294021,
      "loss": 4.6701,
      "mean_token_accuracy": 0.25341845899820326,
      "num_tokens": 50606921.0,
      "step": 22065
    },
    {
      "entropy": 5.349429416656494,
      "epoch": 2.1200768491834774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045519403566774493,
      "loss": 4.8687,
      "mean_token_accuracy": 0.23051507771015167,
      "num_tokens": 50618107.0,
      "step": 22070
    },
    {
      "entropy": 5.321342420578003,
      "epoch": 2.1205571565802113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004551735029225172,
      "loss": 4.7668,
      "mean_token_accuracy": 0.23683675676584243,
      "num_tokens": 50628958.0,
      "step": 22075
    },
    {
      "entropy": 5.343475437164306,
      "epoch": 2.121037463976945,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045515296599419583,
      "loss": 4.7322,
      "mean_token_accuracy": 0.24019158333539964,
      "num_tokens": 50640557.0,
      "step": 22080
    },
    {
      "entropy": 5.360099411010742,
      "epoch": 2.121517771373679,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004551324248832574,
      "loss": 4.7573,
      "mean_token_accuracy": 0.2449215680360794,
      "num_tokens": 50652699.0,
      "step": 22085
    },
    {
      "entropy": 5.303769302368164,
      "epoch": 2.121998078770413,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004551118795901791,
      "loss": 4.6717,
      "mean_token_accuracy": 0.2408018559217453,
      "num_tokens": 50663735.0,
      "step": 22090
    },
    {
      "entropy": 5.280429172515869,
      "epoch": 2.122478386167147,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004550913301154376,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24172598570585252,
      "num_tokens": 50676384.0,
      "step": 22095
    },
    {
      "entropy": 5.350194120407105,
      "epoch": 2.1229586935638807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004550707764595103,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23891211897134781,
      "num_tokens": 50687775.0,
      "step": 22100
    },
    {
      "entropy": 5.36025824546814,
      "epoch": 2.123439000960615,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045505021862287434,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2353914737701416,
      "num_tokens": 50698522.0,
      "step": 22105
    },
    {
      "entropy": 5.283820343017578,
      "epoch": 2.123919308357349,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045502965660600684,
      "loss": 4.7194,
      "mean_token_accuracy": 0.24966657757759095,
      "num_tokens": 50709994.0,
      "step": 22110
    },
    {
      "entropy": 5.2845807552337645,
      "epoch": 2.1243996157540828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004550090904093853,
      "loss": 4.7031,
      "mean_token_accuracy": 0.24091558754444123,
      "num_tokens": 50720777.0,
      "step": 22115
    },
    {
      "entropy": 5.283737516403198,
      "epoch": 2.1248799231508166,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004549885200334872,
      "loss": 4.6882,
      "mean_token_accuracy": 0.24833122193813323,
      "num_tokens": 50732836.0,
      "step": 22120
    },
    {
      "entropy": 5.336600399017334,
      "epoch": 2.1253602305475505,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004549679454787901,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24954746663570404,
      "num_tokens": 50743854.0,
      "step": 22125
    },
    {
      "entropy": 5.242141389846802,
      "epoch": 2.1258405379442844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045494736674577175,
      "loss": 4.6623,
      "mean_token_accuracy": 0.24436326026916505,
      "num_tokens": 50755524.0,
      "step": 22130
    },
    {
      "entropy": 5.2904534339904785,
      "epoch": 2.1263208453410183,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004549267838349099,
      "loss": 4.7019,
      "mean_token_accuracy": 0.24525574892759322,
      "num_tokens": 50766909.0,
      "step": 22135
    },
    {
      "entropy": 5.264840412139892,
      "epoch": 2.126801152737752,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004549061967466823,
      "loss": 4.6782,
      "mean_token_accuracy": 0.24857996255159379,
      "num_tokens": 50778065.0,
      "step": 22140
    },
    {
      "entropy": 5.342424821853638,
      "epoch": 2.127281460134486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004548856054815671,
      "loss": 4.7631,
      "mean_token_accuracy": 0.2403675928711891,
      "num_tokens": 50790122.0,
      "step": 22145
    },
    {
      "entropy": 5.301620531082153,
      "epoch": 2.12776176753122,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045486501004004225,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23365176022052764,
      "num_tokens": 50802661.0,
      "step": 22150
    },
    {
      "entropy": 5.340081024169922,
      "epoch": 2.128242074927954,
      "grad_norm": 1.046875,
      "learning_rate": 0.000454844410422586,
      "loss": 4.7865,
      "mean_token_accuracy": 0.24084324687719344,
      "num_tokens": 50814279.0,
      "step": 22155
    },
    {
      "entropy": 5.266798686981201,
      "epoch": 2.1287223823246877,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045482380662967655,
      "loss": 4.6303,
      "mean_token_accuracy": 0.2545595452189445,
      "num_tokens": 50826186.0,
      "step": 22160
    },
    {
      "entropy": 5.33102765083313,
      "epoch": 2.1292026897214216,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004548031986617923,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2317605048418045,
      "num_tokens": 50838185.0,
      "step": 22165
    },
    {
      "entropy": 5.290328693389893,
      "epoch": 2.1296829971181555,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004547825865194117,
      "loss": 4.7179,
      "mean_token_accuracy": 0.24740959703922272,
      "num_tokens": 50850461.0,
      "step": 22170
    },
    {
      "entropy": 5.363900756835937,
      "epoch": 2.1301633045148893,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045476197020301323,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23892285227775573,
      "num_tokens": 50861859.0,
      "step": 22175
    },
    {
      "entropy": 5.3025449275970455,
      "epoch": 2.1306436119116237,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045474134971307554,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24590679705142976,
      "num_tokens": 50872776.0,
      "step": 22180
    },
    {
      "entropy": 5.306111288070679,
      "epoch": 2.1311239193083575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004547207250500775,
      "loss": 4.6994,
      "mean_token_accuracy": 0.24347657710313797,
      "num_tokens": 50885333.0,
      "step": 22185
    },
    {
      "entropy": 5.336797475814819,
      "epoch": 2.1316042267050914,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000454700096214498,
      "loss": 4.7115,
      "mean_token_accuracy": 0.23617301434278487,
      "num_tokens": 50897256.0,
      "step": 22190
    },
    {
      "entropy": 5.28154239654541,
      "epoch": 2.1320845341018253,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045467946320681567,
      "loss": 4.6999,
      "mean_token_accuracy": 0.24231407046318054,
      "num_tokens": 50909558.0,
      "step": 22195
    },
    {
      "entropy": 5.310467624664307,
      "epoch": 2.132564841498559,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004546588260275098,
      "loss": 4.7458,
      "mean_token_accuracy": 0.24467129707336427,
      "num_tokens": 50920617.0,
      "step": 22200
    },
    {
      "entropy": 5.316555166244507,
      "epoch": 2.133045148895293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045463818467705955,
      "loss": 4.7243,
      "mean_token_accuracy": 0.24242780208587647,
      "num_tokens": 50931323.0,
      "step": 22205
    },
    {
      "entropy": 5.270558023452759,
      "epoch": 2.133525456292027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000454617539155944,
      "loss": 4.6216,
      "mean_token_accuracy": 0.25468734502792356,
      "num_tokens": 50942798.0,
      "step": 22210
    },
    {
      "entropy": 5.233743000030517,
      "epoch": 2.134005763688761,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045459688946464255,
      "loss": 4.6154,
      "mean_token_accuracy": 0.24316914826631547,
      "num_tokens": 50954877.0,
      "step": 22215
    },
    {
      "entropy": 5.35350775718689,
      "epoch": 2.1344860710854947,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004545762356036346,
      "loss": 4.6774,
      "mean_token_accuracy": 0.24728553593158722,
      "num_tokens": 50966133.0,
      "step": 22220
    },
    {
      "entropy": 5.325016212463379,
      "epoch": 2.1349663784822286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545555775733998,
      "loss": 4.6834,
      "mean_token_accuracy": 0.25661215335130694,
      "num_tokens": 50977334.0,
      "step": 22225
    },
    {
      "entropy": 5.326789140701294,
      "epoch": 2.1354466858789625,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045453491537441747,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2393910527229309,
      "num_tokens": 50988743.0,
      "step": 22230
    },
    {
      "entropy": 5.3205994129180905,
      "epoch": 2.1359269932756964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045451424900716763,
      "loss": 4.7123,
      "mean_token_accuracy": 0.24441358894109727,
      "num_tokens": 51000591.0,
      "step": 22235
    },
    {
      "entropy": 5.434566593170166,
      "epoch": 2.1364073006724302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045449357847212994,
      "loss": 4.8965,
      "mean_token_accuracy": 0.23482633084058763,
      "num_tokens": 51011565.0,
      "step": 22240
    },
    {
      "entropy": 5.343700551986695,
      "epoch": 2.136887608069164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004544729037697844,
      "loss": 4.7657,
      "mean_token_accuracy": 0.23950437754392623,
      "num_tokens": 51022755.0,
      "step": 22245
    },
    {
      "entropy": 5.372124052047729,
      "epoch": 2.137367915465898,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045445222490061093,
      "loss": 4.813,
      "mean_token_accuracy": 0.23771845549345016,
      "num_tokens": 51032839.0,
      "step": 22250
    },
    {
      "entropy": 5.337783288955689,
      "epoch": 2.1378482228626323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004544315418650897,
      "loss": 4.7575,
      "mean_token_accuracy": 0.24587953388690947,
      "num_tokens": 51045020.0,
      "step": 22255
    },
    {
      "entropy": 5.290575695037842,
      "epoch": 2.138328530259366,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004544108546637008,
      "loss": 4.6613,
      "mean_token_accuracy": 0.2512912377715111,
      "num_tokens": 51056765.0,
      "step": 22260
    },
    {
      "entropy": 5.275262022018433,
      "epoch": 2.1388088376561,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004543901632969247,
      "loss": 4.6816,
      "mean_token_accuracy": 0.24599525481462478,
      "num_tokens": 51069559.0,
      "step": 22265
    },
    {
      "entropy": 5.298547029495239,
      "epoch": 2.139289145052834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045436946776524157,
      "loss": 4.7328,
      "mean_token_accuracy": 0.24131276607513427,
      "num_tokens": 51080972.0,
      "step": 22270
    },
    {
      "entropy": 5.2581686019897464,
      "epoch": 2.139769452449568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045434876806913204,
      "loss": 4.6635,
      "mean_token_accuracy": 0.24881070405244826,
      "num_tokens": 51092101.0,
      "step": 22275
    },
    {
      "entropy": 5.400112056732178,
      "epoch": 2.1402497598463017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004543280642090767,
      "loss": 4.8745,
      "mean_token_accuracy": 0.23714352399110794,
      "num_tokens": 51103372.0,
      "step": 22280
    },
    {
      "entropy": 5.344463062286377,
      "epoch": 2.1407300672430356,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004543073561855562,
      "loss": 4.7661,
      "mean_token_accuracy": 0.23569402247667312,
      "num_tokens": 51114856.0,
      "step": 22285
    },
    {
      "entropy": 5.42238597869873,
      "epoch": 2.1412103746397695,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004542866439990513,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23754600286483765,
      "num_tokens": 51126354.0,
      "step": 22290
    },
    {
      "entropy": 5.377756404876709,
      "epoch": 2.1416906820365034,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004542659276500429,
      "loss": 4.7888,
      "mean_token_accuracy": 0.24248652756214142,
      "num_tokens": 51138341.0,
      "step": 22295
    },
    {
      "entropy": 5.25858883857727,
      "epoch": 2.1421709894332372,
      "grad_norm": 1.0,
      "learning_rate": 0.00045424520713901204,
      "loss": 4.6594,
      "mean_token_accuracy": 0.2514404088258743,
      "num_tokens": 51150969.0,
      "step": 22300
    },
    {
      "entropy": 5.283135175704956,
      "epoch": 2.142651296829971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004542244824664396,
      "loss": 4.6994,
      "mean_token_accuracy": 0.24365271925926207,
      "num_tokens": 51162199.0,
      "step": 22305
    },
    {
      "entropy": 5.319918489456176,
      "epoch": 2.143131604226705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045420375363280696,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23767067193984986,
      "num_tokens": 51173129.0,
      "step": 22310
    },
    {
      "entropy": 5.256084632873535,
      "epoch": 2.143611911623439,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045418302063859526,
      "loss": 4.6879,
      "mean_token_accuracy": 0.24828515350818633,
      "num_tokens": 51184593.0,
      "step": 22315
    },
    {
      "entropy": 5.312399959564209,
      "epoch": 2.1440922190201728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045416228348428583,
      "loss": 4.7197,
      "mean_token_accuracy": 0.24911215156316757,
      "num_tokens": 51196524.0,
      "step": 22320
    },
    {
      "entropy": 5.326993179321289,
      "epoch": 2.1445725264169067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045414154217036023,
      "loss": 4.7954,
      "mean_token_accuracy": 0.2428452581167221,
      "num_tokens": 51209307.0,
      "step": 22325
    },
    {
      "entropy": 5.2294508934021,
      "epoch": 2.1450528338136405,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045412079669730006,
      "loss": 4.6998,
      "mean_token_accuracy": 0.24746936559677124,
      "num_tokens": 51220919.0,
      "step": 22330
    },
    {
      "entropy": 5.291134405136108,
      "epoch": 2.1455331412103744,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004541000470655867,
      "loss": 4.6608,
      "mean_token_accuracy": 0.24957104176282882,
      "num_tokens": 51232930.0,
      "step": 22335
    },
    {
      "entropy": 5.292617130279541,
      "epoch": 2.1460134486071087,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045407929327570215,
      "loss": 4.7465,
      "mean_token_accuracy": 0.2441573604941368,
      "num_tokens": 51244697.0,
      "step": 22340
    },
    {
      "entropy": 5.281714725494385,
      "epoch": 2.1464937560038426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004540585353281282,
      "loss": 4.7515,
      "mean_token_accuracy": 0.2484479933977127,
      "num_tokens": 51255333.0,
      "step": 22345
    },
    {
      "entropy": 5.362227821350098,
      "epoch": 2.1469740634005765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004540377732233467,
      "loss": 4.741,
      "mean_token_accuracy": 0.24192020893096924,
      "num_tokens": 51267649.0,
      "step": 22350
    },
    {
      "entropy": 5.349464082717896,
      "epoch": 2.1474543707973104,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004540170069618397,
      "loss": 4.6556,
      "mean_token_accuracy": 0.24899276494979858,
      "num_tokens": 51279576.0,
      "step": 22355
    },
    {
      "entropy": 5.357388925552368,
      "epoch": 2.1479346781940443,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045399623654408946,
      "loss": 4.7326,
      "mean_token_accuracy": 0.23733433932065964,
      "num_tokens": 51290974.0,
      "step": 22360
    },
    {
      "entropy": 5.413909292221069,
      "epoch": 2.148414985590778,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000453975461970578,
      "loss": 4.8014,
      "mean_token_accuracy": 0.2366686537861824,
      "num_tokens": 51301405.0,
      "step": 22365
    },
    {
      "entropy": 5.439300775527954,
      "epoch": 2.148895292987512,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004539546832417879,
      "loss": 4.8519,
      "mean_token_accuracy": 0.23393606543540954,
      "num_tokens": 51313158.0,
      "step": 22370
    },
    {
      "entropy": 5.370654487609864,
      "epoch": 2.149375600384246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045393390035820136,
      "loss": 4.7398,
      "mean_token_accuracy": 0.2504728898406029,
      "num_tokens": 51323789.0,
      "step": 22375
    },
    {
      "entropy": 5.271580171585083,
      "epoch": 2.14985590778098,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000453913113320301,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23867260068655013,
      "num_tokens": 51334864.0,
      "step": 22380
    },
    {
      "entropy": 5.374039268493652,
      "epoch": 2.1503362151777137,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004538923221285694,
      "loss": 4.673,
      "mean_token_accuracy": 0.24237383008003235,
      "num_tokens": 51345473.0,
      "step": 22385
    },
    {
      "entropy": 5.4331896781921385,
      "epoch": 2.1508165225744476,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004538715267834893,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2380006268620491,
      "num_tokens": 51356755.0,
      "step": 22390
    },
    {
      "entropy": 5.198922777175904,
      "epoch": 2.1512968299711814,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004538507272855434,
      "loss": 4.6253,
      "mean_token_accuracy": 0.2474164441227913,
      "num_tokens": 51368351.0,
      "step": 22395
    },
    {
      "entropy": 5.376632452011108,
      "epoch": 2.1517771373679153,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045382992363521486,
      "loss": 4.873,
      "mean_token_accuracy": 0.23568562269210816,
      "num_tokens": 51379144.0,
      "step": 22400
    },
    {
      "entropy": 5.318851327896118,
      "epoch": 2.152257444764649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045380911583298633,
      "loss": 4.6754,
      "mean_token_accuracy": 0.24694026112556458,
      "num_tokens": 51390732.0,
      "step": 22405
    },
    {
      "entropy": 5.2883344173431395,
      "epoch": 2.152737752161383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045378830387934123,
      "loss": 4.6596,
      "mean_token_accuracy": 0.25084725171327593,
      "num_tokens": 51401185.0,
      "step": 22410
    },
    {
      "entropy": 5.356513023376465,
      "epoch": 2.1532180595581174,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004537674877747626,
      "loss": 4.7591,
      "mean_token_accuracy": 0.23870777040719987,
      "num_tokens": 51412598.0,
      "step": 22415
    },
    {
      "entropy": 5.327118253707885,
      "epoch": 2.1536983669548513,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045374666751973365,
      "loss": 4.6712,
      "mean_token_accuracy": 0.24329276978969575,
      "num_tokens": 51423248.0,
      "step": 22420
    },
    {
      "entropy": 5.27250566482544,
      "epoch": 2.154178674351585,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045372584311473784,
      "loss": 4.6205,
      "mean_token_accuracy": 0.2544675514101982,
      "num_tokens": 51433862.0,
      "step": 22425
    },
    {
      "entropy": 5.278048896789551,
      "epoch": 2.154658981748319,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004537050145602587,
      "loss": 4.7214,
      "mean_token_accuracy": 0.24564757347106933,
      "num_tokens": 51445311.0,
      "step": 22430
    },
    {
      "entropy": 5.257189512252808,
      "epoch": 2.155139289145053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004536841818567798,
      "loss": 4.6436,
      "mean_token_accuracy": 0.2511398658156395,
      "num_tokens": 51456550.0,
      "step": 22435
    },
    {
      "entropy": 5.314164495468139,
      "epoch": 2.155619596541787,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004536633450047847,
      "loss": 4.7262,
      "mean_token_accuracy": 0.24852168262004853,
      "num_tokens": 51467645.0,
      "step": 22440
    },
    {
      "entropy": 5.308267211914062,
      "epoch": 2.1560999039385207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045364250400475734,
      "loss": 4.7193,
      "mean_token_accuracy": 0.24909869879484176,
      "num_tokens": 51478275.0,
      "step": 22445
    },
    {
      "entropy": 5.289500761032104,
      "epoch": 2.1565802113352546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004536216588571814,
      "loss": 4.694,
      "mean_token_accuracy": 0.24813806116580964,
      "num_tokens": 51489634.0,
      "step": 22450
    },
    {
      "entropy": 5.256476926803589,
      "epoch": 2.1570605187319885,
      "grad_norm": 1.203125,
      "learning_rate": 0.000453600809562541,
      "loss": 4.7047,
      "mean_token_accuracy": 0.25240506380796435,
      "num_tokens": 51500245.0,
      "step": 22455
    },
    {
      "entropy": 5.2995795726776125,
      "epoch": 2.1575408261287223,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004535799561213202,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24533067643642426,
      "num_tokens": 51511996.0,
      "step": 22460
    },
    {
      "entropy": 5.3323485374450685,
      "epoch": 2.158021133525456,
      "grad_norm": 0.921875,
      "learning_rate": 0.000453559098534003,
      "loss": 4.6892,
      "mean_token_accuracy": 0.24780250042676927,
      "num_tokens": 51523522.0,
      "step": 22465
    },
    {
      "entropy": 5.351586484909058,
      "epoch": 2.15850144092219,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004535382368010738,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24179369509220122,
      "num_tokens": 51535261.0,
      "step": 22470
    },
    {
      "entropy": 5.279510354995727,
      "epoch": 2.158981748318924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045351737092301676,
      "loss": 4.6554,
      "mean_token_accuracy": 0.24808209538459777,
      "num_tokens": 51546965.0,
      "step": 22475
    },
    {
      "entropy": 5.291923475265503,
      "epoch": 2.159462055715658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004534965009003165,
      "loss": 4.681,
      "mean_token_accuracy": 0.2467667803168297,
      "num_tokens": 51558976.0,
      "step": 22480
    },
    {
      "entropy": 5.40265007019043,
      "epoch": 2.1599423631123917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004534756267334576,
      "loss": 4.8015,
      "mean_token_accuracy": 0.23777738362550735,
      "num_tokens": 51571625.0,
      "step": 22485
    },
    {
      "entropy": 5.41196117401123,
      "epoch": 2.160422670509126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045345474842292455,
      "loss": 4.8224,
      "mean_token_accuracy": 0.2399259850382805,
      "num_tokens": 51582732.0,
      "step": 22490
    },
    {
      "entropy": 5.301330327987671,
      "epoch": 2.16090297790586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004534338659692022,
      "loss": 4.7188,
      "mean_token_accuracy": 0.2415502890944481,
      "num_tokens": 51594346.0,
      "step": 22495
    },
    {
      "entropy": 5.335553073883057,
      "epoch": 2.161383285302594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004534129793727753,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24448925405740737,
      "num_tokens": 51606909.0,
      "step": 22500
    },
    {
      "entropy": 5.309983539581299,
      "epoch": 2.1618635926993277,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004533920886341288,
      "loss": 4.7132,
      "mean_token_accuracy": 0.24355805814266204,
      "num_tokens": 51617596.0,
      "step": 22505
    },
    {
      "entropy": 5.25330753326416,
      "epoch": 2.1623439000960616,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004533711937537477,
      "loss": 4.5926,
      "mean_token_accuracy": 0.25561256110668185,
      "num_tokens": 51627573.0,
      "step": 22510
    },
    {
      "entropy": 5.266866207122803,
      "epoch": 2.1628242074927955,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004533502947321171,
      "loss": 4.6738,
      "mean_token_accuracy": 0.23748915940523146,
      "num_tokens": 51640070.0,
      "step": 22515
    },
    {
      "entropy": 5.325762748718262,
      "epoch": 2.1633045148895294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004533293915697223,
      "loss": 4.621,
      "mean_token_accuracy": 0.24940133541822435,
      "num_tokens": 51651139.0,
      "step": 22520
    },
    {
      "entropy": 5.390152502059936,
      "epoch": 2.1637848222862632,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045330848426704853,
      "loss": 4.8027,
      "mean_token_accuracy": 0.24498158991336821,
      "num_tokens": 51661893.0,
      "step": 22525
    },
    {
      "entropy": 5.307491588592529,
      "epoch": 2.164265129682997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004532875728245813,
      "loss": 4.7101,
      "mean_token_accuracy": 0.24339240789413452,
      "num_tokens": 51674266.0,
      "step": 22530
    },
    {
      "entropy": 5.322489261627197,
      "epoch": 2.164745437079731,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045326665724280594,
      "loss": 4.7439,
      "mean_token_accuracy": 0.25177913159132004,
      "num_tokens": 51685450.0,
      "step": 22535
    },
    {
      "entropy": 5.330208444595337,
      "epoch": 2.165225744476465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045324573752220814,
      "loss": 4.7383,
      "mean_token_accuracy": 0.249210324883461,
      "num_tokens": 51696779.0,
      "step": 22540
    },
    {
      "entropy": 5.38946270942688,
      "epoch": 2.1657060518731988,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045322481366327365,
      "loss": 4.7599,
      "mean_token_accuracy": 0.2393000155687332,
      "num_tokens": 51709387.0,
      "step": 22545
    },
    {
      "entropy": 5.32688422203064,
      "epoch": 2.1661863592699326,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004532038856664882,
      "loss": 4.7303,
      "mean_token_accuracy": 0.24115523099899291,
      "num_tokens": 51720978.0,
      "step": 22550
    },
    {
      "entropy": 5.422133159637451,
      "epoch": 2.1666666666666665,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004531829535323376,
      "loss": 4.8392,
      "mean_token_accuracy": 0.23532693833112717,
      "num_tokens": 51733510.0,
      "step": 22555
    },
    {
      "entropy": 5.2415238380432125,
      "epoch": 2.1671469740634004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000453162017261308,
      "loss": 4.6392,
      "mean_token_accuracy": 0.2555921941995621,
      "num_tokens": 51743660.0,
      "step": 22560
    },
    {
      "entropy": 5.31256856918335,
      "epoch": 2.1676272814601343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004531410768538854,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24580712020397186,
      "num_tokens": 51754592.0,
      "step": 22565
    },
    {
      "entropy": 5.302414417266846,
      "epoch": 2.168107588856868,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045312013231055596,
      "loss": 4.6714,
      "mean_token_accuracy": 0.25421386659145356,
      "num_tokens": 51766241.0,
      "step": 22570
    },
    {
      "entropy": 5.370648956298828,
      "epoch": 2.1685878962536025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045309918363180593,
      "loss": 4.7034,
      "mean_token_accuracy": 0.23981606364250183,
      "num_tokens": 51777718.0,
      "step": 22575
    },
    {
      "entropy": 5.202638006210327,
      "epoch": 2.1690682036503364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045307823081812166,
      "loss": 4.6358,
      "mean_token_accuracy": 0.2531831279397011,
      "num_tokens": 51788290.0,
      "step": 22580
    },
    {
      "entropy": 5.269257116317749,
      "epoch": 2.1695485110470702,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045305727386998977,
      "loss": 4.6652,
      "mean_token_accuracy": 0.24787796884775162,
      "num_tokens": 51800080.0,
      "step": 22585
    },
    {
      "entropy": 5.345980262756347,
      "epoch": 2.170028818443804,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004530363127878966,
      "loss": 4.7377,
      "mean_token_accuracy": 0.2434770166873932,
      "num_tokens": 51810155.0,
      "step": 22590
    },
    {
      "entropy": 5.327930641174317,
      "epoch": 2.170509125840538,
      "grad_norm": 1.0,
      "learning_rate": 0.00045301534757232885,
      "loss": 4.7485,
      "mean_token_accuracy": 0.2450922355055809,
      "num_tokens": 51823082.0,
      "step": 22595
    },
    {
      "entropy": 5.362551927566528,
      "epoch": 2.170989433237272,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004529943782237735,
      "loss": 4.7963,
      "mean_token_accuracy": 0.24219521582126619,
      "num_tokens": 51833280.0,
      "step": 22600
    },
    {
      "entropy": 5.315093183517456,
      "epoch": 2.1714697406340058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045297340474271717,
      "loss": 4.7186,
      "mean_token_accuracy": 0.2432359740138054,
      "num_tokens": 51845105.0,
      "step": 22605
    },
    {
      "entropy": 5.366456413269043,
      "epoch": 2.1719500480307397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004529524271296468,
      "loss": 4.7035,
      "mean_token_accuracy": 0.252385550737381,
      "num_tokens": 51856232.0,
      "step": 22610
    },
    {
      "entropy": 5.323475933074951,
      "epoch": 2.1724303554274735,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045293144538504943,
      "loss": 4.7294,
      "mean_token_accuracy": 0.24988365471363067,
      "num_tokens": 51866218.0,
      "step": 22615
    },
    {
      "entropy": 5.27262167930603,
      "epoch": 2.1729106628242074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004529104595094124,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24463127553462982,
      "num_tokens": 51877425.0,
      "step": 22620
    },
    {
      "entropy": 5.325945043563843,
      "epoch": 2.1733909702209413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045288946950322264,
      "loss": 4.744,
      "mean_token_accuracy": 0.24236492067575455,
      "num_tokens": 51888818.0,
      "step": 22625
    },
    {
      "entropy": 5.335425996780396,
      "epoch": 2.173871277617675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004528684753669677,
      "loss": 4.7015,
      "mean_token_accuracy": 0.24094209372997283,
      "num_tokens": 51899875.0,
      "step": 22630
    },
    {
      "entropy": 5.342225122451782,
      "epoch": 2.174351585014409,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004528474771011349,
      "loss": 4.719,
      "mean_token_accuracy": 0.23986474573612213,
      "num_tokens": 51911084.0,
      "step": 22635
    },
    {
      "entropy": 5.309115409851074,
      "epoch": 2.174831892411143,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045282647470621176,
      "loss": 4.7406,
      "mean_token_accuracy": 0.2421930029988289,
      "num_tokens": 51922165.0,
      "step": 22640
    },
    {
      "entropy": 5.313009881973267,
      "epoch": 2.175312199807877,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045280546818268595,
      "loss": 4.7383,
      "mean_token_accuracy": 0.24547616690397261,
      "num_tokens": 51934827.0,
      "step": 22645
    },
    {
      "entropy": 5.354127359390259,
      "epoch": 2.175792507204611,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004527844575310452,
      "loss": 4.761,
      "mean_token_accuracy": 0.23750930279493332,
      "num_tokens": 51947997.0,
      "step": 22650
    },
    {
      "entropy": 5.325842952728271,
      "epoch": 2.176272814601345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045276344275177715,
      "loss": 4.7783,
      "mean_token_accuracy": 0.24136972278356553,
      "num_tokens": 51959639.0,
      "step": 22655
    },
    {
      "entropy": 5.359508752822876,
      "epoch": 2.176753121998079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045274242384536984,
      "loss": 4.8009,
      "mean_token_accuracy": 0.24006113409996033,
      "num_tokens": 51970716.0,
      "step": 22660
    },
    {
      "entropy": 5.354462385177612,
      "epoch": 2.177233429394813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527214008123113,
      "loss": 4.71,
      "mean_token_accuracy": 0.2472544178366661,
      "num_tokens": 51982556.0,
      "step": 22665
    },
    {
      "entropy": 5.361033201217651,
      "epoch": 2.1777137367915467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527003736530895,
      "loss": 4.7035,
      "mean_token_accuracy": 0.2522361233830452,
      "num_tokens": 51994894.0,
      "step": 22670
    },
    {
      "entropy": 5.313359069824219,
      "epoch": 2.1781940441882806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045267934236819265,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23234739154577255,
      "num_tokens": 52007128.0,
      "step": 22675
    },
    {
      "entropy": 5.321500444412232,
      "epoch": 2.1786743515850144,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004526583069581091,
      "loss": 4.8795,
      "mean_token_accuracy": 0.23352739661931993,
      "num_tokens": 52019598.0,
      "step": 22680
    },
    {
      "entropy": 5.37825756072998,
      "epoch": 2.1791546589817483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004526372674233272,
      "loss": 4.7534,
      "mean_token_accuracy": 0.24207944124937059,
      "num_tokens": 52031445.0,
      "step": 22685
    },
    {
      "entropy": 5.396061611175537,
      "epoch": 2.179634966378482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045261622376433543,
      "loss": 4.7797,
      "mean_token_accuracy": 0.2418875738978386,
      "num_tokens": 52042548.0,
      "step": 22690
    },
    {
      "entropy": 5.292735385894775,
      "epoch": 2.180115273775216,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045259517598162237,
      "loss": 4.6948,
      "mean_token_accuracy": 0.24668315649032593,
      "num_tokens": 52054147.0,
      "step": 22695
    },
    {
      "entropy": 5.293334054946899,
      "epoch": 2.18059558117195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004525741240756766,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2498183161020279,
      "num_tokens": 52065556.0,
      "step": 22700
    },
    {
      "entropy": 5.3786266326904295,
      "epoch": 2.181075888568684,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004525530680469871,
      "loss": 4.7744,
      "mean_token_accuracy": 0.24442436248064042,
      "num_tokens": 52077543.0,
      "step": 22705
    },
    {
      "entropy": 5.310362720489502,
      "epoch": 2.1815561959654177,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045253200789604245,
      "loss": 4.7208,
      "mean_token_accuracy": 0.2430674910545349,
      "num_tokens": 52088376.0,
      "step": 22710
    },
    {
      "entropy": 5.2983006000518795,
      "epoch": 2.1820365033621516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045251094362333186,
      "loss": 4.7412,
      "mean_token_accuracy": 0.24067662805318832,
      "num_tokens": 52100019.0,
      "step": 22715
    },
    {
      "entropy": 5.295056247711182,
      "epoch": 2.1825168107588855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004524898752293441,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24142597168684005,
      "num_tokens": 52110643.0,
      "step": 22720
    },
    {
      "entropy": 5.247287893295288,
      "epoch": 2.18299711815562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045246880271456857,
      "loss": 4.6352,
      "mean_token_accuracy": 0.2525670349597931,
      "num_tokens": 52123092.0,
      "step": 22725
    },
    {
      "entropy": 5.39801516532898,
      "epoch": 2.1834774255523537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004524477260794944,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23809388130903245,
      "num_tokens": 52135371.0,
      "step": 22730
    },
    {
      "entropy": 5.33134183883667,
      "epoch": 2.1839577329490876,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045242664532461094,
      "loss": 4.6823,
      "mean_token_accuracy": 0.25072041898965836,
      "num_tokens": 52145877.0,
      "step": 22735
    },
    {
      "entropy": 5.277892065048218,
      "epoch": 2.1844380403458215,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045240556045040767,
      "loss": 4.728,
      "mean_token_accuracy": 0.24744197726249695,
      "num_tokens": 52158228.0,
      "step": 22740
    },
    {
      "entropy": 5.398694562911987,
      "epoch": 2.1849183477425553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045238447145737397,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22686052173376084,
      "num_tokens": 52169515.0,
      "step": 22745
    },
    {
      "entropy": 5.369239568710327,
      "epoch": 2.185398655139289,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045236337834599966,
      "loss": 4.7628,
      "mean_token_accuracy": 0.2380661204457283,
      "num_tokens": 52179846.0,
      "step": 22750
    },
    {
      "entropy": 5.200771284103394,
      "epoch": 2.185878962536023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045234228111677434,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2482399210333824,
      "num_tokens": 52191095.0,
      "step": 22755
    },
    {
      "entropy": 5.352427434921265,
      "epoch": 2.186359269932757,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004523211797701878,
      "loss": 4.784,
      "mean_token_accuracy": 0.23832864314317703,
      "num_tokens": 52202289.0,
      "step": 22760
    },
    {
      "entropy": 5.2880363941192625,
      "epoch": 2.186839577329491,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045230007430673014,
      "loss": 4.7218,
      "mean_token_accuracy": 0.24718229621648788,
      "num_tokens": 52214424.0,
      "step": 22765
    },
    {
      "entropy": 5.385380601882934,
      "epoch": 2.1873198847262247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004522789647268911,
      "loss": 4.7614,
      "mean_token_accuracy": 0.23876139968633653,
      "num_tokens": 52225446.0,
      "step": 22770
    },
    {
      "entropy": 5.338502502441406,
      "epoch": 2.1878001921229586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004522578510311609,
      "loss": 4.6893,
      "mean_token_accuracy": 0.2425274521112442,
      "num_tokens": 52236063.0,
      "step": 22775
    },
    {
      "entropy": 5.2680158615112305,
      "epoch": 2.1882804995196925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045223673322002984,
      "loss": 4.6874,
      "mean_token_accuracy": 0.24625545889139175,
      "num_tokens": 52248618.0,
      "step": 22780
    },
    {
      "entropy": 5.337622594833374,
      "epoch": 2.1887608069164264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045221561129398804,
      "loss": 4.8279,
      "mean_token_accuracy": 0.23445827066898345,
      "num_tokens": 52261913.0,
      "step": 22785
    },
    {
      "entropy": 5.374995565414428,
      "epoch": 2.1892411143131603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000452194485253526,
      "loss": 4.7072,
      "mean_token_accuracy": 0.24202573895454407,
      "num_tokens": 52273838.0,
      "step": 22790
    },
    {
      "entropy": 5.322381448745728,
      "epoch": 2.189721421709894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004521733550991342,
      "loss": 4.7554,
      "mean_token_accuracy": 0.24670039415359496,
      "num_tokens": 52285464.0,
      "step": 22795
    },
    {
      "entropy": 5.292813110351562,
      "epoch": 2.1902017291066285,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045215222083130316,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24720058739185333,
      "num_tokens": 52295800.0,
      "step": 22800
    },
    {
      "entropy": 5.362011194229126,
      "epoch": 2.1906820365033624,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004521310824505236,
      "loss": 4.7974,
      "mean_token_accuracy": 0.24249051958322526,
      "num_tokens": 52307243.0,
      "step": 22805
    },
    {
      "entropy": 5.325633716583252,
      "epoch": 2.1911623439000962,
      "grad_norm": 1.125,
      "learning_rate": 0.00045210993995728623,
      "loss": 4.7144,
      "mean_token_accuracy": 0.24473314136266708,
      "num_tokens": 52317065.0,
      "step": 22810
    },
    {
      "entropy": 5.326696872711182,
      "epoch": 2.19164265129683,
      "grad_norm": 1.109375,
      "learning_rate": 0.000452088793352082,
      "loss": 4.7289,
      "mean_token_accuracy": 0.24036959707736968,
      "num_tokens": 52328385.0,
      "step": 22815
    },
    {
      "entropy": 5.254281377792358,
      "epoch": 2.192122958693564,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004520676426354018,
      "loss": 4.6708,
      "mean_token_accuracy": 0.24988327771425248,
      "num_tokens": 52340046.0,
      "step": 22820
    },
    {
      "entropy": 5.247295475006103,
      "epoch": 2.192603266090298,
      "grad_norm": 1.125,
      "learning_rate": 0.0004520464878077368,
      "loss": 4.6818,
      "mean_token_accuracy": 0.24903482794761658,
      "num_tokens": 52352071.0,
      "step": 22825
    },
    {
      "entropy": 5.413037538528442,
      "epoch": 2.1930835734870318,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045202532886957805,
      "loss": 4.798,
      "mean_token_accuracy": 0.2431217074394226,
      "num_tokens": 52364173.0,
      "step": 22830
    },
    {
      "entropy": 5.289604377746582,
      "epoch": 2.1935638808837656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045200416582141676,
      "loss": 4.647,
      "mean_token_accuracy": 0.2484308511018753,
      "num_tokens": 52374690.0,
      "step": 22835
    },
    {
      "entropy": 5.373140954971314,
      "epoch": 2.1940441882804995,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004519829986637444,
      "loss": 4.896,
      "mean_token_accuracy": 0.23140619397163392,
      "num_tokens": 52385661.0,
      "step": 22840
    },
    {
      "entropy": 5.332276344299316,
      "epoch": 2.1945244956772334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004519618273970523,
      "loss": 4.7896,
      "mean_token_accuracy": 0.2418289825320244,
      "num_tokens": 52396759.0,
      "step": 22845
    },
    {
      "entropy": 5.419453048706055,
      "epoch": 2.1950048030739673,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045194065202183205,
      "loss": 4.8582,
      "mean_token_accuracy": 0.23576005846261977,
      "num_tokens": 52409122.0,
      "step": 22850
    },
    {
      "entropy": 5.48240385055542,
      "epoch": 2.195485110470701,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004519194725385752,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23353031575679778,
      "num_tokens": 52421318.0,
      "step": 22855
    },
    {
      "entropy": 5.25208306312561,
      "epoch": 2.195965417867435,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045189828894777364,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24486909806728363,
      "num_tokens": 52432595.0,
      "step": 22860
    },
    {
      "entropy": 5.277630186080932,
      "epoch": 2.196445725264169,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045187710124991904,
      "loss": 4.6965,
      "mean_token_accuracy": 0.2436639577150345,
      "num_tokens": 52444423.0,
      "step": 22865
    },
    {
      "entropy": 5.380270147323609,
      "epoch": 2.196926032660903,
      "grad_norm": 1.0,
      "learning_rate": 0.0004518559094455034,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2357235312461853,
      "num_tokens": 52457183.0,
      "step": 22870
    },
    {
      "entropy": 5.359071969985962,
      "epoch": 2.1974063400576367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004518347135350187,
      "loss": 4.7428,
      "mean_token_accuracy": 0.24182765930891037,
      "num_tokens": 52469624.0,
      "step": 22875
    },
    {
      "entropy": 5.391725826263428,
      "epoch": 2.1978866474543706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045181351351895703,
      "loss": 4.8517,
      "mean_token_accuracy": 0.23393048644065856,
      "num_tokens": 52481369.0,
      "step": 22880
    },
    {
      "entropy": 5.324646854400635,
      "epoch": 2.198366954851105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004517923093978106,
      "loss": 4.6647,
      "mean_token_accuracy": 0.24931210726499559,
      "num_tokens": 52493315.0,
      "step": 22885
    },
    {
      "entropy": 5.320997476577759,
      "epoch": 2.1988472622478388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004517711011720717,
      "loss": 4.7597,
      "mean_token_accuracy": 0.24816303551197053,
      "num_tokens": 52504929.0,
      "step": 22890
    },
    {
      "entropy": 5.3022784233093265,
      "epoch": 2.1993275696445727,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004517498888422327,
      "loss": 4.6486,
      "mean_token_accuracy": 0.24320167303085327,
      "num_tokens": 52518364.0,
      "step": 22895
    },
    {
      "entropy": 5.322622203826905,
      "epoch": 2.1998078770413065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004517286724087862,
      "loss": 4.7138,
      "mean_token_accuracy": 0.24447246342897416,
      "num_tokens": 52530359.0,
      "step": 22900
    },
    {
      "entropy": 5.343125867843628,
      "epoch": 2.2002881844380404,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004517074518722247,
      "loss": 4.7427,
      "mean_token_accuracy": 0.24339924305677413,
      "num_tokens": 52542242.0,
      "step": 22905
    },
    {
      "entropy": 5.407409048080444,
      "epoch": 2.2007684918347743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045168622723304084,
      "loss": 4.7826,
      "mean_token_accuracy": 0.23958794474601747,
      "num_tokens": 52552838.0,
      "step": 22910
    },
    {
      "entropy": 5.339862251281739,
      "epoch": 2.201248799231508,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004516649984917274,
      "loss": 4.7063,
      "mean_token_accuracy": 0.24153437167406083,
      "num_tokens": 52564568.0,
      "step": 22915
    },
    {
      "entropy": 5.3604785919189455,
      "epoch": 2.201729106628242,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045164376564877734,
      "loss": 4.7875,
      "mean_token_accuracy": 0.2342643767595291,
      "num_tokens": 52577563.0,
      "step": 22920
    },
    {
      "entropy": 5.374658632278442,
      "epoch": 2.202209414024976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045162252870468354,
      "loss": 4.7903,
      "mean_token_accuracy": 0.23996607214212418,
      "num_tokens": 52590295.0,
      "step": 22925
    },
    {
      "entropy": 5.33376727104187,
      "epoch": 2.20268972142171,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004516012876599391,
      "loss": 4.7196,
      "mean_token_accuracy": 0.249732081592083,
      "num_tokens": 52602727.0,
      "step": 22930
    },
    {
      "entropy": 5.31065092086792,
      "epoch": 2.2031700288184437,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045158004251503715,
      "loss": 4.6468,
      "mean_token_accuracy": 0.2560227707028389,
      "num_tokens": 52613422.0,
      "step": 22935
    },
    {
      "entropy": 5.24128155708313,
      "epoch": 2.2036503362151776,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045155879327047087,
      "loss": 4.6685,
      "mean_token_accuracy": 0.2538875937461853,
      "num_tokens": 52624907.0,
      "step": 22940
    },
    {
      "entropy": 5.387598180770874,
      "epoch": 2.2041306436119115,
      "grad_norm": 1.0,
      "learning_rate": 0.0004515375399267338,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23810601979494095,
      "num_tokens": 52635864.0,
      "step": 22945
    },
    {
      "entropy": 5.391049766540528,
      "epoch": 2.2046109510086453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045151628248431925,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23799483627080917,
      "num_tokens": 52647023.0,
      "step": 22950
    },
    {
      "entropy": 5.373448896408081,
      "epoch": 2.2050912584053792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045149502094372077,
      "loss": 4.7932,
      "mean_token_accuracy": 0.2409151777625084,
      "num_tokens": 52658846.0,
      "step": 22955
    },
    {
      "entropy": 5.265308666229248,
      "epoch": 2.2055715658021136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045147375530543195,
      "loss": 4.579,
      "mean_token_accuracy": 0.2594130873680115,
      "num_tokens": 52671363.0,
      "step": 22960
    },
    {
      "entropy": 5.370745468139648,
      "epoch": 2.2060518731988474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045145248556994653,
      "loss": 4.809,
      "mean_token_accuracy": 0.23949736803770066,
      "num_tokens": 52682734.0,
      "step": 22965
    },
    {
      "entropy": 5.311919450759888,
      "epoch": 2.2065321805955813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004514312117377584,
      "loss": 4.6793,
      "mean_token_accuracy": 0.24672659635543823,
      "num_tokens": 52693272.0,
      "step": 22970
    },
    {
      "entropy": 5.3359016418457035,
      "epoch": 2.207012487992315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004514099338093613,
      "loss": 4.7565,
      "mean_token_accuracy": 0.2369769498705864,
      "num_tokens": 52703395.0,
      "step": 22975
    },
    {
      "entropy": 5.303762149810791,
      "epoch": 2.207492795389049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004513886517852496,
      "loss": 4.6512,
      "mean_token_accuracy": 0.2457731693983078,
      "num_tokens": 52715870.0,
      "step": 22980
    },
    {
      "entropy": 5.359052753448486,
      "epoch": 2.207973102785783,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000451367365665917,
      "loss": 4.7431,
      "mean_token_accuracy": 0.2492431178689003,
      "num_tokens": 52726922.0,
      "step": 22985
    },
    {
      "entropy": 5.354035472869873,
      "epoch": 2.208453410182517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045134607545185785,
      "loss": 4.746,
      "mean_token_accuracy": 0.24392662942409515,
      "num_tokens": 52738211.0,
      "step": 22990
    },
    {
      "entropy": 5.3639555931091305,
      "epoch": 2.2089337175792507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004513247811435666,
      "loss": 4.7296,
      "mean_token_accuracy": 0.24407717287540437,
      "num_tokens": 52749267.0,
      "step": 22995
    },
    {
      "entropy": 5.239447069168091,
      "epoch": 2.2094140249759846,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045130348274153735,
      "loss": 4.7029,
      "mean_token_accuracy": 0.24124905616044998,
      "num_tokens": 52760520.0,
      "step": 23000
    },
    {
      "entropy": 5.271156883239746,
      "epoch": 2.2098943323727185,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045128218024626486,
      "loss": 4.6838,
      "mean_token_accuracy": 0.2554504096508026,
      "num_tokens": 52772865.0,
      "step": 23005
    },
    {
      "entropy": 5.378840017318725,
      "epoch": 2.2103746397694524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004512608736582436,
      "loss": 4.7133,
      "mean_token_accuracy": 0.24741441160440444,
      "num_tokens": 52783935.0,
      "step": 23010
    },
    {
      "entropy": 5.252227067947388,
      "epoch": 2.2108549471661862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004512395629779682,
      "loss": 4.6781,
      "mean_token_accuracy": 0.2514296039938927,
      "num_tokens": 52794980.0,
      "step": 23015
    },
    {
      "entropy": 5.380865430831909,
      "epoch": 2.21133525456292,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004512182482059335,
      "loss": 4.825,
      "mean_token_accuracy": 0.23143983483314515,
      "num_tokens": 52806426.0,
      "step": 23020
    },
    {
      "entropy": 5.264483022689819,
      "epoch": 2.211815561959654,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004511969293426343,
      "loss": 4.6659,
      "mean_token_accuracy": 0.251725572347641,
      "num_tokens": 52816832.0,
      "step": 23025
    },
    {
      "entropy": 5.309011554718017,
      "epoch": 2.212295869356388,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045117560638856567,
      "loss": 4.715,
      "mean_token_accuracy": 0.24549970626831055,
      "num_tokens": 52829913.0,
      "step": 23030
    },
    {
      "entropy": 5.371594572067261,
      "epoch": 2.212776176753122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004511542793442225,
      "loss": 4.7676,
      "mean_token_accuracy": 0.2428943932056427,
      "num_tokens": 52841300.0,
      "step": 23035
    },
    {
      "entropy": 5.3253766059875485,
      "epoch": 2.213256484149856,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004511329482101001,
      "loss": 4.6784,
      "mean_token_accuracy": 0.24532926827669144,
      "num_tokens": 52852862.0,
      "step": 23040
    },
    {
      "entropy": 5.353502082824707,
      "epoch": 2.21373679154659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004511116129866936,
      "loss": 4.7444,
      "mean_token_accuracy": 0.24086736142635345,
      "num_tokens": 52863499.0,
      "step": 23045
    },
    {
      "entropy": 5.361634349822998,
      "epoch": 2.214217098943324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045109027367449845,
      "loss": 4.7901,
      "mean_token_accuracy": 0.24840784072875977,
      "num_tokens": 52875459.0,
      "step": 23050
    },
    {
      "entropy": 5.3112444400787355,
      "epoch": 2.2146974063400577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045106893027400995,
      "loss": 4.6888,
      "mean_token_accuracy": 0.24491393566131592,
      "num_tokens": 52886758.0,
      "step": 23055
    },
    {
      "entropy": 5.344529724121093,
      "epoch": 2.2151777137367916,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045104758278572375,
      "loss": 4.6883,
      "mean_token_accuracy": 0.24715795367956161,
      "num_tokens": 52898703.0,
      "step": 23060
    },
    {
      "entropy": 5.367285966873169,
      "epoch": 2.2156580211335255,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004510262312101355,
      "loss": 4.7324,
      "mean_token_accuracy": 0.2395428776741028,
      "num_tokens": 52910455.0,
      "step": 23065
    },
    {
      "entropy": 5.240573453903198,
      "epoch": 2.2161383285302594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004510048755477407,
      "loss": 4.6388,
      "mean_token_accuracy": 0.24654879420995712,
      "num_tokens": 52921992.0,
      "step": 23070
    },
    {
      "entropy": 5.333510446548462,
      "epoch": 2.2166186359269933,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004509835157990354,
      "loss": 4.7654,
      "mean_token_accuracy": 0.24097324907779694,
      "num_tokens": 52934994.0,
      "step": 23075
    },
    {
      "entropy": 5.386312341690063,
      "epoch": 2.217098943323727,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045096215196451547,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23370649814605712,
      "num_tokens": 52945745.0,
      "step": 23080
    },
    {
      "entropy": 5.385443878173828,
      "epoch": 2.217579250720461,
      "grad_norm": 1.0,
      "learning_rate": 0.00045094078404467683,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23386477530002595,
      "num_tokens": 52957470.0,
      "step": 23085
    },
    {
      "entropy": 5.4093241691589355,
      "epoch": 2.218059558117195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045091941204001564,
      "loss": 4.7478,
      "mean_token_accuracy": 0.24565932005643845,
      "num_tokens": 52968584.0,
      "step": 23090
    },
    {
      "entropy": 5.345236492156983,
      "epoch": 2.218539865513929,
      "grad_norm": 1.03125,
      "learning_rate": 0.000450898035951028,
      "loss": 4.7388,
      "mean_token_accuracy": 0.24143145382404327,
      "num_tokens": 52980187.0,
      "step": 23095
    },
    {
      "entropy": 5.333146667480468,
      "epoch": 2.2190201729106627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045087665577821034,
      "loss": 4.687,
      "mean_token_accuracy": 0.24871020764112473,
      "num_tokens": 52990478.0,
      "step": 23100
    },
    {
      "entropy": 5.340623188018799,
      "epoch": 2.2195004803073966,
      "grad_norm": 1.15625,
      "learning_rate": 0.000450855271522059,
      "loss": 4.8159,
      "mean_token_accuracy": 0.2326240658760071,
      "num_tokens": 53001130.0,
      "step": 23105
    },
    {
      "entropy": 5.390967226028442,
      "epoch": 2.219980787704131,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045083388318307044,
      "loss": 4.7913,
      "mean_token_accuracy": 0.2414316400885582,
      "num_tokens": 53013057.0,
      "step": 23110
    },
    {
      "entropy": 5.432517623901367,
      "epoch": 2.2204610951008648,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004508124907617411,
      "loss": 4.815,
      "mean_token_accuracy": 0.2367015391588211,
      "num_tokens": 53024333.0,
      "step": 23115
    },
    {
      "entropy": 5.306591796875,
      "epoch": 2.2209414024975986,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004507910942585679,
      "loss": 4.7414,
      "mean_token_accuracy": 0.24271784126758575,
      "num_tokens": 53036723.0,
      "step": 23120
    },
    {
      "entropy": 5.374690008163452,
      "epoch": 2.2214217098943325,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507696936740475,
      "loss": 4.7974,
      "mean_token_accuracy": 0.24477449655532837,
      "num_tokens": 53048183.0,
      "step": 23125
    },
    {
      "entropy": 5.348724985122681,
      "epoch": 2.2219020172910664,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004507482890086767,
      "loss": 4.7516,
      "mean_token_accuracy": 0.2494152456521988,
      "num_tokens": 53061467.0,
      "step": 23130
    },
    {
      "entropy": 5.330927324295044,
      "epoch": 2.2223823246878003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004507268802629525,
      "loss": 4.7421,
      "mean_token_accuracy": 0.24790896624326705,
      "num_tokens": 53072833.0,
      "step": 23135
    },
    {
      "entropy": 5.352190399169922,
      "epoch": 2.222862632084534,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004507054674373719,
      "loss": 4.6972,
      "mean_token_accuracy": 0.24278714507818222,
      "num_tokens": 53084109.0,
      "step": 23140
    },
    {
      "entropy": 5.346156978607178,
      "epoch": 2.223342939481268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045068405053243216,
      "loss": 4.7349,
      "mean_token_accuracy": 0.24685042649507521,
      "num_tokens": 53095997.0,
      "step": 23145
    },
    {
      "entropy": 5.289363670349121,
      "epoch": 2.223823246878002,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004506626295486304,
      "loss": 4.6618,
      "mean_token_accuracy": 0.25436079800128936,
      "num_tokens": 53107518.0,
      "step": 23150
    },
    {
      "entropy": 5.321631813049317,
      "epoch": 2.224303554274736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045064120448646405,
      "loss": 4.7303,
      "mean_token_accuracy": 0.24335800260305404,
      "num_tokens": 53118908.0,
      "step": 23155
    },
    {
      "entropy": 5.349285030364991,
      "epoch": 2.2247838616714697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004506197753464304,
      "loss": 4.7158,
      "mean_token_accuracy": 0.24847399145364762,
      "num_tokens": 53129404.0,
      "step": 23160
    },
    {
      "entropy": 5.386111116409301,
      "epoch": 2.2252641690682036,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045059834212902707,
      "loss": 4.8077,
      "mean_token_accuracy": 0.23944285660982131,
      "num_tokens": 53141364.0,
      "step": 23165
    },
    {
      "entropy": 5.395674085617065,
      "epoch": 2.2257444764649374,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045057690483475167,
      "loss": 4.7102,
      "mean_token_accuracy": 0.2496615782380104,
      "num_tokens": 53153617.0,
      "step": 23170
    },
    {
      "entropy": 5.366702651977539,
      "epoch": 2.2262247838616713,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004505554634641019,
      "loss": 4.8097,
      "mean_token_accuracy": 0.23488749861717223,
      "num_tokens": 53166537.0,
      "step": 23175
    },
    {
      "entropy": 5.337429428100586,
      "epoch": 2.226705091258405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045053401801757554,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24137877970933913,
      "num_tokens": 53178979.0,
      "step": 23180
    },
    {
      "entropy": 5.360771703720093,
      "epoch": 2.227185398655139,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045051256849567054,
      "loss": 4.7597,
      "mean_token_accuracy": 0.24592569023370742,
      "num_tokens": 53190177.0,
      "step": 23185
    },
    {
      "entropy": 5.34479341506958,
      "epoch": 2.227665706051873,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045049111489888486,
      "loss": 4.7257,
      "mean_token_accuracy": 0.24291598051786423,
      "num_tokens": 53201398.0,
      "step": 23190
    },
    {
      "entropy": 5.3729438304901125,
      "epoch": 2.2281460134486073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004504696572277165,
      "loss": 4.7667,
      "mean_token_accuracy": 0.23934225291013717,
      "num_tokens": 53213795.0,
      "step": 23195
    },
    {
      "entropy": 5.335114049911499,
      "epoch": 2.228626320845341,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045044819548266385,
      "loss": 4.8064,
      "mean_token_accuracy": 0.23378079980611802,
      "num_tokens": 53226831.0,
      "step": 23200
    },
    {
      "entropy": 5.425396347045899,
      "epoch": 2.229106628242075,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045042672966422506,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2387322187423706,
      "num_tokens": 53238558.0,
      "step": 23205
    },
    {
      "entropy": 5.369731187820435,
      "epoch": 2.229586935638809,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045040525977289847,
      "loss": 4.7344,
      "mean_token_accuracy": 0.24020297229290008,
      "num_tokens": 53250376.0,
      "step": 23210
    },
    {
      "entropy": 5.328532934188843,
      "epoch": 2.230067243035543,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004503837858091826,
      "loss": 4.7202,
      "mean_token_accuracy": 0.24482488483190537,
      "num_tokens": 53260872.0,
      "step": 23215
    },
    {
      "entropy": 5.335174179077148,
      "epoch": 2.2305475504322767,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045036230777357604,
      "loss": 4.7134,
      "mean_token_accuracy": 0.24352657794952393,
      "num_tokens": 53271363.0,
      "step": 23220
    },
    {
      "entropy": 5.403095722198486,
      "epoch": 2.2310278578290106,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004503408256665774,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2339477136731148,
      "num_tokens": 53282356.0,
      "step": 23225
    },
    {
      "entropy": 5.338918733596802,
      "epoch": 2.2315081652257445,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045031933948868545,
      "loss": 4.7584,
      "mean_token_accuracy": 0.24337849915027618,
      "num_tokens": 53293675.0,
      "step": 23230
    },
    {
      "entropy": 5.276245737075806,
      "epoch": 2.2319884726224783,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045029784924039903,
      "loss": 4.6597,
      "mean_token_accuracy": 0.2549406155943871,
      "num_tokens": 53305193.0,
      "step": 23235
    },
    {
      "entropy": 5.356119680404663,
      "epoch": 2.2324687800192122,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045027635492221716,
      "loss": 4.6617,
      "mean_token_accuracy": 0.2534694135189056,
      "num_tokens": 53315729.0,
      "step": 23240
    },
    {
      "entropy": 5.318604373931885,
      "epoch": 2.232949087415946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045025485653463866,
      "loss": 4.7368,
      "mean_token_accuracy": 0.24454630464315413,
      "num_tokens": 53327220.0,
      "step": 23245
    },
    {
      "entropy": 5.304111289978027,
      "epoch": 2.23342939481268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004502333540781628,
      "loss": 4.7734,
      "mean_token_accuracy": 0.23982256799936294,
      "num_tokens": 53337565.0,
      "step": 23250
    },
    {
      "entropy": 5.382833528518677,
      "epoch": 2.233909702209414,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004502118475532888,
      "loss": 4.753,
      "mean_token_accuracy": 0.24307248294353484,
      "num_tokens": 53348791.0,
      "step": 23255
    },
    {
      "entropy": 5.313071155548096,
      "epoch": 2.2343900096061478,
      "grad_norm": 1.078125,
      "learning_rate": 0.000450190336960516,
      "loss": 4.695,
      "mean_token_accuracy": 0.2454131841659546,
      "num_tokens": 53361341.0,
      "step": 23260
    },
    {
      "entropy": 5.33474817276001,
      "epoch": 2.2348703170028816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004501688223003438,
      "loss": 4.772,
      "mean_token_accuracy": 0.24170411229133607,
      "num_tokens": 53372085.0,
      "step": 23265
    },
    {
      "entropy": 5.329816007614136,
      "epoch": 2.235350624399616,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004501473035732717,
      "loss": 4.7153,
      "mean_token_accuracy": 0.2425371825695038,
      "num_tokens": 53383665.0,
      "step": 23270
    },
    {
      "entropy": 5.320078897476196,
      "epoch": 2.23583093179635,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004501257807797993,
      "loss": 4.733,
      "mean_token_accuracy": 0.23795249611139296,
      "num_tokens": 53394697.0,
      "step": 23275
    },
    {
      "entropy": 5.333355951309204,
      "epoch": 2.2363112391930837,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045010425392042624,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24439499527215958,
      "num_tokens": 53405538.0,
      "step": 23280
    },
    {
      "entropy": 5.452707767486572,
      "epoch": 2.2367915465898176,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004500827229956524,
      "loss": 4.8456,
      "mean_token_accuracy": 0.23163097500801086,
      "num_tokens": 53416515.0,
      "step": 23285
    },
    {
      "entropy": 5.339598608016968,
      "epoch": 2.2372718539865515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045006118800597757,
      "loss": 4.6969,
      "mean_token_accuracy": 0.23797043710947036,
      "num_tokens": 53426530.0,
      "step": 23290
    },
    {
      "entropy": 5.3270806789398195,
      "epoch": 2.2377521613832854,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045003964895190177,
      "loss": 4.8047,
      "mean_token_accuracy": 0.24178521782159806,
      "num_tokens": 53438514.0,
      "step": 23295
    },
    {
      "entropy": 5.299796009063721,
      "epoch": 2.2382324687800192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004500181058339251,
      "loss": 4.6671,
      "mean_token_accuracy": 0.2517859622836113,
      "num_tokens": 53448845.0,
      "step": 23300
    },
    {
      "entropy": 5.363061237335205,
      "epoch": 2.238712776176753,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004499965586525478,
      "loss": 4.7735,
      "mean_token_accuracy": 0.24334441274404525,
      "num_tokens": 53460526.0,
      "step": 23305
    },
    {
      "entropy": 5.456817150115967,
      "epoch": 2.239193083573487,
      "grad_norm": 1.125,
      "learning_rate": 0.00044997500740826993,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23884449005126954,
      "num_tokens": 53471997.0,
      "step": 23310
    },
    {
      "entropy": 5.367876768112183,
      "epoch": 2.239673390970221,
      "grad_norm": 1.109375,
      "learning_rate": 0.000449953452101592,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23704839050769805,
      "num_tokens": 53483097.0,
      "step": 23315
    },
    {
      "entropy": 5.31870436668396,
      "epoch": 2.2401536983669548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044993189273301445,
      "loss": 4.6948,
      "mean_token_accuracy": 0.2552516385912895,
      "num_tokens": 53494877.0,
      "step": 23320
    },
    {
      "entropy": 5.313142442703247,
      "epoch": 2.2406340057636887,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004499103293030377,
      "loss": 4.6994,
      "mean_token_accuracy": 0.24704933911561966,
      "num_tokens": 53507065.0,
      "step": 23325
    },
    {
      "entropy": 5.40010781288147,
      "epoch": 2.2411143131604225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004498887618121625,
      "loss": 4.7783,
      "mean_token_accuracy": 0.24233101755380632,
      "num_tokens": 53518188.0,
      "step": 23330
    },
    {
      "entropy": 5.295528364181519,
      "epoch": 2.2415946205571564,
      "grad_norm": 1.125,
      "learning_rate": 0.0004498671902608897,
      "loss": 4.6447,
      "mean_token_accuracy": 0.2481561154127121,
      "num_tokens": 53529872.0,
      "step": 23335
    },
    {
      "entropy": 5.343093204498291,
      "epoch": 2.2420749279538903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004498456146497198,
      "loss": 4.7301,
      "mean_token_accuracy": 0.23688321709632873,
      "num_tokens": 53540175.0,
      "step": 23340
    },
    {
      "entropy": 5.276356029510498,
      "epoch": 2.2425552353506246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044982403497915405,
      "loss": 4.6575,
      "mean_token_accuracy": 0.24505036622285842,
      "num_tokens": 53552520.0,
      "step": 23345
    },
    {
      "entropy": 5.347867679595947,
      "epoch": 2.2430355427473585,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044980245124969333,
      "loss": 4.7461,
      "mean_token_accuracy": 0.24161538034677504,
      "num_tokens": 53563937.0,
      "step": 23350
    },
    {
      "entropy": 5.357225561141968,
      "epoch": 2.2435158501440924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004497808634618386,
      "loss": 4.7074,
      "mean_token_accuracy": 0.2407086670398712,
      "num_tokens": 53574654.0,
      "step": 23355
    },
    {
      "entropy": 5.34192385673523,
      "epoch": 2.2439961575408263,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004497592716160913,
      "loss": 4.7503,
      "mean_token_accuracy": 0.24492717087268828,
      "num_tokens": 53586848.0,
      "step": 23360
    },
    {
      "entropy": 5.319420099258423,
      "epoch": 2.24447646493756,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044973767571295273,
      "loss": 4.7316,
      "mean_token_accuracy": 0.2461989238858223,
      "num_tokens": 53598200.0,
      "step": 23365
    },
    {
      "entropy": 5.331506013870239,
      "epoch": 2.244956772334294,
      "grad_norm": 1.0,
      "learning_rate": 0.0004497160757529241,
      "loss": 4.6906,
      "mean_token_accuracy": 0.24866973459720612,
      "num_tokens": 53609957.0,
      "step": 23370
    },
    {
      "entropy": 5.350235843658448,
      "epoch": 2.245437079731028,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044969447173650695,
      "loss": 4.752,
      "mean_token_accuracy": 0.2506559893488884,
      "num_tokens": 53621029.0,
      "step": 23375
    },
    {
      "entropy": 5.337652683258057,
      "epoch": 2.245917387127762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000449672863664203,
      "loss": 4.7335,
      "mean_token_accuracy": 0.24626590311527252,
      "num_tokens": 53633279.0,
      "step": 23380
    },
    {
      "entropy": 5.355899047851563,
      "epoch": 2.2463976945244957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044965125153651375,
      "loss": 4.7145,
      "mean_token_accuracy": 0.24732736498117447,
      "num_tokens": 53643959.0,
      "step": 23385
    },
    {
      "entropy": 5.31103024482727,
      "epoch": 2.2468780019212296,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004496296353539411,
      "loss": 4.7189,
      "mean_token_accuracy": 0.24452867209911347,
      "num_tokens": 53655215.0,
      "step": 23390
    },
    {
      "entropy": 5.369206523895263,
      "epoch": 2.2473583093179634,
      "grad_norm": 1.125,
      "learning_rate": 0.0004496080151169869,
      "loss": 4.7291,
      "mean_token_accuracy": 0.24546034187078475,
      "num_tokens": 53666557.0,
      "step": 23395
    },
    {
      "entropy": 5.383982610702515,
      "epoch": 2.2478386167146973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044958639082615294,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23893296718597412,
      "num_tokens": 53679626.0,
      "step": 23400
    },
    {
      "entropy": 5.325354862213135,
      "epoch": 2.248318924111431,
      "grad_norm": 1.0,
      "learning_rate": 0.0004495647624819415,
      "loss": 4.7274,
      "mean_token_accuracy": 0.24797680824995041,
      "num_tokens": 53691681.0,
      "step": 23405
    },
    {
      "entropy": 5.325044965744018,
      "epoch": 2.248799231508165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044954313008485457,
      "loss": 4.6875,
      "mean_token_accuracy": 0.24806791096925734,
      "num_tokens": 53703320.0,
      "step": 23410
    },
    {
      "entropy": 5.300332641601562,
      "epoch": 2.249279538904899,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044952149363539453,
      "loss": 4.6773,
      "mean_token_accuracy": 0.24678549766540528,
      "num_tokens": 53714867.0,
      "step": 23415
    },
    {
      "entropy": 5.338631105422974,
      "epoch": 2.249759846301633,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004494998531340635,
      "loss": 4.786,
      "mean_token_accuracy": 0.24428430646657945,
      "num_tokens": 53726121.0,
      "step": 23420
    },
    {
      "entropy": 5.378246307373047,
      "epoch": 2.2502401536983667,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004494782085813641,
      "loss": 4.7764,
      "mean_token_accuracy": 0.23898318111896516,
      "num_tokens": 53736966.0,
      "step": 23425
    },
    {
      "entropy": 5.303269243240356,
      "epoch": 2.250720461095101,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004494565599777988,
      "loss": 4.6733,
      "mean_token_accuracy": 0.251683734357357,
      "num_tokens": 53747446.0,
      "step": 23430
    },
    {
      "entropy": 5.3215718269348145,
      "epoch": 2.251200768491835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044943490732387025,
      "loss": 4.7908,
      "mean_token_accuracy": 0.2306261509656906,
      "num_tokens": 53759588.0,
      "step": 23435
    },
    {
      "entropy": 5.3648905754089355,
      "epoch": 2.251681075888569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004494132506200811,
      "loss": 4.7897,
      "mean_token_accuracy": 0.2385075196623802,
      "num_tokens": 53770314.0,
      "step": 23440
    },
    {
      "entropy": 5.331382942199707,
      "epoch": 2.2521613832853027,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004493915898669341,
      "loss": 4.7039,
      "mean_token_accuracy": 0.24989967942237853,
      "num_tokens": 53781251.0,
      "step": 23445
    },
    {
      "entropy": 5.424556303024292,
      "epoch": 2.2526416906820366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493699250649323,
      "loss": 4.8135,
      "mean_token_accuracy": 0.23567416816949843,
      "num_tokens": 53792676.0,
      "step": 23450
    },
    {
      "entropy": 5.427659034729004,
      "epoch": 2.2531219980787704,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004493482562145786,
      "loss": 4.804,
      "mean_token_accuracy": 0.23704078495502473,
      "num_tokens": 53802714.0,
      "step": 23455
    },
    {
      "entropy": 5.309996032714844,
      "epoch": 2.2536023054755043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044932658331637605,
      "loss": 4.7675,
      "mean_token_accuracy": 0.2380969762802124,
      "num_tokens": 53815043.0,
      "step": 23460
    },
    {
      "entropy": 5.353374099731445,
      "epoch": 2.254082612872238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004493049063708279,
      "loss": 4.7517,
      "mean_token_accuracy": 0.23445727527141572,
      "num_tokens": 53827903.0,
      "step": 23465
    },
    {
      "entropy": 5.41792893409729,
      "epoch": 2.254562920268972,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044928322537843746,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23375332057476045,
      "num_tokens": 53839574.0,
      "step": 23470
    },
    {
      "entropy": 5.353459787368775,
      "epoch": 2.255043227665706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044926154033970793,
      "loss": 4.6862,
      "mean_token_accuracy": 0.2532571882009506,
      "num_tokens": 53850488.0,
      "step": 23475
    },
    {
      "entropy": 5.354104614257812,
      "epoch": 2.25552353506244,
      "grad_norm": 1.109375,
      "learning_rate": 0.000449239851255143,
      "loss": 4.7008,
      "mean_token_accuracy": 0.24493561685085297,
      "num_tokens": 53860766.0,
      "step": 23480
    },
    {
      "entropy": 5.282110595703125,
      "epoch": 2.2560038424591737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044921815812524606,
      "loss": 4.7116,
      "mean_token_accuracy": 0.24870155304670333,
      "num_tokens": 53872088.0,
      "step": 23485
    },
    {
      "entropy": 5.382993078231811,
      "epoch": 2.2564841498559076,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044919646095052077,
      "loss": 4.7523,
      "mean_token_accuracy": 0.24406830221414566,
      "num_tokens": 53883302.0,
      "step": 23490
    },
    {
      "entropy": 5.236488199234008,
      "epoch": 2.2569644572526415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000449174759731471,
      "loss": 4.6367,
      "mean_token_accuracy": 0.2569802224636078,
      "num_tokens": 53894310.0,
      "step": 23495
    },
    {
      "entropy": 5.344521188735962,
      "epoch": 2.2574447646493754,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044915305446860046,
      "loss": 4.7739,
      "mean_token_accuracy": 0.23661978244781495,
      "num_tokens": 53906760.0,
      "step": 23500
    },
    {
      "entropy": 5.328992033004761,
      "epoch": 2.2579250720461097,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044913134516241305,
      "loss": 4.7642,
      "mean_token_accuracy": 0.241257044672966,
      "num_tokens": 53918758.0,
      "step": 23505
    },
    {
      "entropy": 5.298448705673218,
      "epoch": 2.2584053794428436,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004491096318134129,
      "loss": 4.681,
      "mean_token_accuracy": 0.2518308058381081,
      "num_tokens": 53932574.0,
      "step": 23510
    },
    {
      "entropy": 5.375859498977661,
      "epoch": 2.2588856868395775,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004490879144221042,
      "loss": 4.7933,
      "mean_token_accuracy": 0.24558283090591432,
      "num_tokens": 53943801.0,
      "step": 23515
    },
    {
      "entropy": 5.314622402191162,
      "epoch": 2.2593659942363113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044906619298899097,
      "loss": 4.7215,
      "mean_token_accuracy": 0.2518452301621437,
      "num_tokens": 53955458.0,
      "step": 23520
    },
    {
      "entropy": 5.367046260833741,
      "epoch": 2.2598463016330452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004490444675145777,
      "loss": 4.7629,
      "mean_token_accuracy": 0.24050410687923432,
      "num_tokens": 53967242.0,
      "step": 23525
    },
    {
      "entropy": 5.366106176376343,
      "epoch": 2.260326609029779,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004490227379993686,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2427014857530594,
      "num_tokens": 53978953.0,
      "step": 23530
    },
    {
      "entropy": 5.369247961044311,
      "epoch": 2.260806916426513,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004490010044438683,
      "loss": 4.753,
      "mean_token_accuracy": 0.24166400581598282,
      "num_tokens": 53990553.0,
      "step": 23535
    },
    {
      "entropy": 5.261774301528931,
      "epoch": 2.261287223823247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044897926684858133,
      "loss": 4.7756,
      "mean_token_accuracy": 0.2432629033923149,
      "num_tokens": 54002187.0,
      "step": 23540
    },
    {
      "entropy": 5.369328737258911,
      "epoch": 2.2617675312199808,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044895752521401246,
      "loss": 4.7398,
      "mean_token_accuracy": 0.23905674368143082,
      "num_tokens": 54014202.0,
      "step": 23545
    },
    {
      "entropy": 5.349704313278198,
      "epoch": 2.2622478386167146,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004489357795406663,
      "loss": 4.7406,
      "mean_token_accuracy": 0.245615653693676,
      "num_tokens": 54026286.0,
      "step": 23550
    },
    {
      "entropy": 5.32173433303833,
      "epoch": 2.2627281460134485,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004489140298290479,
      "loss": 4.7531,
      "mean_token_accuracy": 0.24413614422082902,
      "num_tokens": 54038845.0,
      "step": 23555
    },
    {
      "entropy": 5.430362558364868,
      "epoch": 2.2632084534101824,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044889227607966217,
      "loss": 4.8139,
      "mean_token_accuracy": 0.24234387427568435,
      "num_tokens": 54049991.0,
      "step": 23560
    },
    {
      "entropy": 5.349924564361572,
      "epoch": 2.2636887608069163,
      "grad_norm": 1.125,
      "learning_rate": 0.00044887051829301406,
      "loss": 4.7104,
      "mean_token_accuracy": 0.24156416654586793,
      "num_tokens": 54061408.0,
      "step": 23565
    },
    {
      "entropy": 5.351504135131836,
      "epoch": 2.26416906820365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044884875646960886,
      "loss": 4.7119,
      "mean_token_accuracy": 0.241303688287735,
      "num_tokens": 54072670.0,
      "step": 23570
    },
    {
      "entropy": 5.34902811050415,
      "epoch": 2.264649375600384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044882699060995175,
      "loss": 4.7239,
      "mean_token_accuracy": 0.2450388863682747,
      "num_tokens": 54084555.0,
      "step": 23575
    },
    {
      "entropy": 5.3010035991668705,
      "epoch": 2.2651296829971184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004488052207145481,
      "loss": 4.6919,
      "mean_token_accuracy": 0.2530975595116615,
      "num_tokens": 54096764.0,
      "step": 23580
    },
    {
      "entropy": 5.426944637298584,
      "epoch": 2.2656099903938522,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044878344678390324,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24917005598545075,
      "num_tokens": 54109101.0,
      "step": 23585
    },
    {
      "entropy": 5.41650710105896,
      "epoch": 2.266090297790586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044876166881852286,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23615064918994905,
      "num_tokens": 54120217.0,
      "step": 23590
    },
    {
      "entropy": 5.403183031082153,
      "epoch": 2.26657060518732,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004487398868189125,
      "loss": 4.7577,
      "mean_token_accuracy": 0.24235844761133193,
      "num_tokens": 54132059.0,
      "step": 23595
    },
    {
      "entropy": 5.407739686965942,
      "epoch": 2.267050912584054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044871810078557777,
      "loss": 4.8087,
      "mean_token_accuracy": 0.23408226668834686,
      "num_tokens": 54141928.0,
      "step": 23600
    },
    {
      "entropy": 5.360576391220093,
      "epoch": 2.2675312199807878,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004486963107190247,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23751098960638045,
      "num_tokens": 54153052.0,
      "step": 23605
    },
    {
      "entropy": 5.438371467590332,
      "epoch": 2.2680115273775217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044867451661975894,
      "loss": 4.7781,
      "mean_token_accuracy": 0.24041403084993362,
      "num_tokens": 54163539.0,
      "step": 23610
    },
    {
      "entropy": 5.301913738250732,
      "epoch": 2.2684918347742555,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044865271848828673,
      "loss": 4.7142,
      "mean_token_accuracy": 0.24435058385133743,
      "num_tokens": 54176313.0,
      "step": 23615
    },
    {
      "entropy": 5.388664674758911,
      "epoch": 2.2689721421709894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000448630916325114,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23428455740213394,
      "num_tokens": 54187080.0,
      "step": 23620
    },
    {
      "entropy": 5.3749284744262695,
      "epoch": 2.2694524495677233,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448609110130747,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2403099939227104,
      "num_tokens": 54197856.0,
      "step": 23625
    },
    {
      "entropy": 5.401267194747925,
      "epoch": 2.269932756964457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044858729990569193,
      "loss": 4.7681,
      "mean_token_accuracy": 0.23720307797193527,
      "num_tokens": 54210865.0,
      "step": 23630
    },
    {
      "entropy": 5.290365743637085,
      "epoch": 2.270413064361191,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044856548565045523,
      "loss": 4.6719,
      "mean_token_accuracy": 0.24578242003917694,
      "num_tokens": 54222719.0,
      "step": 23635
    },
    {
      "entropy": 5.364580774307251,
      "epoch": 2.270893371757925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044854366736554323,
      "loss": 4.7189,
      "mean_token_accuracy": 0.24885116815567015,
      "num_tokens": 54233874.0,
      "step": 23640
    },
    {
      "entropy": 5.287834692001343,
      "epoch": 2.271373679154659,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044852184505146274,
      "loss": 4.6333,
      "mean_token_accuracy": 0.24748673886060715,
      "num_tokens": 54245966.0,
      "step": 23645
    },
    {
      "entropy": 5.334021186828613,
      "epoch": 2.2718539865513927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004485000187087202,
      "loss": 4.7101,
      "mean_token_accuracy": 0.24738356173038484,
      "num_tokens": 54256468.0,
      "step": 23650
    },
    {
      "entropy": 5.300715398788452,
      "epoch": 2.272334293948127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004484781883378224,
      "loss": 4.6632,
      "mean_token_accuracy": 0.23927871286869049,
      "num_tokens": 54266362.0,
      "step": 23655
    },
    {
      "entropy": 5.3692946434021,
      "epoch": 2.2728146013448605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044845635393927623,
      "loss": 4.8472,
      "mean_token_accuracy": 0.23559024930000305,
      "num_tokens": 54278365.0,
      "step": 23660
    },
    {
      "entropy": 5.36572265625,
      "epoch": 2.273294908741595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004484345155135886,
      "loss": 4.8343,
      "mean_token_accuracy": 0.23914045840501785,
      "num_tokens": 54289310.0,
      "step": 23665
    },
    {
      "entropy": 5.320562744140625,
      "epoch": 2.2737752161383287,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004484126730612664,
      "loss": 4.688,
      "mean_token_accuracy": 0.24679160863161087,
      "num_tokens": 54300421.0,
      "step": 23670
    },
    {
      "entropy": 5.31551456451416,
      "epoch": 2.2742555235350626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000448390826582817,
      "loss": 4.7441,
      "mean_token_accuracy": 0.2465657040476799,
      "num_tokens": 54312803.0,
      "step": 23675
    },
    {
      "entropy": 5.381476402282715,
      "epoch": 2.2747358309317964,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044836897607874744,
      "loss": 4.7802,
      "mean_token_accuracy": 0.2416289657354355,
      "num_tokens": 54324651.0,
      "step": 23680
    },
    {
      "entropy": 5.3361241817474365,
      "epoch": 2.2752161383285303,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000448347121549565,
      "loss": 4.6802,
      "mean_token_accuracy": 0.25255422592163085,
      "num_tokens": 54334841.0,
      "step": 23685
    },
    {
      "entropy": 5.345450019836425,
      "epoch": 2.275696445725264,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004483252629957771,
      "loss": 4.7247,
      "mean_token_accuracy": 0.24403428435325622,
      "num_tokens": 54346685.0,
      "step": 23690
    },
    {
      "entropy": 5.313568830490112,
      "epoch": 2.276176753121998,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044830340041789133,
      "loss": 4.7752,
      "mean_token_accuracy": 0.2363591730594635,
      "num_tokens": 54358710.0,
      "step": 23695
    },
    {
      "entropy": 5.343270111083984,
      "epoch": 2.276657060518732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004482815338164152,
      "loss": 4.7928,
      "mean_token_accuracy": 0.24469198286533356,
      "num_tokens": 54369695.0,
      "step": 23700
    },
    {
      "entropy": 5.2951884269714355,
      "epoch": 2.277137367915466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004482596631918564,
      "loss": 4.7076,
      "mean_token_accuracy": 0.24356389045715332,
      "num_tokens": 54381605.0,
      "step": 23705
    },
    {
      "entropy": 5.360122108459473,
      "epoch": 2.2776176753121997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044823778854472267,
      "loss": 4.7403,
      "mean_token_accuracy": 0.24809530079364778,
      "num_tokens": 54392309.0,
      "step": 23710
    },
    {
      "entropy": 5.3477396965026855,
      "epoch": 2.2780979827089336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004482159098755219,
      "loss": 4.7545,
      "mean_token_accuracy": 0.24452533423900605,
      "num_tokens": 54403195.0,
      "step": 23715
    },
    {
      "entropy": 5.27873387336731,
      "epoch": 2.2785782901056675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004481940271847621,
      "loss": 4.702,
      "mean_token_accuracy": 0.2540816828608513,
      "num_tokens": 54415105.0,
      "step": 23720
    },
    {
      "entropy": 5.300930166244507,
      "epoch": 2.2790585975024014,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004481721404729513,
      "loss": 4.7242,
      "mean_token_accuracy": 0.24208489656448365,
      "num_tokens": 54425798.0,
      "step": 23725
    },
    {
      "entropy": 5.366203641891479,
      "epoch": 2.2795389048991357,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481502497405975,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23961416631937027,
      "num_tokens": 54437254.0,
      "step": 23730
    },
    {
      "entropy": 5.345062160491944,
      "epoch": 2.280019212295869,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004481283549882091,
      "loss": 4.7482,
      "mean_token_accuracy": 0.23063942193984985,
      "num_tokens": 54448152.0,
      "step": 23735
    },
    {
      "entropy": 5.352545928955078,
      "epoch": 2.2804995196926034,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044810645621629443,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2393631160259247,
      "num_tokens": 54460066.0,
      "step": 23740
    },
    {
      "entropy": 5.395090866088867,
      "epoch": 2.2809798270893373,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044808455342536176,
      "loss": 4.802,
      "mean_token_accuracy": 0.23902925997972488,
      "num_tokens": 54471769.0,
      "step": 23745
    },
    {
      "entropy": 5.285472631454468,
      "epoch": 2.281460134486071,
      "grad_norm": 1.125,
      "learning_rate": 0.00044806264661591976,
      "loss": 4.6343,
      "mean_token_accuracy": 0.25429265648126603,
      "num_tokens": 54483630.0,
      "step": 23750
    },
    {
      "entropy": 5.364241027832032,
      "epoch": 2.281940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004480407357884771,
      "loss": 4.7442,
      "mean_token_accuracy": 0.24333203434944153,
      "num_tokens": 54495696.0,
      "step": 23755
    },
    {
      "entropy": 5.345284700393677,
      "epoch": 2.282420749279539,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044801882094354226,
      "loss": 4.7609,
      "mean_token_accuracy": 0.23992364406585692,
      "num_tokens": 54507153.0,
      "step": 23760
    },
    {
      "entropy": 5.383969497680664,
      "epoch": 2.282901056676273,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004479969020816242,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23768450766801835,
      "num_tokens": 54519922.0,
      "step": 23765
    },
    {
      "entropy": 5.329695177078247,
      "epoch": 2.2833813640730067,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044797497920323175,
      "loss": 4.7134,
      "mean_token_accuracy": 0.24701463729143142,
      "num_tokens": 54531857.0,
      "step": 23770
    },
    {
      "entropy": 5.350997447967529,
      "epoch": 2.2838616714697406,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004479530523088739,
      "loss": 4.7114,
      "mean_token_accuracy": 0.24995382726192475,
      "num_tokens": 54542079.0,
      "step": 23775
    },
    {
      "entropy": 5.286996793746948,
      "epoch": 2.2843419788664745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004479311213990598,
      "loss": 4.6991,
      "mean_token_accuracy": 0.2431483745574951,
      "num_tokens": 54552722.0,
      "step": 23780
    },
    {
      "entropy": 5.268832540512085,
      "epoch": 2.2848222862632084,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044790918647429854,
      "loss": 4.6897,
      "mean_token_accuracy": 0.24681592732667923,
      "num_tokens": 54564224.0,
      "step": 23785
    },
    {
      "entropy": 5.3308392524719235,
      "epoch": 2.2853025936599423,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044788724753509935,
      "loss": 4.6958,
      "mean_token_accuracy": 0.2464079663157463,
      "num_tokens": 54575553.0,
      "step": 23790
    },
    {
      "entropy": 5.395200300216675,
      "epoch": 2.285782901056676,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004478653045819717,
      "loss": 4.839,
      "mean_token_accuracy": 0.2356802597641945,
      "num_tokens": 54587315.0,
      "step": 23795
    },
    {
      "entropy": 5.3898317337036135,
      "epoch": 2.28626320845341,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004478433576154249,
      "loss": 4.7585,
      "mean_token_accuracy": 0.24312819689512252,
      "num_tokens": 54598776.0,
      "step": 23800
    },
    {
      "entropy": 5.437557649612427,
      "epoch": 2.286743515850144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004478214066359687,
      "loss": 4.8444,
      "mean_token_accuracy": 0.22870088815689088,
      "num_tokens": 54610272.0,
      "step": 23805
    },
    {
      "entropy": 5.370971965789795,
      "epoch": 2.287223823246878,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044779945164411254,
      "loss": 4.786,
      "mean_token_accuracy": 0.2406330332159996,
      "num_tokens": 54622105.0,
      "step": 23810
    },
    {
      "entropy": 5.393275928497315,
      "epoch": 2.287704130643612,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004477774926403662,
      "loss": 4.7568,
      "mean_token_accuracy": 0.2479838028550148,
      "num_tokens": 54631771.0,
      "step": 23815
    },
    {
      "entropy": 5.324653196334839,
      "epoch": 2.288184438040346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004477555296252396,
      "loss": 4.684,
      "mean_token_accuracy": 0.2485036388039589,
      "num_tokens": 54642724.0,
      "step": 23820
    },
    {
      "entropy": 5.288230562210083,
      "epoch": 2.28866474543708,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044773356259924255,
      "loss": 4.7084,
      "mean_token_accuracy": 0.24496182501316072,
      "num_tokens": 54655126.0,
      "step": 23825
    },
    {
      "entropy": 5.277210474014282,
      "epoch": 2.2891450528338138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044771159156288505,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2514006271958351,
      "num_tokens": 54665334.0,
      "step": 23830
    },
    {
      "entropy": 5.321018505096435,
      "epoch": 2.2896253602305476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004476896165166772,
      "loss": 4.6697,
      "mean_token_accuracy": 0.24406477063894272,
      "num_tokens": 54676205.0,
      "step": 23835
    },
    {
      "entropy": 5.382255887985229,
      "epoch": 2.2901056676272815,
      "grad_norm": 1.125,
      "learning_rate": 0.00044766763746112936,
      "loss": 4.7573,
      "mean_token_accuracy": 0.24403820484876632,
      "num_tokens": 54686586.0,
      "step": 23840
    },
    {
      "entropy": 5.368847990036011,
      "epoch": 2.2905859750240154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004476456543967517,
      "loss": 4.734,
      "mean_token_accuracy": 0.2444797232747078,
      "num_tokens": 54698879.0,
      "step": 23845
    },
    {
      "entropy": 5.32628607749939,
      "epoch": 2.2910662824207493,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044762366732405454,
      "loss": 4.7093,
      "mean_token_accuracy": 0.24288822412490846,
      "num_tokens": 54710140.0,
      "step": 23850
    },
    {
      "entropy": 5.266655445098877,
      "epoch": 2.291546589817483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004476016762435484,
      "loss": 4.6435,
      "mean_token_accuracy": 0.25382926911115644,
      "num_tokens": 54721317.0,
      "step": 23855
    },
    {
      "entropy": 5.422670698165893,
      "epoch": 2.292026897214217,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004475796811557439,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23487236052751542,
      "num_tokens": 54733329.0,
      "step": 23860
    },
    {
      "entropy": 5.4189173698425295,
      "epoch": 2.292507204610951,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044755768206115155,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23477657586336137,
      "num_tokens": 54745230.0,
      "step": 23865
    },
    {
      "entropy": 5.36537013053894,
      "epoch": 2.292987512007685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004475356789602823,
      "loss": 4.7328,
      "mean_token_accuracy": 0.24435377568006517,
      "num_tokens": 54757428.0,
      "step": 23870
    },
    {
      "entropy": 5.274198818206787,
      "epoch": 2.2934678194044187,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044751367185364696,
      "loss": 4.6446,
      "mean_token_accuracy": 0.2520040228962898,
      "num_tokens": 54768886.0,
      "step": 23875
    },
    {
      "entropy": 5.407549524307251,
      "epoch": 2.2939481268011526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044749166074175634,
      "loss": 4.8614,
      "mean_token_accuracy": 0.23235903829336166,
      "num_tokens": 54780243.0,
      "step": 23880
    },
    {
      "entropy": 5.327964830398559,
      "epoch": 2.2944284341978864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044746964562512154,
      "loss": 4.6454,
      "mean_token_accuracy": 0.24786392748355865,
      "num_tokens": 54792462.0,
      "step": 23885
    },
    {
      "entropy": 5.390325975418091,
      "epoch": 2.2949087415946208,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044744762650425376,
      "loss": 4.7795,
      "mean_token_accuracy": 0.23678498417139054,
      "num_tokens": 54804412.0,
      "step": 23890
    },
    {
      "entropy": 5.3675461769104,
      "epoch": 2.2953890489913547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044742560337966415,
      "loss": 4.723,
      "mean_token_accuracy": 0.2534508973360062,
      "num_tokens": 54814910.0,
      "step": 23895
    },
    {
      "entropy": 5.322027254104614,
      "epoch": 2.2958693563880885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000447403576251864,
      "loss": 4.7128,
      "mean_token_accuracy": 0.24548129141330718,
      "num_tokens": 54826500.0,
      "step": 23900
    },
    {
      "entropy": 5.336364984512329,
      "epoch": 2.2963496637848224,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004473815451213648,
      "loss": 4.7452,
      "mean_token_accuracy": 0.24318389892578124,
      "num_tokens": 54837682.0,
      "step": 23905
    },
    {
      "entropy": 5.302153062820435,
      "epoch": 2.2968299711815563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473595099886779,
      "loss": 4.6844,
      "mean_token_accuracy": 0.24647459685802459,
      "num_tokens": 54848937.0,
      "step": 23910
    },
    {
      "entropy": 5.3565184593200685,
      "epoch": 2.29731027857829,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004473374708543151,
      "loss": 4.6513,
      "mean_token_accuracy": 0.25430897027254107,
      "num_tokens": 54859384.0,
      "step": 23915
    },
    {
      "entropy": 5.2942791938781735,
      "epoch": 2.297790585975024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004473154277187878,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2497648850083351,
      "num_tokens": 54871280.0,
      "step": 23920
    },
    {
      "entropy": 5.24927453994751,
      "epoch": 2.298270893371758,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044729338058260805,
      "loss": 4.66,
      "mean_token_accuracy": 0.24797671139240265,
      "num_tokens": 54882699.0,
      "step": 23925
    },
    {
      "entropy": 5.35925326347351,
      "epoch": 2.298751200768492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004472713294462876,
      "loss": 4.7855,
      "mean_token_accuracy": 0.2443837508559227,
      "num_tokens": 54894041.0,
      "step": 23930
    },
    {
      "entropy": 5.422861099243164,
      "epoch": 2.2992315081652257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044724927431033843,
      "loss": 4.7792,
      "mean_token_accuracy": 0.23623384535312653,
      "num_tokens": 54905007.0,
      "step": 23935
    },
    {
      "entropy": 5.3999340534210205,
      "epoch": 2.2997118155619596,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004472272151752725,
      "loss": 4.7821,
      "mean_token_accuracy": 0.2415631502866745,
      "num_tokens": 54915735.0,
      "step": 23940
    },
    {
      "entropy": 5.260418558120728,
      "epoch": 2.3001921229586935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004472051520416022,
      "loss": 4.6086,
      "mean_token_accuracy": 0.25003030747175214,
      "num_tokens": 54926350.0,
      "step": 23945
    },
    {
      "entropy": 5.329766035079956,
      "epoch": 2.3006724303554273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004471830849098395,
      "loss": 4.7722,
      "mean_token_accuracy": 0.23836778849363327,
      "num_tokens": 54938354.0,
      "step": 23950
    },
    {
      "entropy": 5.380752515792847,
      "epoch": 2.3011527377521612,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044716101378049683,
      "loss": 4.765,
      "mean_token_accuracy": 0.2379504084587097,
      "num_tokens": 54949440.0,
      "step": 23955
    },
    {
      "entropy": 5.352747964859009,
      "epoch": 2.301633045148895,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044713893865408667,
      "loss": 4.7002,
      "mean_token_accuracy": 0.2493894249200821,
      "num_tokens": 54961751.0,
      "step": 23960
    },
    {
      "entropy": 5.393712949752808,
      "epoch": 2.3021133525456294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004471168595311215,
      "loss": 4.8795,
      "mean_token_accuracy": 0.22469589561223985,
      "num_tokens": 54975415.0,
      "step": 23965
    },
    {
      "entropy": 5.380953598022461,
      "epoch": 2.302593659942363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044709477641211395,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24965124875307082,
      "num_tokens": 54986599.0,
      "step": 23970
    },
    {
      "entropy": 5.347928857803344,
      "epoch": 2.303073967339097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004470726892975767,
      "loss": 4.7117,
      "mean_token_accuracy": 0.2445939898490906,
      "num_tokens": 54997428.0,
      "step": 23975
    },
    {
      "entropy": 5.364629793167114,
      "epoch": 2.303554274735831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044705059818802255,
      "loss": 4.7977,
      "mean_token_accuracy": 0.23377760499715805,
      "num_tokens": 55008434.0,
      "step": 23980
    },
    {
      "entropy": 5.2761985778808596,
      "epoch": 2.304034582132565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004470285030839644,
      "loss": 4.7201,
      "mean_token_accuracy": 0.24520672708749772,
      "num_tokens": 55019834.0,
      "step": 23985
    },
    {
      "entropy": 5.3234587669372555,
      "epoch": 2.304514889529299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044700640398591526,
      "loss": 4.6994,
      "mean_token_accuracy": 0.2489137604832649,
      "num_tokens": 55030432.0,
      "step": 23990
    },
    {
      "entropy": 5.42881441116333,
      "epoch": 2.3049951969260327,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004469843008943881,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23660646975040436,
      "num_tokens": 55041482.0,
      "step": 23995
    },
    {
      "entropy": 5.361020469665528,
      "epoch": 2.3054755043227666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004469621938098962,
      "loss": 4.639,
      "mean_token_accuracy": 0.2540226235985756,
      "num_tokens": 55051894.0,
      "step": 24000
    },
    {
      "epoch": 2.3054755043227666,
      "eval_entropy": 5.123255483933276,
      "eval_loss": 4.832446575164795,
      "eval_mean_token_accuracy": 0.24735420806990024,
      "eval_num_tokens": 55051894.0,
      "eval_runtime": 20.5095,
      "eval_samples_per_second": 1599.992,
      "eval_steps_per_second": 200.005,
      "step": 24000
    },
    {
      "entropy": 5.386053657531738,
      "epoch": 2.3059558117195005,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004469400827329528,
      "loss": 4.7456,
      "mean_token_accuracy": 0.25078885108232496,
      "num_tokens": 55063907.0,
      "step": 24005
    },
    {
      "entropy": 5.333146953582764,
      "epoch": 2.3064361191162344,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004469179676640712,
      "loss": 4.7287,
      "mean_token_accuracy": 0.24336486160755158,
      "num_tokens": 55076004.0,
      "step": 24010
    },
    {
      "entropy": 5.2794280529022215,
      "epoch": 2.3069164265129682,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468958486037649,
      "loss": 4.6477,
      "mean_token_accuracy": 0.25901556462049485,
      "num_tokens": 55087506.0,
      "step": 24015
    },
    {
      "entropy": 5.361242246627808,
      "epoch": 2.307396733909702,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468737255525474,
      "loss": 4.7361,
      "mean_token_accuracy": 0.23945986926555635,
      "num_tokens": 55099406.0,
      "step": 24020
    },
    {
      "entropy": 5.458822774887085,
      "epoch": 2.307877041306436,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004468515985109324,
      "loss": 4.8216,
      "mean_token_accuracy": 0.23134316354990006,
      "num_tokens": 55111733.0,
      "step": 24025
    },
    {
      "entropy": 5.343498277664184,
      "epoch": 2.30835734870317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004468294674794335,
      "loss": 4.722,
      "mean_token_accuracy": 0.24240820705890656,
      "num_tokens": 55123455.0,
      "step": 24030
    },
    {
      "entropy": 5.369562482833862,
      "epoch": 2.3088376560999038,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004468073324585647,
      "loss": 4.7592,
      "mean_token_accuracy": 0.23715390264987946,
      "num_tokens": 55134284.0,
      "step": 24035
    },
    {
      "entropy": 5.36471700668335,
      "epoch": 2.309317963496638,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004467851934488398,
      "loss": 4.7257,
      "mean_token_accuracy": 0.24842528849840165,
      "num_tokens": 55147224.0,
      "step": 24040
    },
    {
      "entropy": 5.317326450347901,
      "epoch": 2.3097982708933715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004467630504507727,
      "loss": 4.6943,
      "mean_token_accuracy": 0.2417423039674759,
      "num_tokens": 55158144.0,
      "step": 24045
    },
    {
      "entropy": 5.335883235931396,
      "epoch": 2.310278578290106,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004467409034648776,
      "loss": 4.7982,
      "mean_token_accuracy": 0.24636651277542115,
      "num_tokens": 55171522.0,
      "step": 24050
    },
    {
      "entropy": 5.33984603881836,
      "epoch": 2.3107588856868397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004467187524916688,
      "loss": 4.7132,
      "mean_token_accuracy": 0.24982267171144484,
      "num_tokens": 55182097.0,
      "step": 24055
    },
    {
      "entropy": 5.3632231712341305,
      "epoch": 2.3112391930835736,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004466965975316604,
      "loss": 4.7713,
      "mean_token_accuracy": 0.23980385959148406,
      "num_tokens": 55192287.0,
      "step": 24060
    },
    {
      "entropy": 5.36647481918335,
      "epoch": 2.3117195004803075,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044667443858536685,
      "loss": 4.7838,
      "mean_token_accuracy": 0.24320458769798278,
      "num_tokens": 55204049.0,
      "step": 24065
    },
    {
      "entropy": 5.274859857559204,
      "epoch": 2.3121998078770414,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004466522756533026,
      "loss": 4.5928,
      "mean_token_accuracy": 0.25288294404745104,
      "num_tokens": 55214969.0,
      "step": 24070
    },
    {
      "entropy": 5.40179762840271,
      "epoch": 2.3126801152737753,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004466301087359822,
      "loss": 4.8213,
      "mean_token_accuracy": 0.24400303810834884,
      "num_tokens": 55226362.0,
      "step": 24075
    },
    {
      "entropy": 5.415043735504151,
      "epoch": 2.313160422670509,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044660793783392035,
      "loss": 4.8015,
      "mean_token_accuracy": 0.2416181281208992,
      "num_tokens": 55237345.0,
      "step": 24080
    },
    {
      "entropy": 5.387293148040771,
      "epoch": 2.313640730067243,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004465857629476317,
      "loss": 4.823,
      "mean_token_accuracy": 0.24118077605962754,
      "num_tokens": 55248426.0,
      "step": 24085
    },
    {
      "entropy": 5.294674444198608,
      "epoch": 2.314121037463977,
      "grad_norm": 1.125,
      "learning_rate": 0.0004465635840776312,
      "loss": 4.768,
      "mean_token_accuracy": 0.24255839437246324,
      "num_tokens": 55259017.0,
      "step": 24090
    },
    {
      "entropy": 5.377728748321533,
      "epoch": 2.314601344860711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044654140122443373,
      "loss": 4.8065,
      "mean_token_accuracy": 0.24011656492948533,
      "num_tokens": 55270074.0,
      "step": 24095
    },
    {
      "entropy": 5.346732568740845,
      "epoch": 2.3150816522574447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004465192143885542,
      "loss": 4.7233,
      "mean_token_accuracy": 0.2497160628437996,
      "num_tokens": 55282291.0,
      "step": 24100
    },
    {
      "entropy": 5.390328407287598,
      "epoch": 2.3155619596541785,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044649702357050787,
      "loss": 4.8555,
      "mean_token_accuracy": 0.23886138200759888,
      "num_tokens": 55293113.0,
      "step": 24105
    },
    {
      "entropy": 5.345220708847046,
      "epoch": 2.3160422670509124,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464748287708099,
      "loss": 4.7765,
      "mean_token_accuracy": 0.2414157807826996,
      "num_tokens": 55304182.0,
      "step": 24110
    },
    {
      "entropy": 5.344146060943603,
      "epoch": 2.3165225744476463,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044645262998997557,
      "loss": 4.7114,
      "mean_token_accuracy": 0.24420208781957625,
      "num_tokens": 55315292.0,
      "step": 24115
    },
    {
      "entropy": 5.312312746047974,
      "epoch": 2.31700288184438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044643042722852024,
      "loss": 4.7067,
      "mean_token_accuracy": 0.2453678086400032,
      "num_tokens": 55327000.0,
      "step": 24120
    },
    {
      "entropy": 5.294695043563843,
      "epoch": 2.3174831892411145,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004464082204869595,
      "loss": 4.6629,
      "mean_token_accuracy": 0.25146090388298037,
      "num_tokens": 55337506.0,
      "step": 24125
    },
    {
      "entropy": 5.284445714950562,
      "epoch": 2.3179634966378484,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004463860097658088,
      "loss": 4.6352,
      "mean_token_accuracy": 0.24971071034669876,
      "num_tokens": 55348538.0,
      "step": 24130
    },
    {
      "entropy": 5.402809143066406,
      "epoch": 2.3184438040345823,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004463637950655839,
      "loss": 4.785,
      "mean_token_accuracy": 0.23871984332799911,
      "num_tokens": 55360501.0,
      "step": 24135
    },
    {
      "entropy": 5.343633651733398,
      "epoch": 2.318924111431316,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044634157638680054,
      "loss": 4.7346,
      "mean_token_accuracy": 0.24313539117574692,
      "num_tokens": 55371616.0,
      "step": 24140
    },
    {
      "entropy": 5.3927899360656735,
      "epoch": 2.31940441882805,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044631935372997455,
      "loss": 4.8235,
      "mean_token_accuracy": 0.2349054455757141,
      "num_tokens": 55382702.0,
      "step": 24145
    },
    {
      "entropy": 5.35815691947937,
      "epoch": 2.319884726224784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004462971270956219,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2417499378323555,
      "num_tokens": 55394826.0,
      "step": 24150
    },
    {
      "entropy": 5.281096363067627,
      "epoch": 2.320365033621518,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004462748964842586,
      "loss": 4.6574,
      "mean_token_accuracy": 0.25993620455265043,
      "num_tokens": 55406584.0,
      "step": 24155
    },
    {
      "entropy": 5.332727432250977,
      "epoch": 2.3208453410182517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004462526618964008,
      "loss": 4.7901,
      "mean_token_accuracy": 0.23890080899000168,
      "num_tokens": 55419640.0,
      "step": 24160
    },
    {
      "entropy": 5.32474045753479,
      "epoch": 2.3213256484149856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004462304233325647,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24651306569576265,
      "num_tokens": 55430197.0,
      "step": 24165
    },
    {
      "entropy": 5.3528547286987305,
      "epoch": 2.3218059558117194,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004462081807932666,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24525014460086822,
      "num_tokens": 55441832.0,
      "step": 24170
    },
    {
      "entropy": 5.307464981079102,
      "epoch": 2.3222862632084533,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004461859342790229,
      "loss": 4.6571,
      "mean_token_accuracy": 0.25207762122154237,
      "num_tokens": 55452656.0,
      "step": 24175
    },
    {
      "entropy": 5.23266019821167,
      "epoch": 2.322766570605187,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004461636837903501,
      "loss": 4.6117,
      "mean_token_accuracy": 0.2520767167210579,
      "num_tokens": 55463936.0,
      "step": 24180
    },
    {
      "entropy": 5.3961851596832275,
      "epoch": 2.323246878001921,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004461414293277649,
      "loss": 4.7951,
      "mean_token_accuracy": 0.2460714891552925,
      "num_tokens": 55475885.0,
      "step": 24185
    },
    {
      "entropy": 5.336593914031982,
      "epoch": 2.323727185398655,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461191708917839,
      "loss": 4.6976,
      "mean_token_accuracy": 0.23857541531324386,
      "num_tokens": 55489022.0,
      "step": 24190
    },
    {
      "entropy": 5.336173439025879,
      "epoch": 2.324207492795389,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044609690848292376,
      "loss": 4.7082,
      "mean_token_accuracy": 0.24698256552219391,
      "num_tokens": 55500008.0,
      "step": 24195
    },
    {
      "entropy": 5.356186962127685,
      "epoch": 2.324687800192123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004460746421017016,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22819729745388032,
      "num_tokens": 55511024.0,
      "step": 24200
    },
    {
      "entropy": 5.381625080108643,
      "epoch": 2.325168107588857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044605237174863405,
      "loss": 4.753,
      "mean_token_accuracy": 0.24544637650251389,
      "num_tokens": 55522199.0,
      "step": 24205
    },
    {
      "entropy": 5.379104948043823,
      "epoch": 2.325648414985591,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004460300974242386,
      "loss": 4.7067,
      "mean_token_accuracy": 0.24523767083883286,
      "num_tokens": 55533197.0,
      "step": 24210
    },
    {
      "entropy": 5.333476161956787,
      "epoch": 2.326128722382325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004460078191290319,
      "loss": 4.7343,
      "mean_token_accuracy": 0.24633412659168244,
      "num_tokens": 55544317.0,
      "step": 24215
    },
    {
      "entropy": 5.3505536079406735,
      "epoch": 2.3266090297790587,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044598553686353153,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2398588851094246,
      "num_tokens": 55555388.0,
      "step": 24220
    },
    {
      "entropy": 5.384151744842529,
      "epoch": 2.3270893371757926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044596325062825476,
      "loss": 4.7137,
      "mean_token_accuracy": 0.24527337849140168,
      "num_tokens": 55566189.0,
      "step": 24225
    },
    {
      "entropy": 5.308078384399414,
      "epoch": 2.3275696445725265,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004459409604237189,
      "loss": 4.6917,
      "mean_token_accuracy": 0.2552454248070717,
      "num_tokens": 55578235.0,
      "step": 24230
    },
    {
      "entropy": 5.245363378524781,
      "epoch": 2.3280499519692603,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044591866625044154,
      "loss": 4.682,
      "mean_token_accuracy": 0.24679569602012635,
      "num_tokens": 55590300.0,
      "step": 24235
    },
    {
      "entropy": 5.395663404464722,
      "epoch": 2.3285302593659942,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004458963681089403,
      "loss": 4.7494,
      "mean_token_accuracy": 0.24653081446886063,
      "num_tokens": 55601017.0,
      "step": 24240
    },
    {
      "entropy": 5.435169219970703,
      "epoch": 2.329010566762728,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004458740659997328,
      "loss": 4.7659,
      "mean_token_accuracy": 0.24221744686365126,
      "num_tokens": 55612411.0,
      "step": 24245
    },
    {
      "entropy": 5.39183406829834,
      "epoch": 2.329490874159462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004458517599233369,
      "loss": 4.7953,
      "mean_token_accuracy": 0.24309015572071074,
      "num_tokens": 55623872.0,
      "step": 24250
    },
    {
      "entropy": 5.422986793518066,
      "epoch": 2.329971181556196,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004458294498802706,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23434480279684067,
      "num_tokens": 55634887.0,
      "step": 24255
    },
    {
      "entropy": 5.355221843719482,
      "epoch": 2.3304514889529298,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004458071358710516,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24324539452791213,
      "num_tokens": 55646286.0,
      "step": 24260
    },
    {
      "entropy": 5.3324789047241214,
      "epoch": 2.3309317963496636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004457848178961981,
      "loss": 4.7487,
      "mean_token_accuracy": 0.2409030467271805,
      "num_tokens": 55657647.0,
      "step": 24265
    },
    {
      "entropy": 5.322883367538452,
      "epoch": 2.3314121037463975,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044576249595622833,
      "loss": 4.7282,
      "mean_token_accuracy": 0.24018580466508865,
      "num_tokens": 55669000.0,
      "step": 24270
    },
    {
      "entropy": 5.357516717910767,
      "epoch": 2.331892411143132,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004457401700516603,
      "loss": 4.7274,
      "mean_token_accuracy": 0.2410438060760498,
      "num_tokens": 55680678.0,
      "step": 24275
    },
    {
      "entropy": 5.379160690307617,
      "epoch": 2.3323727185398653,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044571784018301267,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24375710934400557,
      "num_tokens": 55692429.0,
      "step": 24280
    },
    {
      "entropy": 5.393106269836426,
      "epoch": 2.3328530259365996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044569550635080365,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23427775055170058,
      "num_tokens": 55704659.0,
      "step": 24285
    },
    {
      "entropy": 5.262855148315429,
      "epoch": 2.3333333333333335,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044567316855555184,
      "loss": 4.6207,
      "mean_token_accuracy": 0.25315329879522325,
      "num_tokens": 55715584.0,
      "step": 24290
    },
    {
      "entropy": 5.406733846664428,
      "epoch": 2.3338136407300674,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004456508267977759,
      "loss": 4.8335,
      "mean_token_accuracy": 0.23839301615953445,
      "num_tokens": 55726715.0,
      "step": 24295
    },
    {
      "entropy": 5.312929487228393,
      "epoch": 2.3342939481268012,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044562848107799444,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24489210695028304,
      "num_tokens": 55737069.0,
      "step": 24300
    },
    {
      "entropy": 5.333858442306519,
      "epoch": 2.334774255523535,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044560613139672627,
      "loss": 4.6772,
      "mean_token_accuracy": 0.2551694095134735,
      "num_tokens": 55747774.0,
      "step": 24305
    },
    {
      "entropy": 5.372061824798584,
      "epoch": 2.335254562920269,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044558377775449036,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24820799827575685,
      "num_tokens": 55758209.0,
      "step": 24310
    },
    {
      "entropy": 5.420988750457764,
      "epoch": 2.335734870317003,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044556142015180573,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23798633813858033,
      "num_tokens": 55768668.0,
      "step": 24315
    },
    {
      "entropy": 5.503579998016358,
      "epoch": 2.3362151777137368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044553905858919134,
      "loss": 4.8608,
      "mean_token_accuracy": 0.23698278963565828,
      "num_tokens": 55780254.0,
      "step": 24320
    },
    {
      "entropy": 5.391843700408936,
      "epoch": 2.3366954851104706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004455166930671664,
      "loss": 4.7506,
      "mean_token_accuracy": 0.24425568133592607,
      "num_tokens": 55790241.0,
      "step": 24325
    },
    {
      "entropy": 5.424515247344971,
      "epoch": 2.3371757925072045,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044549432358625014,
      "loss": 4.8361,
      "mean_token_accuracy": 0.23830574303865432,
      "num_tokens": 55802040.0,
      "step": 24330
    },
    {
      "entropy": 5.354904270172119,
      "epoch": 2.3376560999039384,
      "grad_norm": 1.125,
      "learning_rate": 0.000445471950146962,
      "loss": 4.7402,
      "mean_token_accuracy": 0.24191735088825225,
      "num_tokens": 55813926.0,
      "step": 24335
    },
    {
      "entropy": 5.3488281726837155,
      "epoch": 2.3381364073006723,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004454495727498214,
      "loss": 4.7344,
      "mean_token_accuracy": 0.2473847046494484,
      "num_tokens": 55825281.0,
      "step": 24340
    },
    {
      "entropy": 5.357888078689575,
      "epoch": 2.338616714697406,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004454271913953478,
      "loss": 4.6948,
      "mean_token_accuracy": 0.25054876804351806,
      "num_tokens": 55836295.0,
      "step": 24345
    },
    {
      "entropy": 5.294942808151245,
      "epoch": 2.3390970220941405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044540480608406093,
      "loss": 4.7101,
      "mean_token_accuracy": 0.24888941645622253,
      "num_tokens": 55847150.0,
      "step": 24350
    },
    {
      "entropy": 5.323403596878052,
      "epoch": 2.339577329490874,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004453824168164804,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23749172687530518,
      "num_tokens": 55858135.0,
      "step": 24355
    },
    {
      "entropy": 5.42527813911438,
      "epoch": 2.3400576368876083,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004453600235931261,
      "loss": 4.8345,
      "mean_token_accuracy": 0.24124438762664796,
      "num_tokens": 55869319.0,
      "step": 24360
    },
    {
      "entropy": 5.3880527973175045,
      "epoch": 2.340537944284342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000445337626414518,
      "loss": 4.8009,
      "mean_token_accuracy": 0.243199922144413,
      "num_tokens": 55881383.0,
      "step": 24365
    },
    {
      "entropy": 5.318470001220703,
      "epoch": 2.341018251681076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044531522528117593,
      "loss": 4.7227,
      "mean_token_accuracy": 0.2427143171429634,
      "num_tokens": 55893726.0,
      "step": 24370
    },
    {
      "entropy": 5.371083068847656,
      "epoch": 2.34149855907781,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044529282019362007,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2465218111872673,
      "num_tokens": 55905260.0,
      "step": 24375
    },
    {
      "entropy": 5.396613931655883,
      "epoch": 2.341978866474544,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044527041115237056,
      "loss": 4.7893,
      "mean_token_accuracy": 0.246485136449337,
      "num_tokens": 55915404.0,
      "step": 24380
    },
    {
      "entropy": 5.328866052627563,
      "epoch": 2.3424591738712777,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004452479981579477,
      "loss": 4.6967,
      "mean_token_accuracy": 0.2414553567767143,
      "num_tokens": 55925972.0,
      "step": 24385
    },
    {
      "entropy": 5.321415424346924,
      "epoch": 2.3429394812680115,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004452255812108719,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2450828358530998,
      "num_tokens": 55937545.0,
      "step": 24390
    },
    {
      "entropy": 5.299268531799316,
      "epoch": 2.3434197886647454,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004452031603116636,
      "loss": 4.6614,
      "mean_token_accuracy": 0.2565068259835243,
      "num_tokens": 55949934.0,
      "step": 24395
    },
    {
      "entropy": 5.372416305541992,
      "epoch": 2.3439000960614793,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044518073546084326,
      "loss": 4.7825,
      "mean_token_accuracy": 0.24939707219600676,
      "num_tokens": 55960414.0,
      "step": 24400
    },
    {
      "entropy": 5.299971675872802,
      "epoch": 2.344380403458213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451583066589316,
      "loss": 4.7017,
      "mean_token_accuracy": 0.24361571222543715,
      "num_tokens": 55970999.0,
      "step": 24405
    },
    {
      "entropy": 5.34810733795166,
      "epoch": 2.344860710854947,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044513587390644925,
      "loss": 4.73,
      "mean_token_accuracy": 0.2426212400197983,
      "num_tokens": 55982801.0,
      "step": 24410
    },
    {
      "entropy": 5.4094420909881595,
      "epoch": 2.345341018251681,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044511343720391724,
      "loss": 4.7692,
      "mean_token_accuracy": 0.2408030018210411,
      "num_tokens": 55995143.0,
      "step": 24415
    },
    {
      "entropy": 5.327280139923095,
      "epoch": 2.345821325648415,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004450909965518563,
      "loss": 4.7676,
      "mean_token_accuracy": 0.23978973031044007,
      "num_tokens": 56006368.0,
      "step": 24420
    },
    {
      "entropy": 5.412396621704102,
      "epoch": 2.3463016330451487,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044506855195078755,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2456873059272766,
      "num_tokens": 56019245.0,
      "step": 24425
    },
    {
      "entropy": 5.3832234859466555,
      "epoch": 2.3467819404418826,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044504610340123185,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23163774013519287,
      "num_tokens": 56030982.0,
      "step": 24430
    },
    {
      "entropy": 5.348171615600586,
      "epoch": 2.347262247838617,
      "grad_norm": 1.125,
      "learning_rate": 0.00044502365090371066,
      "loss": 4.7549,
      "mean_token_accuracy": 0.2408944323658943,
      "num_tokens": 56042871.0,
      "step": 24435
    },
    {
      "entropy": 5.412360143661499,
      "epoch": 2.347742555235351,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004450011944587452,
      "loss": 4.7823,
      "mean_token_accuracy": 0.23634620904922485,
      "num_tokens": 56054099.0,
      "step": 24440
    },
    {
      "entropy": 5.314192628860473,
      "epoch": 2.3482228626320847,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044497873406685673,
      "loss": 4.7193,
      "mean_token_accuracy": 0.24516088664531707,
      "num_tokens": 56064315.0,
      "step": 24445
    },
    {
      "entropy": 5.281051063537598,
      "epoch": 2.3487031700288186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004449562697285669,
      "loss": 4.7103,
      "mean_token_accuracy": 0.2389305993914604,
      "num_tokens": 56076446.0,
      "step": 24450
    },
    {
      "entropy": 5.457502603530884,
      "epoch": 2.3491834774255524,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044493380144439707,
      "loss": 4.83,
      "mean_token_accuracy": 0.24011363536119462,
      "num_tokens": 56089356.0,
      "step": 24455
    },
    {
      "entropy": 5.4204401016235355,
      "epoch": 2.3496637848222863,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000444911329214869,
      "loss": 4.8242,
      "mean_token_accuracy": 0.2370094582438469,
      "num_tokens": 56101245.0,
      "step": 24460
    },
    {
      "entropy": 5.244881916046142,
      "epoch": 2.35014409221902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044488885304050434,
      "loss": 4.6373,
      "mean_token_accuracy": 0.250094598531723,
      "num_tokens": 56112756.0,
      "step": 24465
    },
    {
      "entropy": 5.375077867507935,
      "epoch": 2.350624399615754,
      "grad_norm": 1.046875,
      "learning_rate": 0.000444866372921825,
      "loss": 4.7691,
      "mean_token_accuracy": 0.24543174356222153,
      "num_tokens": 56124316.0,
      "step": 24470
    },
    {
      "entropy": 5.310796880722046,
      "epoch": 2.351104707012488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044484388885935287,
      "loss": 4.6899,
      "mean_token_accuracy": 0.25468487590551375,
      "num_tokens": 56135972.0,
      "step": 24475
    },
    {
      "entropy": 5.3758502960205075,
      "epoch": 2.351585014409222,
      "grad_norm": 1.125,
      "learning_rate": 0.00044482140085361005,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2378128230571747,
      "num_tokens": 56147335.0,
      "step": 24480
    },
    {
      "entropy": 5.342349338531494,
      "epoch": 2.3520653218059557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044479890890511853,
      "loss": 4.6666,
      "mean_token_accuracy": 0.24921713173389434,
      "num_tokens": 56158294.0,
      "step": 24485
    },
    {
      "entropy": 5.30308723449707,
      "epoch": 2.3525456292026896,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044477641301440054,
      "loss": 4.6175,
      "mean_token_accuracy": 0.25588681399822233,
      "num_tokens": 56169468.0,
      "step": 24490
    },
    {
      "entropy": 5.344718599319458,
      "epoch": 2.3530259365994235,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004447539131819784,
      "loss": 4.6936,
      "mean_token_accuracy": 0.2468390569090843,
      "num_tokens": 56180112.0,
      "step": 24495
    },
    {
      "entropy": 5.397200012207032,
      "epoch": 2.3535062439961574,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044473140940837436,
      "loss": 4.729,
      "mean_token_accuracy": 0.25052291452884673,
      "num_tokens": 56191290.0,
      "step": 24500
    },
    {
      "entropy": 5.414031887054444,
      "epoch": 2.3539865513928913,
      "grad_norm": 1.125,
      "learning_rate": 0.00044470890169411107,
      "loss": 4.8285,
      "mean_token_accuracy": 0.23796638697385789,
      "num_tokens": 56202447.0,
      "step": 24505
    },
    {
      "entropy": 5.413350391387939,
      "epoch": 2.3544668587896256,
      "grad_norm": 1.078125,
      "learning_rate": 0.000444686390039711,
      "loss": 4.8027,
      "mean_token_accuracy": 0.232855686545372,
      "num_tokens": 56213802.0,
      "step": 24510
    },
    {
      "entropy": 5.331378316879272,
      "epoch": 2.3549471661863595,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004446638744456968,
      "loss": 4.7216,
      "mean_token_accuracy": 0.24255577772855758,
      "num_tokens": 56225243.0,
      "step": 24515
    },
    {
      "entropy": 5.364056348800659,
      "epoch": 2.3554274735830933,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044464135491259135,
      "loss": 4.6778,
      "mean_token_accuracy": 0.24607327878475188,
      "num_tokens": 56235116.0,
      "step": 24520
    },
    {
      "entropy": 5.332288694381714,
      "epoch": 2.3559077809798272,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004446188314409172,
      "loss": 4.7499,
      "mean_token_accuracy": 0.2425179213285446,
      "num_tokens": 56246654.0,
      "step": 24525
    },
    {
      "entropy": 5.363928365707397,
      "epoch": 2.356388088376561,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004445963040311975,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24636237025260926,
      "num_tokens": 56257787.0,
      "step": 24530
    },
    {
      "entropy": 5.239242887496948,
      "epoch": 2.356868395773295,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044457377268395526,
      "loss": 4.6437,
      "mean_token_accuracy": 0.24638913571834564,
      "num_tokens": 56268881.0,
      "step": 24535
    },
    {
      "entropy": 5.232321405410767,
      "epoch": 2.357348703170029,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044455123739971355,
      "loss": 4.6608,
      "mean_token_accuracy": 0.25262525826692583,
      "num_tokens": 56281515.0,
      "step": 24540
    },
    {
      "entropy": 5.295550060272217,
      "epoch": 2.3578290105667628,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044452869817899554,
      "loss": 4.6478,
      "mean_token_accuracy": 0.2487291380763054,
      "num_tokens": 56293226.0,
      "step": 24545
    },
    {
      "entropy": 5.359803533554077,
      "epoch": 2.3583093179634966,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004445061550223246,
      "loss": 4.7406,
      "mean_token_accuracy": 0.24164066314697266,
      "num_tokens": 56303408.0,
      "step": 24550
    },
    {
      "entropy": 5.33576602935791,
      "epoch": 2.3587896253602305,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044448360793022403,
      "loss": 4.7361,
      "mean_token_accuracy": 0.24081117063760757,
      "num_tokens": 56315379.0,
      "step": 24555
    },
    {
      "entropy": 5.353135108947754,
      "epoch": 2.3592699327569644,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004444610569032174,
      "loss": 4.7302,
      "mean_token_accuracy": 0.24449122995138167,
      "num_tokens": 56326496.0,
      "step": 24560
    },
    {
      "entropy": 5.277901029586792,
      "epoch": 2.3597502401536983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444385019418281,
      "loss": 4.6881,
      "mean_token_accuracy": 0.24500708431005477,
      "num_tokens": 56339416.0,
      "step": 24565
    },
    {
      "entropy": 5.364681816101074,
      "epoch": 2.360230547550432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044441594304658004,
      "loss": 4.7586,
      "mean_token_accuracy": 0.23891223669052125,
      "num_tokens": 56352305.0,
      "step": 24570
    },
    {
      "entropy": 5.32533369064331,
      "epoch": 2.360710854947166,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004443933802179968,
      "loss": 4.6706,
      "mean_token_accuracy": 0.2457134336233139,
      "num_tokens": 56364598.0,
      "step": 24575
    },
    {
      "entropy": 5.404559087753296,
      "epoch": 2.3611911623439,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044437081345660224,
      "loss": 4.8245,
      "mean_token_accuracy": 0.23070899695158004,
      "num_tokens": 56376651.0,
      "step": 24580
    },
    {
      "entropy": 5.290912008285522,
      "epoch": 2.3616714697406342,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004443482427629204,
      "loss": 4.7019,
      "mean_token_accuracy": 0.25246368199586866,
      "num_tokens": 56388038.0,
      "step": 24585
    },
    {
      "entropy": 5.364799928665161,
      "epoch": 2.3621517771373677,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004443256681374751,
      "loss": 4.7185,
      "mean_token_accuracy": 0.24532062113285064,
      "num_tokens": 56398889.0,
      "step": 24590
    },
    {
      "entropy": 5.329250955581665,
      "epoch": 2.362632084534102,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004443030895807907,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24101464301347733,
      "num_tokens": 56411548.0,
      "step": 24595
    },
    {
      "entropy": 5.345274877548218,
      "epoch": 2.363112391930836,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044428050709339117,
      "loss": 4.6483,
      "mean_token_accuracy": 0.24672218561172485,
      "num_tokens": 56423292.0,
      "step": 24600
    },
    {
      "entropy": 5.313033771514893,
      "epoch": 2.3635926993275698,
      "grad_norm": 1.046875,
      "learning_rate": 0.000444257920675801,
      "loss": 4.6733,
      "mean_token_accuracy": 0.25000896006822587,
      "num_tokens": 56434834.0,
      "step": 24605
    },
    {
      "entropy": 5.306661128997803,
      "epoch": 2.3640730067243036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044423533032854454,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23570105731487273,
      "num_tokens": 56445554.0,
      "step": 24610
    },
    {
      "entropy": 5.3086940288543705,
      "epoch": 2.3645533141210375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004442127360521462,
      "loss": 4.7168,
      "mean_token_accuracy": 0.2549698665738106,
      "num_tokens": 56457748.0,
      "step": 24615
    },
    {
      "entropy": 5.29944109916687,
      "epoch": 2.3650336215177714,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004441901378471306,
      "loss": 4.6489,
      "mean_token_accuracy": 0.25077402889728545,
      "num_tokens": 56469650.0,
      "step": 24620
    },
    {
      "entropy": 5.310337829589844,
      "epoch": 2.3655139289145053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044416753571402233,
      "loss": 4.6867,
      "mean_token_accuracy": 0.2465689957141876,
      "num_tokens": 56482061.0,
      "step": 24625
    },
    {
      "entropy": 5.365104484558105,
      "epoch": 2.365994236311239,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004441449296533462,
      "loss": 4.786,
      "mean_token_accuracy": 0.23988629281520843,
      "num_tokens": 56493599.0,
      "step": 24630
    },
    {
      "entropy": 5.391002130508423,
      "epoch": 2.366474543707973,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044412231966562717,
      "loss": 4.8245,
      "mean_token_accuracy": 0.2406530797481537,
      "num_tokens": 56505842.0,
      "step": 24635
    },
    {
      "entropy": 5.295916604995727,
      "epoch": 2.366954851104707,
      "grad_norm": 1.125,
      "learning_rate": 0.00044409970575139,
      "loss": 4.6918,
      "mean_token_accuracy": 0.2477475345134735,
      "num_tokens": 56517688.0,
      "step": 24640
    },
    {
      "entropy": 5.387258148193359,
      "epoch": 2.367435158501441,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004440770879111598,
      "loss": 4.7376,
      "mean_token_accuracy": 0.2486572653055191,
      "num_tokens": 56528665.0,
      "step": 24645
    },
    {
      "entropy": 5.242799615859985,
      "epoch": 2.3679154658981747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044405446614546163,
      "loss": 4.6392,
      "mean_token_accuracy": 0.2499245136976242,
      "num_tokens": 56539097.0,
      "step": 24650
    },
    {
      "entropy": 5.3870409488677975,
      "epoch": 2.3683957732949086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004440318404548208,
      "loss": 4.8095,
      "mean_token_accuracy": 0.24245842397212983,
      "num_tokens": 56550373.0,
      "step": 24655
    },
    {
      "entropy": 5.3550012588500975,
      "epoch": 2.3688760806916425,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044400921083976246,
      "loss": 4.7756,
      "mean_token_accuracy": 0.2380503684282303,
      "num_tokens": 56561289.0,
      "step": 24660
    },
    {
      "entropy": 5.362514925003052,
      "epoch": 2.3693563880883763,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004439865773008122,
      "loss": 4.78,
      "mean_token_accuracy": 0.23967952281236649,
      "num_tokens": 56572331.0,
      "step": 24665
    },
    {
      "entropy": 5.313803577423096,
      "epoch": 2.3698366954851107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004439639398384953,
      "loss": 4.6724,
      "mean_token_accuracy": 0.2530435785651207,
      "num_tokens": 56584375.0,
      "step": 24670
    },
    {
      "entropy": 5.281632900238037,
      "epoch": 2.3703170028818445,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044394129845333756,
      "loss": 4.6972,
      "mean_token_accuracy": 0.24674806147813796,
      "num_tokens": 56595825.0,
      "step": 24675
    },
    {
      "entropy": 5.318034744262695,
      "epoch": 2.3707973102785784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004439186531458645,
      "loss": 4.7696,
      "mean_token_accuracy": 0.24645740836858748,
      "num_tokens": 56607842.0,
      "step": 24680
    },
    {
      "entropy": 5.401943826675415,
      "epoch": 2.3712776176753123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044389600391660185,
      "loss": 4.7979,
      "mean_token_accuracy": 0.24014432579278946,
      "num_tokens": 56618509.0,
      "step": 24685
    },
    {
      "entropy": 5.3097412109375,
      "epoch": 2.371757925072046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044387335076607554,
      "loss": 4.6802,
      "mean_token_accuracy": 0.24712239056825638,
      "num_tokens": 56630293.0,
      "step": 24690
    },
    {
      "entropy": 5.396985960006714,
      "epoch": 2.37223823246878,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004438506936948115,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2376475602388382,
      "num_tokens": 56641852.0,
      "step": 24695
    },
    {
      "entropy": 5.45980429649353,
      "epoch": 2.372718539865514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044382803270333565,
      "loss": 4.7824,
      "mean_token_accuracy": 0.24443149268627168,
      "num_tokens": 56654409.0,
      "step": 24700
    },
    {
      "entropy": 5.386722993850708,
      "epoch": 2.373198847262248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004438053677921743,
      "loss": 4.7284,
      "mean_token_accuracy": 0.24413540661334993,
      "num_tokens": 56665227.0,
      "step": 24705
    },
    {
      "entropy": 5.350616884231568,
      "epoch": 2.3736791546589817,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044378269896185344,
      "loss": 4.7729,
      "mean_token_accuracy": 0.2394689604640007,
      "num_tokens": 56677924.0,
      "step": 24710
    },
    {
      "entropy": 5.329773378372193,
      "epoch": 2.3741594620557156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004437600262128996,
      "loss": 4.6758,
      "mean_token_accuracy": 0.24638461023569108,
      "num_tokens": 56689789.0,
      "step": 24715
    },
    {
      "entropy": 5.404334354400635,
      "epoch": 2.3746397694524495,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000443737349545839,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23753874003887177,
      "num_tokens": 56700595.0,
      "step": 24720
    },
    {
      "entropy": 5.401830530166626,
      "epoch": 2.3751200768491834,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044371466896119823,
      "loss": 4.8363,
      "mean_token_accuracy": 0.23955895453691484,
      "num_tokens": 56712554.0,
      "step": 24725
    },
    {
      "entropy": 5.382164716720581,
      "epoch": 2.3756003842459172,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044369198445950384,
      "loss": 4.7277,
      "mean_token_accuracy": 0.2443491190671921,
      "num_tokens": 56723698.0,
      "step": 24730
    },
    {
      "entropy": 5.357358932495117,
      "epoch": 2.376080691642651,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004436692960412824,
      "loss": 4.7406,
      "mean_token_accuracy": 0.251579712331295,
      "num_tokens": 56734660.0,
      "step": 24735
    },
    {
      "entropy": 5.339284992218017,
      "epoch": 2.376560999039385,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004436466037070608,
      "loss": 4.769,
      "mean_token_accuracy": 0.24159978181123734,
      "num_tokens": 56746150.0,
      "step": 24740
    },
    {
      "entropy": 5.462233209609986,
      "epoch": 2.3770413064361193,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044362390745736585,
      "loss": 4.8466,
      "mean_token_accuracy": 0.234512796998024,
      "num_tokens": 56758018.0,
      "step": 24745
    },
    {
      "entropy": 5.335121488571167,
      "epoch": 2.377521613832853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004436012072927245,
      "loss": 4.672,
      "mean_token_accuracy": 0.2522194340825081,
      "num_tokens": 56769129.0,
      "step": 24750
    },
    {
      "entropy": 5.342170000076294,
      "epoch": 2.378001921229587,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044357850321366375,
      "loss": 4.7494,
      "mean_token_accuracy": 0.24391200989484788,
      "num_tokens": 56780445.0,
      "step": 24755
    },
    {
      "entropy": 5.301720190048218,
      "epoch": 2.378482228626321,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004435557952207107,
      "loss": 4.7288,
      "mean_token_accuracy": 0.24857258945703506,
      "num_tokens": 56792338.0,
      "step": 24760
    },
    {
      "entropy": 5.37716498374939,
      "epoch": 2.378962536023055,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044353308331439257,
      "loss": 4.7613,
      "mean_token_accuracy": 0.24631955623626708,
      "num_tokens": 56802345.0,
      "step": 24765
    },
    {
      "entropy": 5.4142378807067875,
      "epoch": 2.3794428434197887,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004435103674952367,
      "loss": 4.773,
      "mean_token_accuracy": 0.24690648764371873,
      "num_tokens": 56813865.0,
      "step": 24770
    },
    {
      "entropy": 5.305678415298462,
      "epoch": 2.3799231508165226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044348764776377047,
      "loss": 4.6844,
      "mean_token_accuracy": 0.2423412725329399,
      "num_tokens": 56824135.0,
      "step": 24775
    },
    {
      "entropy": 5.28649640083313,
      "epoch": 2.3804034582132565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004434649241205214,
      "loss": 4.713,
      "mean_token_accuracy": 0.2459002137184143,
      "num_tokens": 56836104.0,
      "step": 24780
    },
    {
      "entropy": 5.341209793090821,
      "epoch": 2.3808837656099904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044344219656601704,
      "loss": 4.708,
      "mean_token_accuracy": 0.24826300293207168,
      "num_tokens": 56848308.0,
      "step": 24785
    },
    {
      "entropy": 5.358562660217285,
      "epoch": 2.3813640730067243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000443419465100785,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24429804235696792,
      "num_tokens": 56859351.0,
      "step": 24790
    },
    {
      "entropy": 5.376429080963135,
      "epoch": 2.381844380403458,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004433967297253531,
      "loss": 4.7695,
      "mean_token_accuracy": 0.24291929900646209,
      "num_tokens": 56870645.0,
      "step": 24795
    },
    {
      "entropy": 5.318565750122071,
      "epoch": 2.382324687800192,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044337399044024924,
      "loss": 4.7335,
      "mean_token_accuracy": 0.2422419235110283,
      "num_tokens": 56881246.0,
      "step": 24800
    },
    {
      "entropy": 5.32500376701355,
      "epoch": 2.382804995196926,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004433512472460012,
      "loss": 4.6976,
      "mean_token_accuracy": 0.24535200744867325,
      "num_tokens": 56893477.0,
      "step": 24805
    },
    {
      "entropy": 5.308488273620606,
      "epoch": 2.38328530259366,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044332850014313713,
      "loss": 4.7404,
      "mean_token_accuracy": 0.24391093701124192,
      "num_tokens": 56904813.0,
      "step": 24810
    },
    {
      "entropy": 5.314456033706665,
      "epoch": 2.3837656099903937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004433057491321851,
      "loss": 4.6846,
      "mean_token_accuracy": 0.24584470093250274,
      "num_tokens": 56916160.0,
      "step": 24815
    },
    {
      "entropy": 5.346708488464356,
      "epoch": 2.384245917387128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044328299421367333,
      "loss": 4.694,
      "mean_token_accuracy": 0.24857683777809142,
      "num_tokens": 56927791.0,
      "step": 24820
    },
    {
      "entropy": 5.38744740486145,
      "epoch": 2.3847262247838614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004432602353881302,
      "loss": 4.768,
      "mean_token_accuracy": 0.24255959391593934,
      "num_tokens": 56938848.0,
      "step": 24825
    },
    {
      "entropy": 5.341533041000366,
      "epoch": 2.3852065321805958,
      "grad_norm": 1.0,
      "learning_rate": 0.00044323747265608395,
      "loss": 4.7487,
      "mean_token_accuracy": 0.24757462590932847,
      "num_tokens": 56950116.0,
      "step": 24830
    },
    {
      "entropy": 5.33376841545105,
      "epoch": 2.3856868395773296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004432147060180632,
      "loss": 4.6974,
      "mean_token_accuracy": 0.25077409148216245,
      "num_tokens": 56961911.0,
      "step": 24835
    },
    {
      "entropy": 5.389460468292237,
      "epoch": 2.3861671469740635,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044319193547459645,
      "loss": 4.7666,
      "mean_token_accuracy": 0.23247936815023423,
      "num_tokens": 56974105.0,
      "step": 24840
    },
    {
      "entropy": 5.305670833587646,
      "epoch": 2.3866474543707974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004431691610262124,
      "loss": 4.7142,
      "mean_token_accuracy": 0.24265847057104112,
      "num_tokens": 56985564.0,
      "step": 24845
    },
    {
      "entropy": 5.32907190322876,
      "epoch": 2.3871277617675313,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044314638267343976,
      "loss": 4.6653,
      "mean_token_accuracy": 0.2560614675283432,
      "num_tokens": 56995835.0,
      "step": 24850
    },
    {
      "entropy": 5.231254911422729,
      "epoch": 2.387608069164265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004431236004168075,
      "loss": 4.6219,
      "mean_token_accuracy": 0.24942690432071685,
      "num_tokens": 57008114.0,
      "step": 24855
    },
    {
      "entropy": 5.282888793945313,
      "epoch": 2.388088376560999,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004431008142568444,
      "loss": 4.6903,
      "mean_token_accuracy": 0.24580911844968795,
      "num_tokens": 57018968.0,
      "step": 24860
    },
    {
      "entropy": 5.322457933425904,
      "epoch": 2.388568683957733,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044307802419407954,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24801263362169265,
      "num_tokens": 57031297.0,
      "step": 24865
    },
    {
      "entropy": 5.38571310043335,
      "epoch": 2.389048991354467,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004430552302290421,
      "loss": 4.7529,
      "mean_token_accuracy": 0.24063215255737305,
      "num_tokens": 57043457.0,
      "step": 24870
    },
    {
      "entropy": 5.308950042724609,
      "epoch": 2.3895292987512007,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004430324323622611,
      "loss": 4.6796,
      "mean_token_accuracy": 0.2440730482339859,
      "num_tokens": 57054474.0,
      "step": 24875
    },
    {
      "entropy": 5.316192531585694,
      "epoch": 2.3900096061479346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044300963059426605,
      "loss": 4.7136,
      "mean_token_accuracy": 0.2470712497830391,
      "num_tokens": 57065120.0,
      "step": 24880
    },
    {
      "entropy": 5.302127981185913,
      "epoch": 2.3904899135446684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044298682492558637,
      "loss": 4.7287,
      "mean_token_accuracy": 0.24428345263004303,
      "num_tokens": 57076712.0,
      "step": 24885
    },
    {
      "entropy": 5.4396059036254885,
      "epoch": 2.3909702209414023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044296401535675136,
      "loss": 4.8267,
      "mean_token_accuracy": 0.24023930579423905,
      "num_tokens": 57087894.0,
      "step": 24890
    },
    {
      "entropy": 5.416431903839111,
      "epoch": 2.3914505283381366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044294120188829056,
      "loss": 4.8936,
      "mean_token_accuracy": 0.2272513762116432,
      "num_tokens": 57101029.0,
      "step": 24895
    },
    {
      "entropy": 5.408068895339966,
      "epoch": 2.39193083573487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004429183845207339,
      "loss": 4.7547,
      "mean_token_accuracy": 0.23975684195756913,
      "num_tokens": 57112893.0,
      "step": 24900
    },
    {
      "entropy": 5.356100606918335,
      "epoch": 2.3924111431316044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004428955632546108,
      "loss": 4.7867,
      "mean_token_accuracy": 0.2478405252099037,
      "num_tokens": 57123899.0,
      "step": 24905
    },
    {
      "entropy": 5.338412141799926,
      "epoch": 2.3928914505283383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004428727380904514,
      "loss": 4.7926,
      "mean_token_accuracy": 0.24104450047016143,
      "num_tokens": 57134896.0,
      "step": 24910
    },
    {
      "entropy": 5.217646598815918,
      "epoch": 2.393371757925072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044284990902878545,
      "loss": 4.587,
      "mean_token_accuracy": 0.25557810813188553,
      "num_tokens": 57146278.0,
      "step": 24915
    },
    {
      "entropy": 5.344899654388428,
      "epoch": 2.393852065321806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044282707607014304,
      "loss": 4.7242,
      "mean_token_accuracy": 0.2467268392443657,
      "num_tokens": 57157663.0,
      "step": 24920
    },
    {
      "entropy": 5.352066898345948,
      "epoch": 2.39433237271854,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044280423921505427,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2532550483942032,
      "num_tokens": 57169131.0,
      "step": 24925
    },
    {
      "entropy": 5.361490201950073,
      "epoch": 2.394812680115274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004427813984640493,
      "loss": 4.7216,
      "mean_token_accuracy": 0.23870352357625962,
      "num_tokens": 57181653.0,
      "step": 24930
    },
    {
      "entropy": 5.2900713920593265,
      "epoch": 2.3952929875120077,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004427585538176585,
      "loss": 4.7149,
      "mean_token_accuracy": 0.24808821678161622,
      "num_tokens": 57192817.0,
      "step": 24935
    },
    {
      "entropy": 5.366632461547852,
      "epoch": 2.3957732949087416,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044273570527641223,
      "loss": 4.771,
      "mean_token_accuracy": 0.23833743929862977,
      "num_tokens": 57203570.0,
      "step": 24940
    },
    {
      "entropy": 5.310745048522949,
      "epoch": 2.3962536023054755,
      "grad_norm": 1.125,
      "learning_rate": 0.00044271285284084097,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2476769730448723,
      "num_tokens": 57214388.0,
      "step": 24945
    },
    {
      "entropy": 5.3664422035217285,
      "epoch": 2.3967339097022093,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004426899965114752,
      "loss": 4.7517,
      "mean_token_accuracy": 0.24169837087392806,
      "num_tokens": 57225004.0,
      "step": 24950
    },
    {
      "entropy": 5.266486692428589,
      "epoch": 2.3972142170989432,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044266713628884566,
      "loss": 4.6775,
      "mean_token_accuracy": 0.2511877536773682,
      "num_tokens": 57236643.0,
      "step": 24955
    },
    {
      "entropy": 5.3246317386627195,
      "epoch": 2.397694524495677,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044264427217348315,
      "loss": 4.7117,
      "mean_token_accuracy": 0.247183720767498,
      "num_tokens": 57247502.0,
      "step": 24960
    },
    {
      "entropy": 5.42149133682251,
      "epoch": 2.398174831892411,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004426214041659184,
      "loss": 4.8539,
      "mean_token_accuracy": 0.24000805318355561,
      "num_tokens": 57259539.0,
      "step": 24965
    },
    {
      "entropy": 5.4566504001617435,
      "epoch": 2.398655139289145,
      "grad_norm": 1.25,
      "learning_rate": 0.0004425985322666824,
      "loss": 4.8154,
      "mean_token_accuracy": 0.23904582411050795,
      "num_tokens": 57270946.0,
      "step": 24970
    },
    {
      "entropy": 5.334535026550293,
      "epoch": 2.3991354466858787,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004425756564763061,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23950349241495134,
      "num_tokens": 57282855.0,
      "step": 24975
    },
    {
      "entropy": 5.389076566696167,
      "epoch": 2.399615754082613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044255277679532075,
      "loss": 4.7966,
      "mean_token_accuracy": 0.2350280225276947,
      "num_tokens": 57292638.0,
      "step": 24980
    },
    {
      "entropy": 5.391885423660279,
      "epoch": 2.400096061479347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044252989322425735,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24212955087423324,
      "num_tokens": 57304088.0,
      "step": 24985
    },
    {
      "entropy": 5.386612176895142,
      "epoch": 2.400576368876081,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044250700576364734,
      "loss": 4.7158,
      "mean_token_accuracy": 0.2477840632200241,
      "num_tokens": 57314534.0,
      "step": 24990
    },
    {
      "entropy": 5.339500617980957,
      "epoch": 2.4010566762728147,
      "grad_norm": 1.125,
      "learning_rate": 0.000442484114414022,
      "loss": 4.8089,
      "mean_token_accuracy": 0.24128586947917938,
      "num_tokens": 57326787.0,
      "step": 24995
    },
    {
      "entropy": 5.304520034790039,
      "epoch": 2.4015369836695486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004424612191759129,
      "loss": 4.7164,
      "mean_token_accuracy": 0.23970824629068374,
      "num_tokens": 57338935.0,
      "step": 25000
    },
    {
      "entropy": 5.413946342468262,
      "epoch": 2.4020172910662825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004424383200498515,
      "loss": 4.7567,
      "mean_token_accuracy": 0.24426374733448028,
      "num_tokens": 57349963.0,
      "step": 25005
    },
    {
      "entropy": 5.368399000167846,
      "epoch": 2.4024975984630164,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004424154170363696,
      "loss": 4.7627,
      "mean_token_accuracy": 0.2436523601412773,
      "num_tokens": 57361305.0,
      "step": 25010
    },
    {
      "entropy": 5.31254506111145,
      "epoch": 2.4029779058597502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004423925101359987,
      "loss": 4.7217,
      "mean_token_accuracy": 0.2395223468542099,
      "num_tokens": 57371582.0,
      "step": 25015
    },
    {
      "entropy": 5.3458497524261475,
      "epoch": 2.403458213256484,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004423695993492709,
      "loss": 4.7107,
      "mean_token_accuracy": 0.24219976514577865,
      "num_tokens": 57382222.0,
      "step": 25020
    },
    {
      "entropy": 5.312402772903442,
      "epoch": 2.403938520653218,
      "grad_norm": 1.125,
      "learning_rate": 0.0004423466846767179,
      "loss": 4.7086,
      "mean_token_accuracy": 0.24521177411079406,
      "num_tokens": 57393939.0,
      "step": 25025
    },
    {
      "entropy": 5.361353921890259,
      "epoch": 2.404418828049952,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044232376611887185,
      "loss": 4.7169,
      "mean_token_accuracy": 0.24595606625080108,
      "num_tokens": 57404889.0,
      "step": 25030
    },
    {
      "entropy": 5.378064727783203,
      "epoch": 2.4048991354466858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044230084367626477,
      "loss": 4.7938,
      "mean_token_accuracy": 0.2415386274456978,
      "num_tokens": 57416296.0,
      "step": 25035
    },
    {
      "entropy": 5.294056224822998,
      "epoch": 2.4053794428434196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004422779173494288,
      "loss": 4.7571,
      "mean_token_accuracy": 0.24544522762298585,
      "num_tokens": 57427926.0,
      "step": 25040
    },
    {
      "entropy": 5.345160055160522,
      "epoch": 2.4058597502401535,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004422549871388965,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2493927001953125,
      "num_tokens": 57439397.0,
      "step": 25045
    },
    {
      "entropy": 5.382217025756836,
      "epoch": 2.4063400576368874,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044223205304519994,
      "loss": 4.792,
      "mean_token_accuracy": 0.24049362540245056,
      "num_tokens": 57449976.0,
      "step": 25050
    },
    {
      "entropy": 5.4080099105834964,
      "epoch": 2.4068203650336217,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004422091150688717,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23763509094715118,
      "num_tokens": 57461578.0,
      "step": 25055
    },
    {
      "entropy": 5.3359606742858885,
      "epoch": 2.4073006724303556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004421861732104443,
      "loss": 4.7664,
      "mean_token_accuracy": 0.24203770309686662,
      "num_tokens": 57471832.0,
      "step": 25060
    },
    {
      "entropy": 5.40374608039856,
      "epoch": 2.4077809798270895,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004421632274704504,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23926813900470734,
      "num_tokens": 57482854.0,
      "step": 25065
    },
    {
      "entropy": 5.370753002166748,
      "epoch": 2.4082612872238234,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004421402778494227,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23666036278009414,
      "num_tokens": 57494243.0,
      "step": 25070
    },
    {
      "entropy": 5.355899715423584,
      "epoch": 2.4087415946205573,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004421173243478941,
      "loss": 4.7026,
      "mean_token_accuracy": 0.24700437486171722,
      "num_tokens": 57506185.0,
      "step": 25075
    },
    {
      "entropy": 5.368545913696289,
      "epoch": 2.409221902017291,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044209436696639745,
      "loss": 4.734,
      "mean_token_accuracy": 0.25121677815914156,
      "num_tokens": 57517980.0,
      "step": 25080
    },
    {
      "entropy": 5.274747467041015,
      "epoch": 2.409702209414025,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044207140570546574,
      "loss": 4.6216,
      "mean_token_accuracy": 0.25365189611911776,
      "num_tokens": 57529331.0,
      "step": 25085
    },
    {
      "entropy": 5.3249180793762205,
      "epoch": 2.410182516810759,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044204844056563216,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24777317494153978,
      "num_tokens": 57540463.0,
      "step": 25090
    },
    {
      "entropy": 5.403650093078613,
      "epoch": 2.410662824207493,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004420254715474297,
      "loss": 4.8424,
      "mean_token_accuracy": 0.23886225372552872,
      "num_tokens": 57550896.0,
      "step": 25095
    },
    {
      "entropy": 5.3283627986907955,
      "epoch": 2.4111431316042267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044200249865139187,
      "loss": 4.7207,
      "mean_token_accuracy": 0.24802268892526627,
      "num_tokens": 57562041.0,
      "step": 25100
    },
    {
      "entropy": 5.436002159118653,
      "epoch": 2.4116234390009605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044197952187805185,
      "loss": 4.8504,
      "mean_token_accuracy": 0.2371052861213684,
      "num_tokens": 57573771.0,
      "step": 25105
    },
    {
      "entropy": 5.348689365386963,
      "epoch": 2.4121037463976944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044195654122794324,
      "loss": 4.7098,
      "mean_token_accuracy": 0.24272936284542085,
      "num_tokens": 57585651.0,
      "step": 25110
    },
    {
      "entropy": 5.325748491287231,
      "epoch": 2.4125840537944283,
      "grad_norm": 1.125,
      "learning_rate": 0.0004419335567015994,
      "loss": 4.7198,
      "mean_token_accuracy": 0.24849774390459062,
      "num_tokens": 57596328.0,
      "step": 25115
    },
    {
      "entropy": 5.287246751785278,
      "epoch": 2.413064361191162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004419105682995542,
      "loss": 4.661,
      "mean_token_accuracy": 0.2491775184869766,
      "num_tokens": 57608012.0,
      "step": 25120
    },
    {
      "entropy": 5.408689641952515,
      "epoch": 2.413544668587896,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004418875760223411,
      "loss": 4.733,
      "mean_token_accuracy": 0.2421518787741661,
      "num_tokens": 57619711.0,
      "step": 25125
    },
    {
      "entropy": 5.377035903930664,
      "epoch": 2.4140249759846304,
      "grad_norm": 1.125,
      "learning_rate": 0.00044186457987049405,
      "loss": 4.7196,
      "mean_token_accuracy": 0.24245551526546477,
      "num_tokens": 57630263.0,
      "step": 25130
    },
    {
      "entropy": 5.296732759475708,
      "epoch": 2.414505283381364,
      "grad_norm": 1.140625,
      "learning_rate": 0.000441841579844547,
      "loss": 4.7315,
      "mean_token_accuracy": 0.24403147995471955,
      "num_tokens": 57640973.0,
      "step": 25135
    },
    {
      "entropy": 5.333933353424072,
      "epoch": 2.414985590778098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004418185759450338,
      "loss": 4.6705,
      "mean_token_accuracy": 0.24400443732738494,
      "num_tokens": 57651698.0,
      "step": 25140
    },
    {
      "entropy": 5.32362732887268,
      "epoch": 2.415465898174832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004417955681724887,
      "loss": 4.7278,
      "mean_token_accuracy": 0.24217445999383927,
      "num_tokens": 57662547.0,
      "step": 25145
    },
    {
      "entropy": 5.396020984649658,
      "epoch": 2.415946205571566,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044177255652744576,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23701879978179932,
      "num_tokens": 57675665.0,
      "step": 25150
    },
    {
      "entropy": 5.314811944961548,
      "epoch": 2.4164265129683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044174954101043926,
      "loss": 4.6886,
      "mean_token_accuracy": 0.24816319197416306,
      "num_tokens": 57687765.0,
      "step": 25155
    },
    {
      "entropy": 5.39875168800354,
      "epoch": 2.4169068203650337,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044172652162200354,
      "loss": 4.783,
      "mean_token_accuracy": 0.23935863822698594,
      "num_tokens": 57699575.0,
      "step": 25160
    },
    {
      "entropy": 5.419455862045288,
      "epoch": 2.4173871277617676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004417034983626731,
      "loss": 4.7853,
      "mean_token_accuracy": 0.24454631209373473,
      "num_tokens": 57710014.0,
      "step": 25165
    },
    {
      "entropy": 5.263224792480469,
      "epoch": 2.4178674351585014,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004416804712329825,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2577709496021271,
      "num_tokens": 57720763.0,
      "step": 25170
    },
    {
      "entropy": 5.343162298202515,
      "epoch": 2.4183477425552353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044165744023346614,
      "loss": 4.7924,
      "mean_token_accuracy": 0.23905838280916214,
      "num_tokens": 57731581.0,
      "step": 25175
    },
    {
      "entropy": 5.409988260269165,
      "epoch": 2.418828049951969,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044163440536465904,
      "loss": 4.8222,
      "mean_token_accuracy": 0.23995472937822343,
      "num_tokens": 57742473.0,
      "step": 25180
    },
    {
      "entropy": 5.416976737976074,
      "epoch": 2.419308357348703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044161136662709577,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23956245332956314,
      "num_tokens": 57754317.0,
      "step": 25185
    },
    {
      "entropy": 5.351978731155396,
      "epoch": 2.419788664745437,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044158832402131133,
      "loss": 4.7377,
      "mean_token_accuracy": 0.24103081971406937,
      "num_tokens": 57765882.0,
      "step": 25190
    },
    {
      "entropy": 5.351353883743286,
      "epoch": 2.420268972142171,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044156527754784066,
      "loss": 4.7089,
      "mean_token_accuracy": 0.2450721800327301,
      "num_tokens": 57776977.0,
      "step": 25195
    },
    {
      "entropy": 5.392636299133301,
      "epoch": 2.4207492795389047,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044154222720721887,
      "loss": 4.7822,
      "mean_token_accuracy": 0.24787696748971938,
      "num_tokens": 57788652.0,
      "step": 25200
    },
    {
      "entropy": 5.422864389419556,
      "epoch": 2.421229586935639,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441519172999981,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23509814590215683,
      "num_tokens": 57799660.0,
      "step": 25205
    },
    {
      "entropy": 5.374416399002075,
      "epoch": 2.4217098943323725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004414961149266625,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2314496800303459,
      "num_tokens": 57811754.0,
      "step": 25210
    },
    {
      "entropy": 5.397156763076782,
      "epoch": 2.422190201729107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044147305298779856,
      "loss": 4.7582,
      "mean_token_accuracy": 0.2479524716734886,
      "num_tokens": 57823470.0,
      "step": 25215
    },
    {
      "entropy": 5.425167179107666,
      "epoch": 2.4226705091258407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004414499871839247,
      "loss": 4.8499,
      "mean_token_accuracy": 0.23419205397367476,
      "num_tokens": 57834917.0,
      "step": 25220
    },
    {
      "entropy": 5.3563618659973145,
      "epoch": 2.4231508165225746,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004414269175155763,
      "loss": 4.7556,
      "mean_token_accuracy": 0.23996043652296067,
      "num_tokens": 57844655.0,
      "step": 25225
    },
    {
      "entropy": 5.385544681549073,
      "epoch": 2.4236311239193085,
      "grad_norm": 1.125,
      "learning_rate": 0.0004414038439832891,
      "loss": 4.7861,
      "mean_token_accuracy": 0.2417859837412834,
      "num_tokens": 57856019.0,
      "step": 25230
    },
    {
      "entropy": 5.3629639625549315,
      "epoch": 2.4241114313160423,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004413807665875988,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2445261612534523,
      "num_tokens": 57867661.0,
      "step": 25235
    },
    {
      "entropy": 5.370154953002929,
      "epoch": 2.4245917387127762,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044135768532904104,
      "loss": 4.7388,
      "mean_token_accuracy": 0.24077003002166747,
      "num_tokens": 57879899.0,
      "step": 25240
    },
    {
      "entropy": 5.42895884513855,
      "epoch": 2.42507204610951,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000441334600208152,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23525258302688598,
      "num_tokens": 57891168.0,
      "step": 25245
    },
    {
      "entropy": 5.382128953933716,
      "epoch": 2.425552353506244,
      "grad_norm": 1.125,
      "learning_rate": 0.00044131151122546724,
      "loss": 4.6929,
      "mean_token_accuracy": 0.23923978060483933,
      "num_tokens": 57901997.0,
      "step": 25250
    },
    {
      "entropy": 5.348763799667358,
      "epoch": 2.426032660902978,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044128841838152313,
      "loss": 4.7446,
      "mean_token_accuracy": 0.24445625096559526,
      "num_tokens": 57913380.0,
      "step": 25255
    },
    {
      "entropy": 5.296038675308227,
      "epoch": 2.4265129682997117,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004412653216768558,
      "loss": 4.655,
      "mean_token_accuracy": 0.2502958685159683,
      "num_tokens": 57923675.0,
      "step": 25260
    },
    {
      "entropy": 5.406459760665894,
      "epoch": 2.4269932756964456,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004412422211120013,
      "loss": 4.7761,
      "mean_token_accuracy": 0.2430829495191574,
      "num_tokens": 57935885.0,
      "step": 25265
    },
    {
      "entropy": 5.361088991165161,
      "epoch": 2.4274735830931795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004412191166874961,
      "loss": 4.7354,
      "mean_token_accuracy": 0.2412078469991684,
      "num_tokens": 57946551.0,
      "step": 25270
    },
    {
      "entropy": 5.311912584304809,
      "epoch": 2.4279538904899134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004411960084038766,
      "loss": 4.6539,
      "mean_token_accuracy": 0.24892996102571488,
      "num_tokens": 57957778.0,
      "step": 25275
    },
    {
      "entropy": 5.337730073928833,
      "epoch": 2.4284341978866473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044117289626167917,
      "loss": 4.7772,
      "mean_token_accuracy": 0.24708016216754913,
      "num_tokens": 57969256.0,
      "step": 25280
    },
    {
      "entropy": 5.356664180755615,
      "epoch": 2.428914505283381,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004411497802614406,
      "loss": 4.7055,
      "mean_token_accuracy": 0.24950391352176665,
      "num_tokens": 57981368.0,
      "step": 25285
    },
    {
      "entropy": 5.353858995437622,
      "epoch": 2.4293948126801155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004411266604036975,
      "loss": 4.8102,
      "mean_token_accuracy": 0.24061287343502044,
      "num_tokens": 57992862.0,
      "step": 25290
    },
    {
      "entropy": 5.284159469604492,
      "epoch": 2.4298751200768494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044110353668898674,
      "loss": 4.6289,
      "mean_token_accuracy": 0.2552827373147011,
      "num_tokens": 58003492.0,
      "step": 25295
    },
    {
      "entropy": 5.290531969070434,
      "epoch": 2.4303554274735832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004410804091178449,
      "loss": 4.6526,
      "mean_token_accuracy": 0.25308151394128797,
      "num_tokens": 58014623.0,
      "step": 25300
    },
    {
      "entropy": 5.404169750213623,
      "epoch": 2.430835734870317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004410572776908092,
      "loss": 4.797,
      "mean_token_accuracy": 0.24135761857032775,
      "num_tokens": 58024886.0,
      "step": 25305
    },
    {
      "entropy": 5.387620544433593,
      "epoch": 2.431316042267051,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044103414240841664,
      "loss": 4.793,
      "mean_token_accuracy": 0.23932377696037294,
      "num_tokens": 58035998.0,
      "step": 25310
    },
    {
      "entropy": 5.3948626041412355,
      "epoch": 2.431796349663785,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004410110032712043,
      "loss": 4.7545,
      "mean_token_accuracy": 0.25149706453084947,
      "num_tokens": 58046577.0,
      "step": 25315
    },
    {
      "entropy": 5.3023604393005375,
      "epoch": 2.4322766570605188,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004409878602797094,
      "loss": 4.6957,
      "mean_token_accuracy": 0.24827914834022521,
      "num_tokens": 58057711.0,
      "step": 25320
    },
    {
      "entropy": 5.355804109573365,
      "epoch": 2.4327569644572526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044096471343446923,
      "loss": 4.6913,
      "mean_token_accuracy": 0.24453192800283433,
      "num_tokens": 58069657.0,
      "step": 25325
    },
    {
      "entropy": 5.370145988464356,
      "epoch": 2.4332372718539865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004409415627360213,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24430697411298752,
      "num_tokens": 58081081.0,
      "step": 25330
    },
    {
      "entropy": 5.388729572296143,
      "epoch": 2.4337175792507204,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044091840818490303,
      "loss": 4.7105,
      "mean_token_accuracy": 0.23696616291999817,
      "num_tokens": 58091561.0,
      "step": 25335
    },
    {
      "entropy": 5.2962933540344235,
      "epoch": 2.4341978866474543,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044089524978165197,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2467866435647011,
      "num_tokens": 58102640.0,
      "step": 25340
    },
    {
      "entropy": 5.36241774559021,
      "epoch": 2.434678194044188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044087208752680577,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23412802219390869,
      "num_tokens": 58113356.0,
      "step": 25345
    },
    {
      "entropy": 5.41907148361206,
      "epoch": 2.435158501440922,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004408489214209023,
      "loss": 4.8449,
      "mean_token_accuracy": 0.23764426559209822,
      "num_tokens": 58125437.0,
      "step": 25350
    },
    {
      "entropy": 5.466281270980835,
      "epoch": 2.435638808837656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004408257514644793,
      "loss": 4.8262,
      "mean_token_accuracy": 0.24030633121728898,
      "num_tokens": 58138491.0,
      "step": 25355
    },
    {
      "entropy": 5.349001216888428,
      "epoch": 2.43611911623439,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044080257765807476,
      "loss": 4.7308,
      "mean_token_accuracy": 0.23579584211111068,
      "num_tokens": 58150200.0,
      "step": 25360
    },
    {
      "entropy": 5.358913516998291,
      "epoch": 2.436599423631124,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004407794000022267,
      "loss": 4.8366,
      "mean_token_accuracy": 0.24192926287651062,
      "num_tokens": 58162467.0,
      "step": 25365
    },
    {
      "entropy": 5.441703414916992,
      "epoch": 2.437079731027858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004407562184974732,
      "loss": 4.8552,
      "mean_token_accuracy": 0.2301675260066986,
      "num_tokens": 58175052.0,
      "step": 25370
    },
    {
      "entropy": 5.391958951950073,
      "epoch": 2.437560038424592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407330331443526,
      "loss": 4.757,
      "mean_token_accuracy": 0.2445457771420479,
      "num_tokens": 58186999.0,
      "step": 25375
    },
    {
      "entropy": 5.324234867095948,
      "epoch": 2.438040345821326,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440709843943403,
      "loss": 4.6498,
      "mean_token_accuracy": 0.25075135827064515,
      "num_tokens": 58197811.0,
      "step": 25380
    },
    {
      "entropy": 5.4375982761383055,
      "epoch": 2.4385206532180597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000440686650895163,
      "loss": 4.8555,
      "mean_token_accuracy": 0.240145680308342,
      "num_tokens": 58209545.0,
      "step": 25385
    },
    {
      "entropy": 5.315494632720947,
      "epoch": 2.4390009606147935,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044066345400017084,
      "loss": 4.6636,
      "mean_token_accuracy": 0.2507217571139336,
      "num_tokens": 58220994.0,
      "step": 25390
    },
    {
      "entropy": 5.274564552307129,
      "epoch": 2.4394812680115274,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044064025325896524,
      "loss": 4.6182,
      "mean_token_accuracy": 0.2543371692299843,
      "num_tokens": 58232778.0,
      "step": 25395
    },
    {
      "entropy": 5.394096326828003,
      "epoch": 2.4399615754082613,
      "grad_norm": 1.125,
      "learning_rate": 0.00044061704867208484,
      "loss": 4.7877,
      "mean_token_accuracy": 0.24770109206438065,
      "num_tokens": 58243889.0,
      "step": 25400
    },
    {
      "entropy": 5.30770320892334,
      "epoch": 2.440441882804995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044059384024006825,
      "loss": 4.6908,
      "mean_token_accuracy": 0.24767297953367234,
      "num_tokens": 58256144.0,
      "step": 25405
    },
    {
      "entropy": 5.346848964691162,
      "epoch": 2.440922190201729,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004405706279634545,
      "loss": 4.7223,
      "mean_token_accuracy": 0.2454131469130516,
      "num_tokens": 58267509.0,
      "step": 25410
    },
    {
      "entropy": 5.317186594009399,
      "epoch": 2.441402497598463,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044054741184278243,
      "loss": 4.671,
      "mean_token_accuracy": 0.25163798332214354,
      "num_tokens": 58278880.0,
      "step": 25415
    },
    {
      "entropy": 5.381020164489746,
      "epoch": 2.441882804995197,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044052419187859095,
      "loss": 4.6976,
      "mean_token_accuracy": 0.24433363527059554,
      "num_tokens": 58290046.0,
      "step": 25420
    },
    {
      "entropy": 5.265408182144165,
      "epoch": 2.4423631123919307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004405009680714193,
      "loss": 4.5777,
      "mean_token_accuracy": 0.2605554908514023,
      "num_tokens": 58301406.0,
      "step": 25425
    },
    {
      "entropy": 5.323939943313599,
      "epoch": 2.4428434197886646,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004404777404218065,
      "loss": 4.6717,
      "mean_token_accuracy": 0.25663247853517535,
      "num_tokens": 58313064.0,
      "step": 25430
    },
    {
      "entropy": 5.3327168941497805,
      "epoch": 2.4433237271853985,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000440454508930292,
      "loss": 4.8079,
      "mean_token_accuracy": 0.23121351450681688,
      "num_tokens": 58324779.0,
      "step": 25435
    },
    {
      "entropy": 5.3939556121826175,
      "epoch": 2.443804034582133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004404312735974152,
      "loss": 4.7947,
      "mean_token_accuracy": 0.24127036780118943,
      "num_tokens": 58336064.0,
      "step": 25440
    },
    {
      "entropy": 5.341655540466308,
      "epoch": 2.4442843419788662,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044040803442371533,
      "loss": 4.6629,
      "mean_token_accuracy": 0.2575659826397896,
      "num_tokens": 58347497.0,
      "step": 25445
    },
    {
      "entropy": 5.271103954315185,
      "epoch": 2.4447646493756006,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004403847914097321,
      "loss": 4.6593,
      "mean_token_accuracy": 0.25597435235977173,
      "num_tokens": 58357716.0,
      "step": 25450
    },
    {
      "entropy": 5.360018062591553,
      "epoch": 2.4452449567723344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044036154455600517,
      "loss": 4.7861,
      "mean_token_accuracy": 0.23873458355665206,
      "num_tokens": 58368451.0,
      "step": 25455
    },
    {
      "entropy": 5.358899211883545,
      "epoch": 2.4457252641690683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004403382938630741,
      "loss": 4.7545,
      "mean_token_accuracy": 0.2389329716563225,
      "num_tokens": 58378915.0,
      "step": 25460
    },
    {
      "entropy": 5.3284101486206055,
      "epoch": 2.446205571565802,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044031503933147887,
      "loss": 4.6555,
      "mean_token_accuracy": 0.2580718472599983,
      "num_tokens": 58390251.0,
      "step": 25465
    },
    {
      "entropy": 5.340762710571289,
      "epoch": 2.446685878962536,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044029178096175934,
      "loss": 4.6857,
      "mean_token_accuracy": 0.25021591782569885,
      "num_tokens": 58401390.0,
      "step": 25470
    },
    {
      "entropy": 5.298350381851196,
      "epoch": 2.44716618635927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004402685187544554,
      "loss": 4.7206,
      "mean_token_accuracy": 0.25300102680921555,
      "num_tokens": 58412530.0,
      "step": 25475
    },
    {
      "entropy": 5.353336477279663,
      "epoch": 2.447646493756004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004402452527101072,
      "loss": 4.7873,
      "mean_token_accuracy": 0.2435840919613838,
      "num_tokens": 58424437.0,
      "step": 25480
    },
    {
      "entropy": 5.40989875793457,
      "epoch": 2.4481268011527377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004402219828292549,
      "loss": 4.7765,
      "mean_token_accuracy": 0.23700784891843796,
      "num_tokens": 58435661.0,
      "step": 25485
    },
    {
      "entropy": 5.432336616516113,
      "epoch": 2.4486071085494716,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004401987091124388,
      "loss": 4.857,
      "mean_token_accuracy": 0.231657275557518,
      "num_tokens": 58445840.0,
      "step": 25490
    },
    {
      "entropy": 5.390316200256348,
      "epoch": 2.4490874159462055,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401754315601992,
      "loss": 4.8115,
      "mean_token_accuracy": 0.231622576713562,
      "num_tokens": 58460160.0,
      "step": 25495
    },
    {
      "entropy": 5.320026445388794,
      "epoch": 2.4495677233429394,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004401521501730765,
      "loss": 4.6396,
      "mean_token_accuracy": 0.2555305853486061,
      "num_tokens": 58471257.0,
      "step": 25500
    },
    {
      "entropy": 5.254549980163574,
      "epoch": 2.4500480307396733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044012886495161144,
      "loss": 4.682,
      "mean_token_accuracy": 0.24694566130638124,
      "num_tokens": 58482830.0,
      "step": 25505
    },
    {
      "entropy": 5.3526997566223145,
      "epoch": 2.450528338136407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004401055758963443,
      "loss": 4.6973,
      "mean_token_accuracy": 0.2534628853201866,
      "num_tokens": 58494278.0,
      "step": 25510
    },
    {
      "entropy": 5.341869831085205,
      "epoch": 2.451008645533141,
      "grad_norm": 1.09375,
      "learning_rate": 0.000440082283007816,
      "loss": 4.7037,
      "mean_token_accuracy": 0.24100592583417893,
      "num_tokens": 58505653.0,
      "step": 25515
    },
    {
      "entropy": 5.385990762710572,
      "epoch": 2.451488952929875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044005898628656734,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2422904834151268,
      "num_tokens": 58516906.0,
      "step": 25520
    },
    {
      "entropy": 5.3828980922698975,
      "epoch": 2.4519692603266092,
      "grad_norm": 1.328125,
      "learning_rate": 0.000440035685733139,
      "loss": 4.7539,
      "mean_token_accuracy": 0.2487964704632759,
      "num_tokens": 58527326.0,
      "step": 25525
    },
    {
      "entropy": 5.340374183654785,
      "epoch": 2.452449567723343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004400123813480722,
      "loss": 4.6878,
      "mean_token_accuracy": 0.2500326827168465,
      "num_tokens": 58537749.0,
      "step": 25530
    },
    {
      "entropy": 5.336763954162597,
      "epoch": 2.452929875120077,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043998907313190787,
      "loss": 4.726,
      "mean_token_accuracy": 0.24130743891000747,
      "num_tokens": 58549479.0,
      "step": 25535
    },
    {
      "entropy": 5.345426940917969,
      "epoch": 2.453410182516811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004399657610851873,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24679375886917115,
      "num_tokens": 58561051.0,
      "step": 25540
    },
    {
      "entropy": 5.3854344367980955,
      "epoch": 2.4538904899135447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043994244520845146,
      "loss": 4.8266,
      "mean_token_accuracy": 0.23767464905977248,
      "num_tokens": 58571482.0,
      "step": 25545
    },
    {
      "entropy": 5.338894319534302,
      "epoch": 2.4543707973102786,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004399191255022418,
      "loss": 4.7018,
      "mean_token_accuracy": 0.2489435240626335,
      "num_tokens": 58584369.0,
      "step": 25550
    },
    {
      "entropy": 5.360405683517456,
      "epoch": 2.4548511047070125,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004398958019670998,
      "loss": 4.6929,
      "mean_token_accuracy": 0.24845885634422302,
      "num_tokens": 58595786.0,
      "step": 25555
    },
    {
      "entropy": 5.2942180156707765,
      "epoch": 2.4553314121037464,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043987247460356696,
      "loss": 4.7055,
      "mean_token_accuracy": 0.24555952847003937,
      "num_tokens": 58608536.0,
      "step": 25560
    },
    {
      "entropy": 5.253662204742431,
      "epoch": 2.4558117195004803,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004398491434121847,
      "loss": 4.6599,
      "mean_token_accuracy": 0.25263796746730804,
      "num_tokens": 58620981.0,
      "step": 25565
    },
    {
      "entropy": 5.415938997268677,
      "epoch": 2.456292026897214,
      "grad_norm": 1.203125,
      "learning_rate": 0.000439825808393495,
      "loss": 4.7326,
      "mean_token_accuracy": 0.2441316068172455,
      "num_tokens": 58632802.0,
      "step": 25570
    },
    {
      "entropy": 5.377977323532105,
      "epoch": 2.456772334293948,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004398024695480394,
      "loss": 4.7834,
      "mean_token_accuracy": 0.24691038429737092,
      "num_tokens": 58644577.0,
      "step": 25575
    },
    {
      "entropy": 5.372489023208618,
      "epoch": 2.457252641690682,
      "grad_norm": 1.125,
      "learning_rate": 0.0004397791268763598,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24405156821012497,
      "num_tokens": 58654970.0,
      "step": 25580
    },
    {
      "entropy": 5.375953102111817,
      "epoch": 2.457732949087416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043975578037899814,
      "loss": 4.8153,
      "mean_token_accuracy": 0.23903334587812425,
      "num_tokens": 58666534.0,
      "step": 25585
    },
    {
      "entropy": 5.356626701354981,
      "epoch": 2.4582132564841497,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004397324300564966,
      "loss": 4.7104,
      "mean_token_accuracy": 0.24787124693393708,
      "num_tokens": 58677451.0,
      "step": 25590
    },
    {
      "entropy": 5.378768110275269,
      "epoch": 2.4586935638808836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004397090759093971,
      "loss": 4.7057,
      "mean_token_accuracy": 0.2515427127480507,
      "num_tokens": 58689179.0,
      "step": 25595
    },
    {
      "entropy": 5.343332195281983,
      "epoch": 2.459173871277618,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043968571793824194,
      "loss": 4.8058,
      "mean_token_accuracy": 0.24442394971847534,
      "num_tokens": 58701939.0,
      "step": 25600
    },
    {
      "entropy": 5.279941558837891,
      "epoch": 2.4596541786743518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004396623561435734,
      "loss": 4.5817,
      "mean_token_accuracy": 0.25992731750011444,
      "num_tokens": 58714302.0,
      "step": 25605
    },
    {
      "entropy": 5.331330919265747,
      "epoch": 2.4601344860710856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004396389905259339,
      "loss": 4.726,
      "mean_token_accuracy": 0.2431240424513817,
      "num_tokens": 58727619.0,
      "step": 25610
    },
    {
      "entropy": 5.390741968154908,
      "epoch": 2.4606147934678195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043961562108586603,
      "loss": 4.791,
      "mean_token_accuracy": 0.23551993072032928,
      "num_tokens": 58738929.0,
      "step": 25615
    },
    {
      "entropy": 5.4434691905975345,
      "epoch": 2.4610951008645534,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043959224782391215,
      "loss": 4.8006,
      "mean_token_accuracy": 0.24478850811719893,
      "num_tokens": 58751771.0,
      "step": 25620
    },
    {
      "entropy": 5.260545825958252,
      "epoch": 2.4615754082612873,
      "grad_norm": 1.1875,
      "learning_rate": 0.000439568870740615,
      "loss": 4.6528,
      "mean_token_accuracy": 0.25793610513210297,
      "num_tokens": 58762304.0,
      "step": 25625
    },
    {
      "entropy": 5.278812837600708,
      "epoch": 2.462055715658021,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004395454898365174,
      "loss": 4.7088,
      "mean_token_accuracy": 0.24562008380889894,
      "num_tokens": 58775222.0,
      "step": 25630
    },
    {
      "entropy": 5.315068864822388,
      "epoch": 2.462536023054755,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043952210511216205,
      "loss": 4.6992,
      "mean_token_accuracy": 0.2496173307299614,
      "num_tokens": 58785904.0,
      "step": 25635
    },
    {
      "entropy": 5.2972088813781735,
      "epoch": 2.463016330451489,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043949871656809205,
      "loss": 4.7104,
      "mean_token_accuracy": 0.2555748924612999,
      "num_tokens": 58796186.0,
      "step": 25640
    },
    {
      "entropy": 5.393298149108887,
      "epoch": 2.463496637848223,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043947532420485024,
      "loss": 4.7734,
      "mean_token_accuracy": 0.24102589339017869,
      "num_tokens": 58807317.0,
      "step": 25645
    },
    {
      "entropy": 5.414260721206665,
      "epoch": 2.4639769452449567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004394519280229798,
      "loss": 4.8368,
      "mean_token_accuracy": 0.24018035382032393,
      "num_tokens": 58819108.0,
      "step": 25650
    },
    {
      "entropy": 5.344442892074585,
      "epoch": 2.4644572526416906,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043942852802302397,
      "loss": 4.7492,
      "mean_token_accuracy": 0.24220120906829834,
      "num_tokens": 58830962.0,
      "step": 25655
    },
    {
      "entropy": 5.399046182632446,
      "epoch": 2.4649375600384245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004394051242055259,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2308042898774147,
      "num_tokens": 58843236.0,
      "step": 25660
    },
    {
      "entropy": 5.3280566215515135,
      "epoch": 2.4654178674351583,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004393817165710291,
      "loss": 4.6886,
      "mean_token_accuracy": 0.24764636605978013,
      "num_tokens": 58853995.0,
      "step": 25665
    },
    {
      "entropy": 5.30343599319458,
      "epoch": 2.465898174831892,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043935830512007687,
      "loss": 4.6915,
      "mean_token_accuracy": 0.2554923191666603,
      "num_tokens": 58866563.0,
      "step": 25670
    },
    {
      "entropy": 5.409355497360229,
      "epoch": 2.4663784822286265,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043933488985321286,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23730689585208892,
      "num_tokens": 58878660.0,
      "step": 25675
    },
    {
      "entropy": 5.3108563899993895,
      "epoch": 2.46685878962536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004393114707709807,
      "loss": 4.6829,
      "mean_token_accuracy": 0.24851684868335724,
      "num_tokens": 58890332.0,
      "step": 25680
    },
    {
      "entropy": 5.358557415008545,
      "epoch": 2.4673390970220943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004392880478739241,
      "loss": 4.7776,
      "mean_token_accuracy": 0.23772649317979813,
      "num_tokens": 58902228.0,
      "step": 25685
    },
    {
      "entropy": 5.372214365005493,
      "epoch": 2.467819404418828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004392646211625869,
      "loss": 4.7643,
      "mean_token_accuracy": 0.2467061847448349,
      "num_tokens": 58914452.0,
      "step": 25690
    },
    {
      "entropy": 5.358060216903686,
      "epoch": 2.468299711815562,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004392411906375129,
      "loss": 4.7372,
      "mean_token_accuracy": 0.23917120844125747,
      "num_tokens": 58926079.0,
      "step": 25695
    },
    {
      "entropy": 5.342823553085327,
      "epoch": 2.468780019212296,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043921775629924615,
      "loss": 4.7039,
      "mean_token_accuracy": 0.25230961591005324,
      "num_tokens": 58937292.0,
      "step": 25700
    },
    {
      "entropy": 5.3268712043762205,
      "epoch": 2.46926032660903,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043919431814833077,
      "loss": 4.7687,
      "mean_token_accuracy": 0.24384741336107255,
      "num_tokens": 58949425.0,
      "step": 25705
    },
    {
      "entropy": 5.384842729568481,
      "epoch": 2.4697406340057637,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043917087618531084,
      "loss": 4.7578,
      "mean_token_accuracy": 0.24562288522720338,
      "num_tokens": 58961620.0,
      "step": 25710
    },
    {
      "entropy": 5.383074140548706,
      "epoch": 2.4702209414024976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004391474304107307,
      "loss": 4.6973,
      "mean_token_accuracy": 0.2513650611042976,
      "num_tokens": 58973043.0,
      "step": 25715
    },
    {
      "entropy": 5.384355068206787,
      "epoch": 2.4707012487992315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043912398082513463,
      "loss": 4.864,
      "mean_token_accuracy": 0.22997657358646392,
      "num_tokens": 58984516.0,
      "step": 25720
    },
    {
      "entropy": 5.341231632232666,
      "epoch": 2.4711815561959654,
      "grad_norm": 1.125,
      "learning_rate": 0.000439100527429067,
      "loss": 4.6753,
      "mean_token_accuracy": 0.2520748183131218,
      "num_tokens": 58996473.0,
      "step": 25725
    },
    {
      "entropy": 5.363528823852539,
      "epoch": 2.4716618635926992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043907707022307243,
      "loss": 4.6748,
      "mean_token_accuracy": 0.24375026822090148,
      "num_tokens": 59009001.0,
      "step": 25730
    },
    {
      "entropy": 5.416208791732788,
      "epoch": 2.472142170989433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043905360920769553,
      "loss": 4.8028,
      "mean_token_accuracy": 0.24367992132902144,
      "num_tokens": 59021238.0,
      "step": 25735
    },
    {
      "entropy": 5.424332189559936,
      "epoch": 2.472622478386167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000439030144383481,
      "loss": 4.7869,
      "mean_token_accuracy": 0.24787338078022003,
      "num_tokens": 59031311.0,
      "step": 25740
    },
    {
      "entropy": 5.330635452270508,
      "epoch": 2.473102785782901,
      "grad_norm": 1.390625,
      "learning_rate": 0.00043900667575097355,
      "loss": 4.6896,
      "mean_token_accuracy": 0.2469698891043663,
      "num_tokens": 59043151.0,
      "step": 25745
    },
    {
      "entropy": 5.434612131118774,
      "epoch": 2.473583093179635,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004389832033107181,
      "loss": 4.8012,
      "mean_token_accuracy": 0.24645957350730896,
      "num_tokens": 59054513.0,
      "step": 25750
    },
    {
      "entropy": 5.379882526397705,
      "epoch": 2.4740634005763686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043895972706325953,
      "loss": 4.6999,
      "mean_token_accuracy": 0.24769505560398103,
      "num_tokens": 59064909.0,
      "step": 25755
    },
    {
      "entropy": 5.33945631980896,
      "epoch": 2.474543707973103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438936247009143,
      "loss": 4.76,
      "mean_token_accuracy": 0.23875679522752763,
      "num_tokens": 59077089.0,
      "step": 25760
    },
    {
      "entropy": 5.361870527267456,
      "epoch": 2.475024015369837,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043891276314891365,
      "loss": 4.7128,
      "mean_token_accuracy": 0.24633420556783675,
      "num_tokens": 59089188.0,
      "step": 25765
    },
    {
      "entropy": 5.411313247680664,
      "epoch": 2.4755043227665707,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004388892754831166,
      "loss": 4.7723,
      "mean_token_accuracy": 0.24078909903764725,
      "num_tokens": 59101384.0,
      "step": 25770
    },
    {
      "entropy": 5.394585800170899,
      "epoch": 2.4759846301633046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004388657840122973,
      "loss": 4.7414,
      "mean_token_accuracy": 0.24101697653532028,
      "num_tokens": 59112327.0,
      "step": 25775
    },
    {
      "entropy": 5.395894956588745,
      "epoch": 2.4764649375600385,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000438842288737001,
      "loss": 4.7788,
      "mean_token_accuracy": 0.23978375643491745,
      "num_tokens": 59123577.0,
      "step": 25780
    },
    {
      "entropy": 5.302930688858032,
      "epoch": 2.4769452449567724,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043881878965777325,
      "loss": 4.6802,
      "mean_token_accuracy": 0.25240428000688553,
      "num_tokens": 59135359.0,
      "step": 25785
    },
    {
      "entropy": 5.320555591583252,
      "epoch": 2.4774255523535063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043879528677515973,
      "loss": 4.6132,
      "mean_token_accuracy": 0.2549236476421356,
      "num_tokens": 59146679.0,
      "step": 25790
    },
    {
      "entropy": 5.335108470916748,
      "epoch": 2.47790585975024,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043877178008970596,
      "loss": 4.7548,
      "mean_token_accuracy": 0.24233800768852234,
      "num_tokens": 59159893.0,
      "step": 25795
    },
    {
      "entropy": 5.283112621307373,
      "epoch": 2.478386167146974,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004387482696019578,
      "loss": 4.6866,
      "mean_token_accuracy": 0.2500202387571335,
      "num_tokens": 59172592.0,
      "step": 25800
    },
    {
      "entropy": 5.3322032451629635,
      "epoch": 2.478866474543708,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043872475531246105,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24618907868862153,
      "num_tokens": 59183650.0,
      "step": 25805
    },
    {
      "entropy": 5.389300346374512,
      "epoch": 2.479346781940442,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043870123722176166,
      "loss": 4.7683,
      "mean_token_accuracy": 0.24707189500331878,
      "num_tokens": 59195423.0,
      "step": 25810
    },
    {
      "entropy": 5.3387116432189945,
      "epoch": 2.4798270893371757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004386777153304056,
      "loss": 4.7656,
      "mean_token_accuracy": 0.23784084618091583,
      "num_tokens": 59208486.0,
      "step": 25815
    },
    {
      "entropy": 5.330440425872803,
      "epoch": 2.4803073967339095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043865418963893896,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24681769460439681,
      "num_tokens": 59220230.0,
      "step": 25820
    },
    {
      "entropy": 5.337818813323975,
      "epoch": 2.4807877041306434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004386306601479081,
      "loss": 4.7624,
      "mean_token_accuracy": 0.2396613910794258,
      "num_tokens": 59231524.0,
      "step": 25825
    },
    {
      "entropy": 5.358001804351806,
      "epoch": 2.4812680115273773,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004386071268578591,
      "loss": 4.829,
      "mean_token_accuracy": 0.23513719737529754,
      "num_tokens": 59243525.0,
      "step": 25830
    },
    {
      "entropy": 5.3942821502685545,
      "epoch": 2.4817483189241116,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043858358976933844,
      "loss": 4.6738,
      "mean_token_accuracy": 0.24635009467601776,
      "num_tokens": 59253774.0,
      "step": 25835
    },
    {
      "entropy": 5.350784778594971,
      "epoch": 2.4822286263208455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043856004888289264,
      "loss": 4.734,
      "mean_token_accuracy": 0.24783222675323485,
      "num_tokens": 59265640.0,
      "step": 25840
    },
    {
      "entropy": 5.32339129447937,
      "epoch": 2.4827089337175794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004385365041990681,
      "loss": 4.789,
      "mean_token_accuracy": 0.2436710327863693,
      "num_tokens": 59277513.0,
      "step": 25845
    },
    {
      "entropy": 5.4353272914886475,
      "epoch": 2.4831892411143133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004385129557184116,
      "loss": 4.753,
      "mean_token_accuracy": 0.24532286524772645,
      "num_tokens": 59289446.0,
      "step": 25850
    },
    {
      "entropy": 5.4332233428955075,
      "epoch": 2.483669548511047,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043848940344146976,
      "loss": 4.7756,
      "mean_token_accuracy": 0.24679714143276216,
      "num_tokens": 59300156.0,
      "step": 25855
    },
    {
      "entropy": 5.287687015533447,
      "epoch": 2.484149855907781,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004384658473687894,
      "loss": 4.7155,
      "mean_token_accuracy": 0.24585206508636476,
      "num_tokens": 59311500.0,
      "step": 25860
    },
    {
      "entropy": 5.423905086517334,
      "epoch": 2.484630163304515,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004384422875009176,
      "loss": 4.8348,
      "mean_token_accuracy": 0.24290449768304825,
      "num_tokens": 59323299.0,
      "step": 25865
    },
    {
      "entropy": 5.41529107093811,
      "epoch": 2.485110470701249,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384187238384011,
      "loss": 4.7779,
      "mean_token_accuracy": 0.24201775342226028,
      "num_tokens": 59335955.0,
      "step": 25870
    },
    {
      "entropy": 5.276340675354004,
      "epoch": 2.4855907780979827,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004383951563817871,
      "loss": 4.6502,
      "mean_token_accuracy": 0.24737419337034225,
      "num_tokens": 59347571.0,
      "step": 25875
    },
    {
      "entropy": 5.342245864868164,
      "epoch": 2.4860710854947166,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004383715851316227,
      "loss": 4.7327,
      "mean_token_accuracy": 0.24469150006771087,
      "num_tokens": 59358932.0,
      "step": 25880
    },
    {
      "entropy": 5.246404695510864,
      "epoch": 2.4865513928914504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043834801008845527,
      "loss": 4.6464,
      "mean_token_accuracy": 0.2549636423587799,
      "num_tokens": 59371425.0,
      "step": 25885
    },
    {
      "entropy": 5.284940910339356,
      "epoch": 2.4870317002881843,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004383244312528321,
      "loss": 4.6831,
      "mean_token_accuracy": 0.2410389244556427,
      "num_tokens": 59382917.0,
      "step": 25890
    },
    {
      "entropy": 5.43507342338562,
      "epoch": 2.487512007684918,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004383008486253006,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23513478487730027,
      "num_tokens": 59395014.0,
      "step": 25895
    },
    {
      "entropy": 5.3926863193511965,
      "epoch": 2.487992315081652,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043827726220640827,
      "loss": 4.7751,
      "mean_token_accuracy": 0.24561142176389694,
      "num_tokens": 59407115.0,
      "step": 25900
    },
    {
      "entropy": 5.474006128311157,
      "epoch": 2.488472622478386,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043825367199670274,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2376790910959244,
      "num_tokens": 59418744.0,
      "step": 25905
    },
    {
      "entropy": 5.329650259017944,
      "epoch": 2.4889529298751203,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004382300779967318,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2530334547162056,
      "num_tokens": 59429950.0,
      "step": 25910
    },
    {
      "entropy": 5.377033424377442,
      "epoch": 2.489433237271854,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043820648020704303,
      "loss": 4.857,
      "mean_token_accuracy": 0.2382788583636284,
      "num_tokens": 59441335.0,
      "step": 25915
    },
    {
      "entropy": 5.4157712936401365,
      "epoch": 2.489913544668588,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043818287862818444,
      "loss": 4.7372,
      "mean_token_accuracy": 0.24605117589235306,
      "num_tokens": 59452325.0,
      "step": 25920
    },
    {
      "entropy": 5.436654901504516,
      "epoch": 2.490393852065322,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000438159273260704,
      "loss": 4.8457,
      "mean_token_accuracy": 0.24331448078155518,
      "num_tokens": 59464352.0,
      "step": 25925
    },
    {
      "entropy": 5.374738073348999,
      "epoch": 2.490874159462056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004381356641051497,
      "loss": 4.7222,
      "mean_token_accuracy": 0.24777292758226394,
      "num_tokens": 59476048.0,
      "step": 25930
    },
    {
      "entropy": 5.422716665267944,
      "epoch": 2.4913544668587897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004381120511620697,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23766153007745744,
      "num_tokens": 59488139.0,
      "step": 25935
    },
    {
      "entropy": 5.37464919090271,
      "epoch": 2.4918347742555236,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043808843443201217,
      "loss": 4.826,
      "mean_token_accuracy": 0.24067238718271255,
      "num_tokens": 59498279.0,
      "step": 25940
    },
    {
      "entropy": 5.390185356140137,
      "epoch": 2.4923150816522575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004380648139155255,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24634429514408113,
      "num_tokens": 59509914.0,
      "step": 25945
    },
    {
      "entropy": 5.499292230606079,
      "epoch": 2.4927953890489913,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004380411896131581,
      "loss": 4.9098,
      "mean_token_accuracy": 0.23042665272951127,
      "num_tokens": 59521441.0,
      "step": 25950
    },
    {
      "entropy": 5.410424518585205,
      "epoch": 2.493275696445725,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043801756152545836,
      "loss": 4.7927,
      "mean_token_accuracy": 0.24300790429115296,
      "num_tokens": 59533086.0,
      "step": 25955
    },
    {
      "entropy": 5.36420578956604,
      "epoch": 2.493756003842459,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043799392965297496,
      "loss": 4.7873,
      "mean_token_accuracy": 0.24190953224897385,
      "num_tokens": 59545165.0,
      "step": 25960
    },
    {
      "entropy": 5.385427522659302,
      "epoch": 2.494236311239193,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004379702939962564,
      "loss": 4.7739,
      "mean_token_accuracy": 0.23124686628580093,
      "num_tokens": 59556016.0,
      "step": 25965
    },
    {
      "entropy": 5.386369228363037,
      "epoch": 2.494716618635927,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004379466545558516,
      "loss": 4.7562,
      "mean_token_accuracy": 0.24131855964660645,
      "num_tokens": 59566431.0,
      "step": 25970
    },
    {
      "entropy": 5.415160989761352,
      "epoch": 2.4951969260326607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043792301133230933,
      "loss": 4.812,
      "mean_token_accuracy": 0.23545338362455367,
      "num_tokens": 59579371.0,
      "step": 25975
    },
    {
      "entropy": 5.385939359664917,
      "epoch": 2.4956772334293946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004378993643261785,
      "loss": 4.7258,
      "mean_token_accuracy": 0.24809639900922775,
      "num_tokens": 59590088.0,
      "step": 25980
    },
    {
      "entropy": 5.333972692489624,
      "epoch": 2.496157540826129,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043787571353800814,
      "loss": 4.6753,
      "mean_token_accuracy": 0.2554552868008614,
      "num_tokens": 59600973.0,
      "step": 25985
    },
    {
      "entropy": 5.34606966972351,
      "epoch": 2.4966378482228624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004378520589683475,
      "loss": 4.7333,
      "mean_token_accuracy": 0.24253453463315963,
      "num_tokens": 59613194.0,
      "step": 25990
    },
    {
      "entropy": 5.3668708324432375,
      "epoch": 2.4971181556195967,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043782840061774544,
      "loss": 4.7477,
      "mean_token_accuracy": 0.24048659652471543,
      "num_tokens": 59623932.0,
      "step": 25995
    },
    {
      "entropy": 5.398875284194946,
      "epoch": 2.4975984630163306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043780473848675143,
      "loss": 4.8026,
      "mean_token_accuracy": 0.24092032462358476,
      "num_tokens": 59635095.0,
      "step": 26000
    },
    {
      "entropy": 5.378239631652832,
      "epoch": 2.4980787704130645,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004377810725759149,
      "loss": 4.7375,
      "mean_token_accuracy": 0.24099632650613784,
      "num_tokens": 59646930.0,
      "step": 26005
    },
    {
      "entropy": 5.3700789451599125,
      "epoch": 2.4985590778097984,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043775740288578516,
      "loss": 4.6857,
      "mean_token_accuracy": 0.24988860934972762,
      "num_tokens": 59658729.0,
      "step": 26010
    },
    {
      "entropy": 5.345280647277832,
      "epoch": 2.4990393852065322,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004377337294169118,
      "loss": 4.7739,
      "mean_token_accuracy": 0.240486042201519,
      "num_tokens": 59670619.0,
      "step": 26015
    },
    {
      "entropy": 5.426921749114991,
      "epoch": 2.499519692603266,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043771005216984457,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23766830414533616,
      "num_tokens": 59682307.0,
      "step": 26020
    },
    {
      "entropy": 5.382787990570068,
      "epoch": 2.5,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000437686371145133,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2440864250063896,
      "num_tokens": 59693515.0,
      "step": 26025
    },
    {
      "entropy": 5.376715087890625,
      "epoch": 2.500480307396734,
      "grad_norm": 1.0625,
      "learning_rate": 0.000437662686343327,
      "loss": 4.7855,
      "mean_token_accuracy": 0.23990519642829894,
      "num_tokens": 59705012.0,
      "step": 26030
    },
    {
      "entropy": 5.4607665061950685,
      "epoch": 2.5009606147934678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004376389977649764,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23858509212732315,
      "num_tokens": 59716776.0,
      "step": 26035
    },
    {
      "entropy": 5.380136251449585,
      "epoch": 2.5014409221902016,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004376153054106313,
      "loss": 4.7473,
      "mean_token_accuracy": 0.24530425369739534,
      "num_tokens": 59727526.0,
      "step": 26040
    },
    {
      "entropy": 5.387203645706177,
      "epoch": 2.5019212295869355,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004375916092808416,
      "loss": 4.7273,
      "mean_token_accuracy": 0.2495861291885376,
      "num_tokens": 59738861.0,
      "step": 26045
    },
    {
      "entropy": 5.4884865283966064,
      "epoch": 2.5024015369836694,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004375679093761575,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22972525507211686,
      "num_tokens": 59751019.0,
      "step": 26050
    },
    {
      "entropy": 5.389583921432495,
      "epoch": 2.5028818443804033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043754420569712925,
      "loss": 4.7427,
      "mean_token_accuracy": 0.2488473579287529,
      "num_tokens": 59762936.0,
      "step": 26055
    },
    {
      "entropy": 5.375458145141602,
      "epoch": 2.5033621517771376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043752049824430736,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24290012270212175,
      "num_tokens": 59772909.0,
      "step": 26060
    },
    {
      "entropy": 5.283071136474609,
      "epoch": 2.503842459173871,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043749678701824197,
      "loss": 4.6775,
      "mean_token_accuracy": 0.25094126909971237,
      "num_tokens": 59784003.0,
      "step": 26065
    },
    {
      "entropy": 5.369092702865601,
      "epoch": 2.5043227665706054,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004374730720194837,
      "loss": 4.7772,
      "mean_token_accuracy": 0.24609546065330506,
      "num_tokens": 59796200.0,
      "step": 26070
    },
    {
      "entropy": 5.367141580581665,
      "epoch": 2.5048030739673393,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004374493532485832,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24533288031816483,
      "num_tokens": 59807559.0,
      "step": 26075
    },
    {
      "entropy": 5.392909717559815,
      "epoch": 2.505283381364073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437425630706091,
      "loss": 4.7871,
      "mean_token_accuracy": 0.24244491159915924,
      "num_tokens": 59818213.0,
      "step": 26080
    },
    {
      "entropy": 5.369356441497803,
      "epoch": 2.505763688760807,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000437401904392558,
      "loss": 4.7643,
      "mean_token_accuracy": 0.24616695046424866,
      "num_tokens": 59829485.0,
      "step": 26085
    },
    {
      "entropy": 5.339328718185425,
      "epoch": 2.506243996157541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043737817430853504,
      "loss": 4.8071,
      "mean_token_accuracy": 0.2462535545229912,
      "num_tokens": 59842187.0,
      "step": 26090
    },
    {
      "entropy": 5.466590166091919,
      "epoch": 2.506724303554275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043735444045457303,
      "loss": 4.882,
      "mean_token_accuracy": 0.23017221838235855,
      "num_tokens": 59852783.0,
      "step": 26095
    },
    {
      "entropy": 5.44015212059021,
      "epoch": 2.5072046109510087,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043733070283122306,
      "loss": 4.7283,
      "mean_token_accuracy": 0.2419237643480301,
      "num_tokens": 59863661.0,
      "step": 26100
    },
    {
      "entropy": 5.3572012424469,
      "epoch": 2.5076849183477425,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043730696143903607,
      "loss": 4.7051,
      "mean_token_accuracy": 0.2492308109998703,
      "num_tokens": 59876109.0,
      "step": 26105
    },
    {
      "entropy": 5.292147159576416,
      "epoch": 2.5081652257444764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004372832162785635,
      "loss": 4.5764,
      "mean_token_accuracy": 0.25470654517412183,
      "num_tokens": 59886322.0,
      "step": 26110
    },
    {
      "entropy": 5.27913875579834,
      "epoch": 2.5086455331412103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004372594673503565,
      "loss": 4.6858,
      "mean_token_accuracy": 0.24693802297115325,
      "num_tokens": 59897307.0,
      "step": 26115
    },
    {
      "entropy": 5.307277393341065,
      "epoch": 2.509125840537944,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004372357146549664,
      "loss": 4.6418,
      "mean_token_accuracy": 0.2518728256225586,
      "num_tokens": 59908709.0,
      "step": 26120
    },
    {
      "entropy": 5.3451121807098385,
      "epoch": 2.509606147934678,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043721195819294487,
      "loss": 4.6888,
      "mean_token_accuracy": 0.2547922104597092,
      "num_tokens": 59920082.0,
      "step": 26125
    },
    {
      "entropy": 5.338413333892822,
      "epoch": 2.510086455331412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004371881979648433,
      "loss": 4.7097,
      "mean_token_accuracy": 0.24194797724485398,
      "num_tokens": 59931632.0,
      "step": 26130
    },
    {
      "entropy": 5.293112707138062,
      "epoch": 2.5105667627281463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004371644339712133,
      "loss": 4.6865,
      "mean_token_accuracy": 0.2462937444448471,
      "num_tokens": 59944093.0,
      "step": 26135
    },
    {
      "entropy": 5.366531181335449,
      "epoch": 2.5110470701248797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004371406662126067,
      "loss": 4.7296,
      "mean_token_accuracy": 0.23833635747432708,
      "num_tokens": 59954879.0,
      "step": 26140
    },
    {
      "entropy": 5.363802194595337,
      "epoch": 2.511527377521614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043711689468957534,
      "loss": 4.7428,
      "mean_token_accuracy": 0.24106235802173615,
      "num_tokens": 59967074.0,
      "step": 26145
    },
    {
      "entropy": 5.441539764404297,
      "epoch": 2.5120076849183475,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043709311940267107,
      "loss": 4.7741,
      "mean_token_accuracy": 0.24537006616592408,
      "num_tokens": 59978939.0,
      "step": 26150
    },
    {
      "entropy": 5.433216238021851,
      "epoch": 2.512487992315082,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004370693403524458,
      "loss": 4.7455,
      "mean_token_accuracy": 0.2438986673951149,
      "num_tokens": 59989251.0,
      "step": 26155
    },
    {
      "entropy": 5.272768592834472,
      "epoch": 2.5129682997118157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004370455575394518,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24561930298805237,
      "num_tokens": 60000267.0,
      "step": 26160
    },
    {
      "entropy": 5.388568162918091,
      "epoch": 2.5134486071085496,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004370217709642411,
      "loss": 4.7599,
      "mean_token_accuracy": 0.2380293995141983,
      "num_tokens": 60011190.0,
      "step": 26165
    },
    {
      "entropy": 5.385072326660156,
      "epoch": 2.5139289145052834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000436997980627366,
      "loss": 4.8214,
      "mean_token_accuracy": 0.2345840737223625,
      "num_tokens": 60023496.0,
      "step": 26170
    },
    {
      "entropy": 5.460495376586914,
      "epoch": 2.5144092219020173,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043697418652937877,
      "loss": 4.827,
      "mean_token_accuracy": 0.2368771329522133,
      "num_tokens": 60034585.0,
      "step": 26175
    },
    {
      "entropy": 5.425562191009521,
      "epoch": 2.514889529298751,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004369503886708319,
      "loss": 4.8033,
      "mean_token_accuracy": 0.2354872316122055,
      "num_tokens": 60046333.0,
      "step": 26180
    },
    {
      "entropy": 5.364349365234375,
      "epoch": 2.515369836695485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043692658705227796,
      "loss": 4.7927,
      "mean_token_accuracy": 0.23835314214229583,
      "num_tokens": 60058744.0,
      "step": 26185
    },
    {
      "entropy": 5.474870204925537,
      "epoch": 2.515850144092219,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043690278167426945,
      "loss": 4.8482,
      "mean_token_accuracy": 0.23759962022304534,
      "num_tokens": 60069303.0,
      "step": 26190
    },
    {
      "entropy": 5.3198693752288815,
      "epoch": 2.516330451488953,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004368789725373591,
      "loss": 4.6949,
      "mean_token_accuracy": 0.24986553341150283,
      "num_tokens": 60080604.0,
      "step": 26195
    },
    {
      "entropy": 5.3904759883880615,
      "epoch": 2.5168107588856867,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043685515964209977,
      "loss": 4.7439,
      "mean_token_accuracy": 0.2473731443285942,
      "num_tokens": 60090151.0,
      "step": 26200
    },
    {
      "entropy": 5.438986921310425,
      "epoch": 2.5172910662824206,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004368313429890441,
      "loss": 4.8354,
      "mean_token_accuracy": 0.23721711933612824,
      "num_tokens": 60102528.0,
      "step": 26205
    },
    {
      "entropy": 5.317896556854248,
      "epoch": 2.517771373679155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004368075225787453,
      "loss": 4.7023,
      "mean_token_accuracy": 0.24659770727157593,
      "num_tokens": 60113078.0,
      "step": 26210
    },
    {
      "entropy": 5.3043070316314695,
      "epoch": 2.5182516810758884,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004367836984117562,
      "loss": 4.6889,
      "mean_token_accuracy": 0.24382904618978501,
      "num_tokens": 60124728.0,
      "step": 26215
    },
    {
      "entropy": 5.2659400463104244,
      "epoch": 2.5187319884726227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043675987048863,
      "loss": 4.6716,
      "mean_token_accuracy": 0.24543798118829727,
      "num_tokens": 60136224.0,
      "step": 26220
    },
    {
      "entropy": 5.3912024974823,
      "epoch": 2.519212295869356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004367360388099201,
      "loss": 4.7817,
      "mean_token_accuracy": 0.2417552262544632,
      "num_tokens": 60148654.0,
      "step": 26225
    },
    {
      "entropy": 5.477967929840088,
      "epoch": 2.5196926032660905,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004367122033761796,
      "loss": 4.7825,
      "mean_token_accuracy": 0.24210938066244125,
      "num_tokens": 60159740.0,
      "step": 26230
    },
    {
      "entropy": 5.395452260971069,
      "epoch": 2.5201729106628243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004366883641879618,
      "loss": 4.793,
      "mean_token_accuracy": 0.24153801500797273,
      "num_tokens": 60171387.0,
      "step": 26235
    },
    {
      "entropy": 5.323436117172241,
      "epoch": 2.520653218059558,
      "grad_norm": 1.125,
      "learning_rate": 0.00043666452124582034,
      "loss": 4.7105,
      "mean_token_accuracy": 0.25131860077381135,
      "num_tokens": 60181562.0,
      "step": 26240
    },
    {
      "entropy": 5.288196182250976,
      "epoch": 2.521133525456292,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004366406745503088,
      "loss": 4.6541,
      "mean_token_accuracy": 0.24673811644315718,
      "num_tokens": 60191187.0,
      "step": 26245
    },
    {
      "entropy": 5.255898189544678,
      "epoch": 2.521613832853026,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004366168241019807,
      "loss": 4.6916,
      "mean_token_accuracy": 0.2548580065369606,
      "num_tokens": 60203792.0,
      "step": 26250
    },
    {
      "entropy": 5.449641942977905,
      "epoch": 2.52209414024976,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004365929699013899,
      "loss": 4.7553,
      "mean_token_accuracy": 0.2485227480530739,
      "num_tokens": 60215523.0,
      "step": 26255
    },
    {
      "entropy": 5.423949575424194,
      "epoch": 2.5225744476464937,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004365691119490902,
      "loss": 4.8016,
      "mean_token_accuracy": 0.24368085861206054,
      "num_tokens": 60226113.0,
      "step": 26260
    },
    {
      "entropy": 5.443356895446778,
      "epoch": 2.5230547550432276,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004365452502456354,
      "loss": 4.8651,
      "mean_token_accuracy": 0.2368111565709114,
      "num_tokens": 60238521.0,
      "step": 26265
    },
    {
      "entropy": 5.285477256774902,
      "epoch": 2.5235350624399615,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004365213847915796,
      "loss": 4.6396,
      "mean_token_accuracy": 0.25231152176856997,
      "num_tokens": 60249717.0,
      "step": 26270
    },
    {
      "entropy": 5.412798690795898,
      "epoch": 2.5240153698366954,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043649751558747695,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23222438842058182,
      "num_tokens": 60261749.0,
      "step": 26275
    },
    {
      "entropy": 5.3914477825164795,
      "epoch": 2.5244956772334293,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043647364263388143,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23938889503479005,
      "num_tokens": 60272925.0,
      "step": 26280
    },
    {
      "entropy": 5.423447465896606,
      "epoch": 2.524975984630163,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004364497659313475,
      "loss": 4.7636,
      "mean_token_accuracy": 0.2394995078444481,
      "num_tokens": 60285081.0,
      "step": 26285
    },
    {
      "entropy": 5.402713298797607,
      "epoch": 2.525456292026897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004364258854804294,
      "loss": 4.7548,
      "mean_token_accuracy": 0.24673074334859849,
      "num_tokens": 60296796.0,
      "step": 26290
    },
    {
      "entropy": 5.389077138900757,
      "epoch": 2.5259365994236314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004364020012816815,
      "loss": 4.8441,
      "mean_token_accuracy": 0.24033734947443008,
      "num_tokens": 60308600.0,
      "step": 26295
    },
    {
      "entropy": 5.353756093978882,
      "epoch": 2.526416906820365,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004363781133356584,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24886123985052108,
      "num_tokens": 60319188.0,
      "step": 26300
    },
    {
      "entropy": 5.424701118469239,
      "epoch": 2.526897214217099,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004363542216429147,
      "loss": 4.8061,
      "mean_token_accuracy": 0.24233652651309967,
      "num_tokens": 60329912.0,
      "step": 26305
    },
    {
      "entropy": 5.417070150375366,
      "epoch": 2.527377521613833,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004363303262040051,
      "loss": 4.775,
      "mean_token_accuracy": 0.24440769851207733,
      "num_tokens": 60339760.0,
      "step": 26310
    },
    {
      "entropy": 5.273773717880249,
      "epoch": 2.527857829010567,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043630642701948446,
      "loss": 4.6254,
      "mean_token_accuracy": 0.25255185216665266,
      "num_tokens": 60349947.0,
      "step": 26315
    },
    {
      "entropy": 5.3426495552062985,
      "epoch": 2.5283381364073008,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043628252408990756,
      "loss": 4.8242,
      "mean_token_accuracy": 0.23513203412294387,
      "num_tokens": 60362620.0,
      "step": 26320
    },
    {
      "entropy": 5.420818853378296,
      "epoch": 2.5288184438040346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043625861741582926,
      "loss": 4.7866,
      "mean_token_accuracy": 0.24109088629484177,
      "num_tokens": 60373729.0,
      "step": 26325
    },
    {
      "entropy": 5.4253926277160645,
      "epoch": 2.5292987512007685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043623470699780483,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24121790677309035,
      "num_tokens": 60386423.0,
      "step": 26330
    },
    {
      "entropy": 5.358440589904785,
      "epoch": 2.5297790585975024,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004362107928363892,
      "loss": 4.7734,
      "mean_token_accuracy": 0.24243559390306474,
      "num_tokens": 60398364.0,
      "step": 26335
    },
    {
      "entropy": 5.360712480545044,
      "epoch": 2.5302593659942363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004361868749321377,
      "loss": 4.793,
      "mean_token_accuracy": 0.2414584994316101,
      "num_tokens": 60408423.0,
      "step": 26340
    },
    {
      "entropy": 5.428333806991577,
      "epoch": 2.53073967339097,
      "grad_norm": 1.0,
      "learning_rate": 0.0004361629532856055,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23811232298612595,
      "num_tokens": 60420229.0,
      "step": 26345
    },
    {
      "entropy": 5.469498538970948,
      "epoch": 2.531219980787704,
      "grad_norm": 1.125,
      "learning_rate": 0.00043613902789734816,
      "loss": 4.8474,
      "mean_token_accuracy": 0.23851505368947984,
      "num_tokens": 60430711.0,
      "step": 26350
    },
    {
      "entropy": 5.435506963729859,
      "epoch": 2.531700288184438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000436115098767921,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23687957972288132,
      "num_tokens": 60442293.0,
      "step": 26355
    },
    {
      "entropy": 5.335764694213867,
      "epoch": 2.532180595581172,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043609116589787974,
      "loss": 4.7198,
      "mean_token_accuracy": 0.24822563230991362,
      "num_tokens": 60452616.0,
      "step": 26360
    },
    {
      "entropy": 5.346296548843384,
      "epoch": 2.5326609029779057,
      "grad_norm": 1.125,
      "learning_rate": 0.0004360672292877799,
      "loss": 4.726,
      "mean_token_accuracy": 0.2416718915104866,
      "num_tokens": 60463254.0,
      "step": 26365
    },
    {
      "entropy": 5.3517180442810055,
      "epoch": 2.53314121037464,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043604328893817726,
      "loss": 4.7417,
      "mean_token_accuracy": 0.24593522399663925,
      "num_tokens": 60473725.0,
      "step": 26370
    },
    {
      "entropy": 5.380674743652344,
      "epoch": 2.5336215177713735,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043601934484962775,
      "loss": 4.7757,
      "mean_token_accuracy": 0.24660785794258117,
      "num_tokens": 60484502.0,
      "step": 26375
    },
    {
      "entropy": 5.323826694488526,
      "epoch": 2.534101825168108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004359953970226871,
      "loss": 4.7076,
      "mean_token_accuracy": 0.2534389182925224,
      "num_tokens": 60496134.0,
      "step": 26380
    },
    {
      "entropy": 5.411583185195923,
      "epoch": 2.5345821325648417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043597144545791134,
      "loss": 4.7831,
      "mean_token_accuracy": 0.245423886179924,
      "num_tokens": 60507236.0,
      "step": 26385
    },
    {
      "entropy": 5.338635492324829,
      "epoch": 2.5350624399615755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004359474901558567,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2500953316688538,
      "num_tokens": 60517499.0,
      "step": 26390
    },
    {
      "entropy": 5.384001684188843,
      "epoch": 2.5355427473583094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004359235311170792,
      "loss": 4.8244,
      "mean_token_accuracy": 0.23402478992938996,
      "num_tokens": 60529434.0,
      "step": 26395
    },
    {
      "entropy": 5.413464117050171,
      "epoch": 2.5360230547550433,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004358995683421352,
      "loss": 4.7977,
      "mean_token_accuracy": 0.24172275513410568,
      "num_tokens": 60542571.0,
      "step": 26400
    },
    {
      "entropy": 5.386654710769653,
      "epoch": 2.536503362151777,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043587560183158095,
      "loss": 4.7243,
      "mean_token_accuracy": 0.24350059628486634,
      "num_tokens": 60553826.0,
      "step": 26405
    },
    {
      "entropy": 5.35264220237732,
      "epoch": 2.536983669548511,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000435851631585973,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2400376409292221,
      "num_tokens": 60565173.0,
      "step": 26410
    },
    {
      "entropy": 5.324710273742676,
      "epoch": 2.537463976945245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004358276576058677,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2491302564740181,
      "num_tokens": 60577267.0,
      "step": 26415
    },
    {
      "entropy": 5.39168963432312,
      "epoch": 2.537944284341979,
      "grad_norm": 1.25,
      "learning_rate": 0.0004358036798918218,
      "loss": 4.766,
      "mean_token_accuracy": 0.23732548356056213,
      "num_tokens": 60587441.0,
      "step": 26420
    },
    {
      "entropy": 5.462089252471924,
      "epoch": 2.5384245917387127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004357796984443919,
      "loss": 4.812,
      "mean_token_accuracy": 0.2408580705523491,
      "num_tokens": 60597173.0,
      "step": 26425
    },
    {
      "entropy": 5.330295515060425,
      "epoch": 2.5389048991354466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043575571326413484,
      "loss": 4.6767,
      "mean_token_accuracy": 0.24542035013437272,
      "num_tokens": 60608299.0,
      "step": 26430
    },
    {
      "entropy": 5.435695695877075,
      "epoch": 2.5393852065321805,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004357317243516075,
      "loss": 4.8641,
      "mean_token_accuracy": 0.23877304494380952,
      "num_tokens": 60619106.0,
      "step": 26435
    },
    {
      "entropy": 5.346229696273804,
      "epoch": 2.5398655139289144,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043570773170736676,
      "loss": 4.6785,
      "mean_token_accuracy": 0.24788910001516343,
      "num_tokens": 60629339.0,
      "step": 26440
    },
    {
      "entropy": 5.405059051513672,
      "epoch": 2.5403458213256487,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043568373533196976,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2509258672595024,
      "num_tokens": 60640162.0,
      "step": 26445
    },
    {
      "entropy": 5.3589214324951175,
      "epoch": 2.540826128722382,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043565973522597344,
      "loss": 4.7195,
      "mean_token_accuracy": 0.24279344379901885,
      "num_tokens": 60652529.0,
      "step": 26450
    },
    {
      "entropy": 5.283762264251709,
      "epoch": 2.5413064361191164,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043563573138993524,
      "loss": 4.6809,
      "mean_token_accuracy": 0.24375426024198532,
      "num_tokens": 60663035.0,
      "step": 26455
    },
    {
      "entropy": 5.333381223678589,
      "epoch": 2.54178674351585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004356117238244123,
      "loss": 4.694,
      "mean_token_accuracy": 0.25034448504447937,
      "num_tokens": 60674430.0,
      "step": 26460
    },
    {
      "entropy": 5.358366966247559,
      "epoch": 2.542267050912584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043558771252996204,
      "loss": 4.6834,
      "mean_token_accuracy": 0.2531615376472473,
      "num_tokens": 60685923.0,
      "step": 26465
    },
    {
      "entropy": 5.3380420207977295,
      "epoch": 2.542747358309318,
      "grad_norm": 1.09375,
      "learning_rate": 0.000435563697507142,
      "loss": 4.7416,
      "mean_token_accuracy": 0.2483152210712433,
      "num_tokens": 60696344.0,
      "step": 26470
    },
    {
      "entropy": 5.3618669509887695,
      "epoch": 2.543227665706052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004355396787565096,
      "loss": 4.7594,
      "mean_token_accuracy": 0.2453264907002449,
      "num_tokens": 60709941.0,
      "step": 26475
    },
    {
      "entropy": 5.384775972366333,
      "epoch": 2.543707973102786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043551565627862257,
      "loss": 4.7279,
      "mean_token_accuracy": 0.24471736401319505,
      "num_tokens": 60721245.0,
      "step": 26480
    },
    {
      "entropy": 5.363555431365967,
      "epoch": 2.5441882804995197,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004354916300740387,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24171093702316285,
      "num_tokens": 60732763.0,
      "step": 26485
    },
    {
      "entropy": 5.331658601760864,
      "epoch": 2.5446685878962536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004354676001433157,
      "loss": 4.7143,
      "mean_token_accuracy": 0.2517920762300491,
      "num_tokens": 60743192.0,
      "step": 26490
    },
    {
      "entropy": 5.279242658615113,
      "epoch": 2.5451488952929875,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004354435664870116,
      "loss": 4.6346,
      "mean_token_accuracy": 0.24794637858867646,
      "num_tokens": 60754115.0,
      "step": 26495
    },
    {
      "entropy": 5.4424234390258786,
      "epoch": 2.5456292026897214,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043541952910568417,
      "loss": 4.7697,
      "mean_token_accuracy": 0.24611559510231018,
      "num_tokens": 60765516.0,
      "step": 26500
    },
    {
      "entropy": 5.404594802856446,
      "epoch": 2.5461095100864553,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004353954879998916,
      "loss": 4.7549,
      "mean_token_accuracy": 0.2418079048395157,
      "num_tokens": 60776206.0,
      "step": 26505
    },
    {
      "entropy": 5.314809370040893,
      "epoch": 2.546589817483189,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004353714431701922,
      "loss": 4.6931,
      "mean_token_accuracy": 0.24768145978450776,
      "num_tokens": 60786992.0,
      "step": 26510
    },
    {
      "entropy": 5.359896850585938,
      "epoch": 2.547070124879923,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004353473946171441,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23851648420095445,
      "num_tokens": 60797765.0,
      "step": 26515
    },
    {
      "entropy": 5.4057821273803714,
      "epoch": 2.547550432276657,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043532334234130547,
      "loss": 4.7925,
      "mean_token_accuracy": 0.2294461488723755,
      "num_tokens": 60808873.0,
      "step": 26520
    },
    {
      "entropy": 5.438994455337524,
      "epoch": 2.5480307396733908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043529928634323503,
      "loss": 4.7972,
      "mean_token_accuracy": 0.2436169609427452,
      "num_tokens": 60820245.0,
      "step": 26525
    },
    {
      "entropy": 5.390586090087891,
      "epoch": 2.548511047070125,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043527522662349113,
      "loss": 4.7662,
      "mean_token_accuracy": 0.24732738435268403,
      "num_tokens": 60832159.0,
      "step": 26530
    },
    {
      "entropy": 5.34813494682312,
      "epoch": 2.5489913544668585,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004352511631826324,
      "loss": 4.7136,
      "mean_token_accuracy": 0.24659736901521684,
      "num_tokens": 60842957.0,
      "step": 26535
    },
    {
      "entropy": 5.324805641174317,
      "epoch": 2.549471661863593,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004352270960212175,
      "loss": 4.7262,
      "mean_token_accuracy": 0.24628336876630783,
      "num_tokens": 60852537.0,
      "step": 26540
    },
    {
      "entropy": 5.4595105171203615,
      "epoch": 2.5499519692603267,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004352030251398052,
      "loss": 4.8702,
      "mean_token_accuracy": 0.23284553587436677,
      "num_tokens": 60865761.0,
      "step": 26545
    },
    {
      "entropy": 5.444684076309204,
      "epoch": 2.5504322766570606,
      "grad_norm": 1.125,
      "learning_rate": 0.00043517895053895434,
      "loss": 4.7345,
      "mean_token_accuracy": 0.2478494182229042,
      "num_tokens": 60877805.0,
      "step": 26550
    },
    {
      "entropy": 5.389783573150635,
      "epoch": 2.5509125840537945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004351548722192239,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2410367414355278,
      "num_tokens": 60889868.0,
      "step": 26555
    },
    {
      "entropy": 5.338828706741333,
      "epoch": 2.5513928914505284,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004351307901811729,
      "loss": 4.7317,
      "mean_token_accuracy": 0.24959045499563218,
      "num_tokens": 60902046.0,
      "step": 26560
    },
    {
      "entropy": 5.374965381622315,
      "epoch": 2.5518731988472623,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351067044253605,
      "loss": 4.7611,
      "mean_token_accuracy": 0.23977685123682022,
      "num_tokens": 60913447.0,
      "step": 26565
    },
    {
      "entropy": 5.389798498153686,
      "epoch": 2.552353506243996,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043508261495234577,
      "loss": 4.7341,
      "mean_token_accuracy": 0.25174477100372317,
      "num_tokens": 60924920.0,
      "step": 26570
    },
    {
      "entropy": 5.329143810272217,
      "epoch": 2.55283381364073,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000435058521762688,
      "loss": 4.6772,
      "mean_token_accuracy": 0.2484119012951851,
      "num_tokens": 60936937.0,
      "step": 26575
    },
    {
      "entropy": 5.333020544052124,
      "epoch": 2.553314121037464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004350344248569467,
      "loss": 4.6967,
      "mean_token_accuracy": 0.2566064640879631,
      "num_tokens": 60947836.0,
      "step": 26580
    },
    {
      "entropy": 5.352782917022705,
      "epoch": 2.553794428434198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004350103242356813,
      "loss": 4.7941,
      "mean_token_accuracy": 0.2442470073699951,
      "num_tokens": 60959380.0,
      "step": 26585
    },
    {
      "entropy": 5.410522079467773,
      "epoch": 2.5542747358309317,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004349862198994513,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23652493953704834,
      "num_tokens": 60970148.0,
      "step": 26590
    },
    {
      "entropy": 5.394122695922851,
      "epoch": 2.5547550432276656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004349621118488163,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2511217787861824,
      "num_tokens": 60981139.0,
      "step": 26595
    },
    {
      "entropy": 5.323498010635376,
      "epoch": 2.5552353506243994,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004349380000843361,
      "loss": 4.7048,
      "mean_token_accuracy": 0.248105426132679,
      "num_tokens": 60992204.0,
      "step": 26600
    },
    {
      "entropy": 5.303554582595825,
      "epoch": 2.5557156580211338,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004349138846065704,
      "loss": 4.7276,
      "mean_token_accuracy": 0.2464417561888695,
      "num_tokens": 61002198.0,
      "step": 26605
    },
    {
      "entropy": 5.419418525695801,
      "epoch": 2.556195965417867,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004348897654160791,
      "loss": 4.8232,
      "mean_token_accuracy": 0.23842440247535707,
      "num_tokens": 61012987.0,
      "step": 26610
    },
    {
      "entropy": 5.32757978439331,
      "epoch": 2.5566762728146015,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004348656425134223,
      "loss": 4.6051,
      "mean_token_accuracy": 0.2559655636548996,
      "num_tokens": 61023668.0,
      "step": 26615
    },
    {
      "entropy": 5.396190738677978,
      "epoch": 2.5571565802113354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043484151589916,
      "loss": 4.6575,
      "mean_token_accuracy": 0.2523059114813805,
      "num_tokens": 61034375.0,
      "step": 26620
    },
    {
      "entropy": 5.369319581985474,
      "epoch": 2.5576368876080693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004348173855738523,
      "loss": 4.7511,
      "mean_token_accuracy": 0.24006447196006775,
      "num_tokens": 61045625.0,
      "step": 26625
    },
    {
      "entropy": 5.360025119781494,
      "epoch": 2.558117195004803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004347932515380595,
      "loss": 4.7881,
      "mean_token_accuracy": 0.2476736694574356,
      "num_tokens": 61056359.0,
      "step": 26630
    },
    {
      "entropy": 5.373827648162842,
      "epoch": 2.558597502401537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004347691137923418,
      "loss": 4.7156,
      "mean_token_accuracy": 0.24716044664382936,
      "num_tokens": 61067194.0,
      "step": 26635
    },
    {
      "entropy": 5.437444400787354,
      "epoch": 2.559077809798271,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004347449723372598,
      "loss": 4.8117,
      "mean_token_accuracy": 0.23616617918014526,
      "num_tokens": 61079341.0,
      "step": 26640
    },
    {
      "entropy": 5.398040008544922,
      "epoch": 2.559558117195005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004347208271733738,
      "loss": 4.8,
      "mean_token_accuracy": 0.24408307075500488,
      "num_tokens": 61090843.0,
      "step": 26645
    },
    {
      "entropy": 5.382343578338623,
      "epoch": 2.5600384245917387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004346966783012445,
      "loss": 4.7202,
      "mean_token_accuracy": 0.24186955988407136,
      "num_tokens": 61101780.0,
      "step": 26650
    },
    {
      "entropy": 5.388761568069458,
      "epoch": 2.5605187319884726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043467252572143247,
      "loss": 4.7917,
      "mean_token_accuracy": 0.24420867413282393,
      "num_tokens": 61113026.0,
      "step": 26655
    },
    {
      "entropy": 5.4001366138458256,
      "epoch": 2.5609990393852065,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043464836943449866,
      "loss": 4.7488,
      "mean_token_accuracy": 0.24670237749814988,
      "num_tokens": 61124772.0,
      "step": 26660
    },
    {
      "entropy": 5.394095802307129,
      "epoch": 2.5614793467819403,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004346242094410036,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24569027870893478,
      "num_tokens": 61136683.0,
      "step": 26665
    },
    {
      "entropy": 5.3560065746307375,
      "epoch": 2.561959654178674,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004346000457415085,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2496449500322342,
      "num_tokens": 61148047.0,
      "step": 26670
    },
    {
      "entropy": 5.353915214538574,
      "epoch": 2.562439961575408,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043457587833657424,
      "loss": 4.6672,
      "mean_token_accuracy": 0.2504213124513626,
      "num_tokens": 61160680.0,
      "step": 26675
    },
    {
      "entropy": 5.396128225326538,
      "epoch": 2.5629202689721424,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043455170722676194,
      "loss": 4.7687,
      "mean_token_accuracy": 0.23810633569955825,
      "num_tokens": 61172624.0,
      "step": 26680
    },
    {
      "entropy": 5.269804811477661,
      "epoch": 2.563400576368876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004345275324126328,
      "loss": 4.6664,
      "mean_token_accuracy": 0.24835258424282075,
      "num_tokens": 61183971.0,
      "step": 26685
    },
    {
      "entropy": 5.356388664245605,
      "epoch": 2.56388088376561,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043450335389474796,
      "loss": 4.6588,
      "mean_token_accuracy": 0.2614606961607933,
      "num_tokens": 61194511.0,
      "step": 26690
    },
    {
      "entropy": 5.337244939804077,
      "epoch": 2.5643611911623436,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004344791716736689,
      "loss": 4.6839,
      "mean_token_accuracy": 0.24554642885923386,
      "num_tokens": 61205235.0,
      "step": 26695
    },
    {
      "entropy": 5.25635027885437,
      "epoch": 2.564841498559078,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043445498574995705,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24843116104602814,
      "num_tokens": 61217460.0,
      "step": 26700
    },
    {
      "entropy": 5.369704675674439,
      "epoch": 2.565321805955812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043443079612417394,
      "loss": 4.7057,
      "mean_token_accuracy": 0.2450148031115532,
      "num_tokens": 61229917.0,
      "step": 26705
    },
    {
      "entropy": 5.441059732437134,
      "epoch": 2.5658021133525457,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004344066027968812,
      "loss": 4.8487,
      "mean_token_accuracy": 0.23366919457912444,
      "num_tokens": 61241826.0,
      "step": 26710
    },
    {
      "entropy": 5.471963357925415,
      "epoch": 2.5662824207492796,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043438240576864034,
      "loss": 4.843,
      "mean_token_accuracy": 0.24126484990119934,
      "num_tokens": 61254095.0,
      "step": 26715
    },
    {
      "entropy": 5.3957137107849125,
      "epoch": 2.5667627281460135,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004343582050400134,
      "loss": 4.7964,
      "mean_token_accuracy": 0.24766053408384323,
      "num_tokens": 61266114.0,
      "step": 26720
    },
    {
      "entropy": 5.344595813751221,
      "epoch": 2.5672430355427474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004343340006115621,
      "loss": 4.7317,
      "mean_token_accuracy": 0.24893167465925217,
      "num_tokens": 61277289.0,
      "step": 26725
    },
    {
      "entropy": 5.491652011871338,
      "epoch": 2.5677233429394812,
      "grad_norm": 1.125,
      "learning_rate": 0.0004343097924838483,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2381291851401329,
      "num_tokens": 61289424.0,
      "step": 26730
    },
    {
      "entropy": 5.4368185043334964,
      "epoch": 2.568203650336215,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004342855806574344,
      "loss": 4.7589,
      "mean_token_accuracy": 0.24426206946372986,
      "num_tokens": 61300821.0,
      "step": 26735
    },
    {
      "entropy": 5.370675373077392,
      "epoch": 2.568683957732949,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004342613651328822,
      "loss": 4.7056,
      "mean_token_accuracy": 0.24955295771360397,
      "num_tokens": 61312097.0,
      "step": 26740
    },
    {
      "entropy": 5.293587875366211,
      "epoch": 2.569164265129683,
      "grad_norm": 1.046875,
      "learning_rate": 0.000434237145910754,
      "loss": 4.724,
      "mean_token_accuracy": 0.25029616355895995,
      "num_tokens": 61324144.0,
      "step": 26745
    },
    {
      "entropy": 5.349129390716553,
      "epoch": 2.5696445725264168,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043421292299161213,
      "loss": 4.7107,
      "mean_token_accuracy": 0.24471861869096756,
      "num_tokens": 61335809.0,
      "step": 26750
    },
    {
      "entropy": 5.357222270965576,
      "epoch": 2.570124879923151,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043418869637601887,
      "loss": 4.7133,
      "mean_token_accuracy": 0.24992190301418304,
      "num_tokens": 61346980.0,
      "step": 26755
    },
    {
      "entropy": 5.4210282325744625,
      "epoch": 2.5706051873198845,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043416446606453686,
      "loss": 4.8505,
      "mean_token_accuracy": 0.24306728988885878,
      "num_tokens": 61358401.0,
      "step": 26760
    },
    {
      "entropy": 5.325794410705567,
      "epoch": 2.571085494716619,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004341402320577285,
      "loss": 4.7109,
      "mean_token_accuracy": 0.24613125920295714,
      "num_tokens": 61369383.0,
      "step": 26765
    },
    {
      "entropy": 5.397389698028564,
      "epoch": 2.5715658021133523,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004341159943561566,
      "loss": 4.7571,
      "mean_token_accuracy": 0.24182094484567643,
      "num_tokens": 61380439.0,
      "step": 26770
    },
    {
      "entropy": 5.385487127304077,
      "epoch": 2.5720461095100866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340917529603837,
      "loss": 4.6861,
      "mean_token_accuracy": 0.2531160175800323,
      "num_tokens": 61391218.0,
      "step": 26775
    },
    {
      "entropy": 5.369384670257569,
      "epoch": 2.5725264169068205,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004340675078709728,
      "loss": 4.6808,
      "mean_token_accuracy": 0.2478260189294815,
      "num_tokens": 61402715.0,
      "step": 26780
    },
    {
      "entropy": 5.438705253601074,
      "epoch": 2.5730067243035544,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004340432590884866,
      "loss": 4.7778,
      "mean_token_accuracy": 0.23726189285516738,
      "num_tokens": 61414361.0,
      "step": 26785
    },
    {
      "entropy": 5.352579498291016,
      "epoch": 2.5734870317002883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043401900661348825,
      "loss": 4.6911,
      "mean_token_accuracy": 0.25157486349344255,
      "num_tokens": 61426817.0,
      "step": 26790
    },
    {
      "entropy": 5.337439298629761,
      "epoch": 2.573967339097022,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043399475044654073,
      "loss": 4.7043,
      "mean_token_accuracy": 0.2429383620619774,
      "num_tokens": 61438142.0,
      "step": 26795
    },
    {
      "entropy": 5.399322366714477,
      "epoch": 2.574447646493756,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043397049058820726,
      "loss": 4.778,
      "mean_token_accuracy": 0.24163489043712616,
      "num_tokens": 61450708.0,
      "step": 26800
    },
    {
      "entropy": 5.39176983833313,
      "epoch": 2.57492795389049,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000433946227039051,
      "loss": 4.7501,
      "mean_token_accuracy": 0.2419169634580612,
      "num_tokens": 61462266.0,
      "step": 26805
    },
    {
      "entropy": 5.376583576202393,
      "epoch": 2.5754082612872238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004339219597996353,
      "loss": 4.746,
      "mean_token_accuracy": 0.24645287096500396,
      "num_tokens": 61473690.0,
      "step": 26810
    },
    {
      "entropy": 5.264575433731079,
      "epoch": 2.5758885686839577,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004338976888705236,
      "loss": 4.6193,
      "mean_token_accuracy": 0.2551027923822403,
      "num_tokens": 61484997.0,
      "step": 26815
    },
    {
      "entropy": 5.302763366699219,
      "epoch": 2.5763688760806915,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043387341425227944,
      "loss": 4.7804,
      "mean_token_accuracy": 0.24342362135648726,
      "num_tokens": 61497792.0,
      "step": 26820
    },
    {
      "entropy": 5.358625268936157,
      "epoch": 2.5768491834774254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004338491359454664,
      "loss": 4.6414,
      "mean_token_accuracy": 0.24885679334402083,
      "num_tokens": 61509060.0,
      "step": 26825
    },
    {
      "entropy": 5.419218969345093,
      "epoch": 2.5773294908741593,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043382485395064796,
      "loss": 4.7631,
      "mean_token_accuracy": 0.25050225406885146,
      "num_tokens": 61519577.0,
      "step": 26830
    },
    {
      "entropy": 5.315019226074218,
      "epoch": 2.577809798270893,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004338005682683881,
      "loss": 4.7044,
      "mean_token_accuracy": 0.24822279810905457,
      "num_tokens": 61531956.0,
      "step": 26835
    },
    {
      "entropy": 5.377149248123169,
      "epoch": 2.5782901056676275,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043377627889925057,
      "loss": 4.7929,
      "mean_token_accuracy": 0.24410223513841628,
      "num_tokens": 61543801.0,
      "step": 26840
    },
    {
      "entropy": 5.38058500289917,
      "epoch": 2.578770413064361,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004337519858437994,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23784735649824143,
      "num_tokens": 61555899.0,
      "step": 26845
    },
    {
      "entropy": 5.313778638839722,
      "epoch": 2.5792507204610953,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004337276891025984,
      "loss": 4.653,
      "mean_token_accuracy": 0.25808094441890717,
      "num_tokens": 61567280.0,
      "step": 26850
    },
    {
      "entropy": 5.343655157089233,
      "epoch": 2.579731027857829,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043370338867621184,
      "loss": 4.763,
      "mean_token_accuracy": 0.2450314298272133,
      "num_tokens": 61578026.0,
      "step": 26855
    },
    {
      "entropy": 5.374447441101074,
      "epoch": 2.580211335254563,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043367908456520387,
      "loss": 4.7472,
      "mean_token_accuracy": 0.2391871765255928,
      "num_tokens": 61589399.0,
      "step": 26860
    },
    {
      "entropy": 5.4201559066772464,
      "epoch": 2.580691642651297,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004336547767701387,
      "loss": 4.7802,
      "mean_token_accuracy": 0.2426774963736534,
      "num_tokens": 61601025.0,
      "step": 26865
    },
    {
      "entropy": 5.457146406173706,
      "epoch": 2.581171950048031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043363046529158077,
      "loss": 4.857,
      "mean_token_accuracy": 0.2368979573249817,
      "num_tokens": 61612529.0,
      "step": 26870
    },
    {
      "entropy": 5.368618488311768,
      "epoch": 2.5816522574447647,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004336061501300944,
      "loss": 4.7913,
      "mean_token_accuracy": 0.24014250040054322,
      "num_tokens": 61624423.0,
      "step": 26875
    },
    {
      "entropy": 5.418833589553833,
      "epoch": 2.5821325648414986,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004335818312862442,
      "loss": 4.824,
      "mean_token_accuracy": 0.2385423555970192,
      "num_tokens": 61635002.0,
      "step": 26880
    },
    {
      "entropy": 5.319018840789795,
      "epoch": 2.5826128722382324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043355750876059485,
      "loss": 4.7095,
      "mean_token_accuracy": 0.24904280602931977,
      "num_tokens": 61645911.0,
      "step": 26885
    },
    {
      "entropy": 5.360689687728882,
      "epoch": 2.5830931796349663,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043353318255371086,
      "loss": 4.7194,
      "mean_token_accuracy": 0.24682482331991196,
      "num_tokens": 61657545.0,
      "step": 26890
    },
    {
      "entropy": 5.442089080810547,
      "epoch": 2.5835734870317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004335088526661571,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23795273900032043,
      "num_tokens": 61669567.0,
      "step": 26895
    },
    {
      "entropy": 5.448243427276611,
      "epoch": 2.584053794428434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043348451909849855,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2468821093440056,
      "num_tokens": 61681725.0,
      "step": 26900
    },
    {
      "entropy": 5.4312032699584964,
      "epoch": 2.584534101825168,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043346018185130006,
      "loss": 4.809,
      "mean_token_accuracy": 0.2468055620789528,
      "num_tokens": 61692503.0,
      "step": 26905
    },
    {
      "entropy": 5.325273132324218,
      "epoch": 2.585014409221902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043343584092512665,
      "loss": 4.6781,
      "mean_token_accuracy": 0.2537978425621986,
      "num_tokens": 61704567.0,
      "step": 26910
    },
    {
      "entropy": 5.361824607849121,
      "epoch": 2.585494716618636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004334114963205435,
      "loss": 4.7002,
      "mean_token_accuracy": 0.24961762875318527,
      "num_tokens": 61716155.0,
      "step": 26915
    },
    {
      "entropy": 5.357460641860962,
      "epoch": 2.5859750240153696,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004333871480381158,
      "loss": 4.8208,
      "mean_token_accuracy": 0.24146770685911179,
      "num_tokens": 61727489.0,
      "step": 26920
    },
    {
      "entropy": 5.391724920272827,
      "epoch": 2.586455331412104,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004333627960784088,
      "loss": 4.7554,
      "mean_token_accuracy": 0.24655339270830154,
      "num_tokens": 61739792.0,
      "step": 26925
    },
    {
      "entropy": 5.436347913742066,
      "epoch": 2.586935638808838,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004333384404419879,
      "loss": 4.7695,
      "mean_token_accuracy": 0.24293783009052278,
      "num_tokens": 61750891.0,
      "step": 26930
    },
    {
      "entropy": 5.334785842895508,
      "epoch": 2.5874159462055717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004333140811294186,
      "loss": 4.6984,
      "mean_token_accuracy": 0.2531579062342644,
      "num_tokens": 61761691.0,
      "step": 26935
    },
    {
      "entropy": 5.378616809844971,
      "epoch": 2.5878962536023056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004332897181412664,
      "loss": 4.7256,
      "mean_token_accuracy": 0.24318442046642302,
      "num_tokens": 61772371.0,
      "step": 26940
    },
    {
      "entropy": 5.355374670028686,
      "epoch": 2.5883765609990395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043326535147809696,
      "loss": 4.7169,
      "mean_token_accuracy": 0.24630044251680375,
      "num_tokens": 61784908.0,
      "step": 26945
    },
    {
      "entropy": 5.409377193450927,
      "epoch": 2.5888568683957733,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043324098114047604,
      "loss": 4.7901,
      "mean_token_accuracy": 0.23832721263170242,
      "num_tokens": 61795948.0,
      "step": 26950
    },
    {
      "entropy": 5.427465343475342,
      "epoch": 2.589337175792507,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004332166071289695,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2423390805721283,
      "num_tokens": 61807528.0,
      "step": 26955
    },
    {
      "entropy": 5.378372001647949,
      "epoch": 2.589817483189241,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043319222944414304,
      "loss": 4.7079,
      "mean_token_accuracy": 0.25000920444726943,
      "num_tokens": 61819971.0,
      "step": 26960
    },
    {
      "entropy": 5.297861528396607,
      "epoch": 2.590297790585975,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043316784808656276,
      "loss": 4.6818,
      "mean_token_accuracy": 0.25243731290102006,
      "num_tokens": 61832284.0,
      "step": 26965
    },
    {
      "entropy": 5.364977312088013,
      "epoch": 2.590778097982709,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043314346305679477,
      "loss": 4.7834,
      "mean_token_accuracy": 0.23972320556640625,
      "num_tokens": 61844350.0,
      "step": 26970
    },
    {
      "entropy": 5.4279238224029545,
      "epoch": 2.5912584053794427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043311907435540517,
      "loss": 4.7134,
      "mean_token_accuracy": 0.2520176529884338,
      "num_tokens": 61855180.0,
      "step": 26975
    },
    {
      "entropy": 5.402017307281494,
      "epoch": 2.5917387127761766,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004330946819829601,
      "loss": 4.7793,
      "mean_token_accuracy": 0.23829298615455627,
      "num_tokens": 61866963.0,
      "step": 26980
    },
    {
      "entropy": 5.426289129257202,
      "epoch": 2.5922190201729105,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043307028594002597,
      "loss": 4.797,
      "mean_token_accuracy": 0.24344509989023208,
      "num_tokens": 61877503.0,
      "step": 26985
    },
    {
      "entropy": 5.355916118621826,
      "epoch": 2.592699327569645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043304588622716924,
      "loss": 4.7056,
      "mean_token_accuracy": 0.2542313113808632,
      "num_tokens": 61888426.0,
      "step": 26990
    },
    {
      "entropy": 5.377012157440186,
      "epoch": 2.5931796349663783,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004330214828449563,
      "loss": 4.711,
      "mean_token_accuracy": 0.24508720636367798,
      "num_tokens": 61899952.0,
      "step": 26995
    },
    {
      "entropy": 5.303738307952881,
      "epoch": 2.5936599423631126,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043299707579395365,
      "loss": 4.7048,
      "mean_token_accuracy": 0.253624452650547,
      "num_tokens": 61911956.0,
      "step": 27000
    },
    {
      "epoch": 2.5936599423631126,
      "eval_entropy": 5.091165599911344,
      "eval_loss": 4.81263542175293,
      "eval_mean_token_accuracy": 0.250323348698174,
      "eval_num_tokens": 61911956.0,
      "eval_runtime": 20.7707,
      "eval_samples_per_second": 1579.866,
      "eval_steps_per_second": 197.489,
      "step": 27000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.40300389343232e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}