ita-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed455/checkpoint-39000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.870583564906709,
  "eval_steps": 3000,
  "global_step": 39000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.931299161911011,
      "epoch": 0.0004962286621675268,
      "grad_norm": 14.6875,
      "learning_rate": 2e-06,
      "loss": 14.4374,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11135.0,
      "step": 5
    },
    {
      "entropy": 4.938097429275513,
      "epoch": 0.0009924573243350536,
      "grad_norm": 15.375,
      "learning_rate": 4.5e-06,
      "loss": 14.4228,
      "mean_token_accuracy": 0.0,
      "num_tokens": 22299.0,
      "step": 10
    },
    {
      "entropy": 4.951365232467651,
      "epoch": 0.0014886859865025804,
      "grad_norm": 16.375,
      "learning_rate": 7e-06,
      "loss": 14.3239,
      "mean_token_accuracy": 0.0,
      "num_tokens": 33670.0,
      "step": 15
    },
    {
      "entropy": 4.991879177093506,
      "epoch": 0.001984914648670107,
      "grad_norm": 18.625,
      "learning_rate": 9.5e-06,
      "loss": 14.1546,
      "mean_token_accuracy": 0.0,
      "num_tokens": 44240.0,
      "step": 20
    },
    {
      "entropy": 5.11100959777832,
      "epoch": 0.002481143310837634,
      "grad_norm": 22.5,
      "learning_rate": 1.2e-05,
      "loss": 13.6792,
      "mean_token_accuracy": 0.00010111223673447967,
      "num_tokens": 54168.0,
      "step": 25
    },
    {
      "entropy": 5.412902069091797,
      "epoch": 0.0029773719730051607,
      "grad_norm": 29.0,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 13.1651,
      "mean_token_accuracy": 8.11688310932368e-05,
      "num_tokens": 64704.0,
      "step": 30
    },
    {
      "entropy": 7.036752700805664,
      "epoch": 0.0034736006351726877,
      "grad_norm": 22.25,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9985,
      "mean_token_accuracy": 0.0,
      "num_tokens": 75766.0,
      "step": 35
    },
    {
      "entropy": 10.185614490509034,
      "epoch": 0.003969829297340214,
      "grad_norm": 3.25,
      "learning_rate": 1.95e-05,
      "loss": 10.9414,
      "mean_token_accuracy": 0.0002966745407320559,
      "num_tokens": 88074.0,
      "step": 40
    },
    {
      "entropy": 10.717006397247314,
      "epoch": 0.004466057959507741,
      "grad_norm": 2.828125,
      "learning_rate": 2.2e-05,
      "loss": 10.725,
      "mean_token_accuracy": 0.00629717658739537,
      "num_tokens": 99313.0,
      "step": 45
    },
    {
      "entropy": 10.743248081207275,
      "epoch": 0.004962286621675268,
      "grad_norm": 2.59375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.6055,
      "mean_token_accuracy": 0.01345829488709569,
      "num_tokens": 110819.0,
      "step": 50
    },
    {
      "entropy": 10.739305686950683,
      "epoch": 0.0054585152838427945,
      "grad_norm": 1.9921875,
      "learning_rate": 2.7e-05,
      "loss": 10.3942,
      "mean_token_accuracy": 0.016125928703695537,
      "num_tokens": 121156.0,
      "step": 55
    },
    {
      "entropy": 10.702740478515626,
      "epoch": 0.0059547439460103215,
      "grad_norm": 2.046875,
      "learning_rate": 2.95e-05,
      "loss": 10.2317,
      "mean_token_accuracy": 0.016230483166873456,
      "num_tokens": 131307.0,
      "step": 60
    },
    {
      "entropy": 10.715532684326172,
      "epoch": 0.0064509726081778485,
      "grad_norm": 1.921875,
      "learning_rate": 3.2e-05,
      "loss": 10.0947,
      "mean_token_accuracy": 0.01627691639587283,
      "num_tokens": 141812.0,
      "step": 65
    },
    {
      "entropy": 10.737543678283691,
      "epoch": 0.0069472012703453755,
      "grad_norm": 1.703125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 10.0193,
      "mean_token_accuracy": 0.01881603915244341,
      "num_tokens": 152425.0,
      "step": 70
    },
    {
      "entropy": 10.738888168334961,
      "epoch": 0.007443429932512902,
      "grad_norm": 1.9140625,
      "learning_rate": 3.7e-05,
      "loss": 9.902,
      "mean_token_accuracy": 0.029196479916572572,
      "num_tokens": 162562.0,
      "step": 75
    },
    {
      "entropy": 10.733177089691162,
      "epoch": 0.007939658594680429,
      "grad_norm": 1.9296875,
      "learning_rate": 3.95e-05,
      "loss": 9.835,
      "mean_token_accuracy": 0.038548733852803704,
      "num_tokens": 173322.0,
      "step": 80
    },
    {
      "entropy": 10.730626487731934,
      "epoch": 0.008435887256847956,
      "grad_norm": 1.765625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.7586,
      "mean_token_accuracy": 0.04156450685113668,
      "num_tokens": 183169.0,
      "step": 85
    },
    {
      "entropy": 10.72908525466919,
      "epoch": 0.008932115919015483,
      "grad_norm": 1.6875,
      "learning_rate": 4.45e-05,
      "loss": 9.7319,
      "mean_token_accuracy": 0.040352378785610196,
      "num_tokens": 194473.0,
      "step": 90
    },
    {
      "entropy": 10.723829555511475,
      "epoch": 0.00942834458118301,
      "grad_norm": 1.6953125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.5914,
      "mean_token_accuracy": 0.041312170401215556,
      "num_tokens": 206166.0,
      "step": 95
    },
    {
      "entropy": 10.710507011413574,
      "epoch": 0.009924573243350537,
      "grad_norm": 1.703125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4691,
      "mean_token_accuracy": 0.04619110412895679,
      "num_tokens": 215923.0,
      "step": 100
    },
    {
      "entropy": 10.687235641479493,
      "epoch": 0.010420801905518062,
      "grad_norm": 1.765625,
      "learning_rate": 5.2e-05,
      "loss": 9.3768,
      "mean_token_accuracy": 0.040744516998529434,
      "num_tokens": 227139.0,
      "step": 105
    },
    {
      "entropy": 10.66772632598877,
      "epoch": 0.010917030567685589,
      "grad_norm": 1.734375,
      "learning_rate": 5.45e-05,
      "loss": 9.2519,
      "mean_token_accuracy": 0.04524887017905712,
      "num_tokens": 237770.0,
      "step": 110
    },
    {
      "entropy": 10.642418003082275,
      "epoch": 0.011413259229853116,
      "grad_norm": 1.5703125,
      "learning_rate": 5.7e-05,
      "loss": 9.1603,
      "mean_token_accuracy": 0.050716237351298335,
      "num_tokens": 249241.0,
      "step": 115
    },
    {
      "entropy": 10.606610870361328,
      "epoch": 0.011909487892020643,
      "grad_norm": 1.6484375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.9869,
      "mean_token_accuracy": 0.046933449991047385,
      "num_tokens": 258562.0,
      "step": 120
    },
    {
      "entropy": 10.56153678894043,
      "epoch": 0.01240571655418817,
      "grad_norm": 1.6328125,
      "learning_rate": 6.2e-05,
      "loss": 8.956,
      "mean_token_accuracy": 0.049542845040559766,
      "num_tokens": 269548.0,
      "step": 125
    },
    {
      "entropy": 10.512712478637695,
      "epoch": 0.012901945216355697,
      "grad_norm": 1.53125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.8768,
      "mean_token_accuracy": 0.0483577411621809,
      "num_tokens": 280965.0,
      "step": 130
    },
    {
      "entropy": 10.447002983093261,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.515625,
      "learning_rate": 6.7e-05,
      "loss": 8.6644,
      "mean_token_accuracy": 0.05167540945112705,
      "num_tokens": 290063.0,
      "step": 135
    },
    {
      "entropy": 10.371509552001953,
      "epoch": 0.013894402540690751,
      "grad_norm": 1.5546875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.6306,
      "mean_token_accuracy": 0.047542725875973704,
      "num_tokens": 301013.0,
      "step": 140
    },
    {
      "entropy": 10.292595195770264,
      "epoch": 0.014390631202858276,
      "grad_norm": 1.4921875,
      "learning_rate": 7.2e-05,
      "loss": 8.5113,
      "mean_token_accuracy": 0.049639220163226125,
      "num_tokens": 313351.0,
      "step": 145
    },
    {
      "entropy": 10.18106632232666,
      "epoch": 0.014886859865025803,
      "grad_norm": 1.3359375,
      "learning_rate": 7.45e-05,
      "loss": 8.4824,
      "mean_token_accuracy": 0.047823553532361986,
      "num_tokens": 324871.0,
      "step": 150
    },
    {
      "entropy": 10.063039779663086,
      "epoch": 0.01538308852719333,
      "grad_norm": 1.1875,
      "learning_rate": 7.7e-05,
      "loss": 8.2836,
      "mean_token_accuracy": 0.048729633539915086,
      "num_tokens": 335856.0,
      "step": 155
    },
    {
      "entropy": 9.891231155395507,
      "epoch": 0.015879317189360857,
      "grad_norm": 1.28125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.2022,
      "mean_token_accuracy": 0.05146194100379944,
      "num_tokens": 345410.0,
      "step": 160
    },
    {
      "entropy": 9.697038269042968,
      "epoch": 0.016375545851528384,
      "grad_norm": 1.421875,
      "learning_rate": 8.2e-05,
      "loss": 8.1555,
      "mean_token_accuracy": 0.048843025043606755,
      "num_tokens": 356370.0,
      "step": 165
    },
    {
      "entropy": 9.49948616027832,
      "epoch": 0.01687177451369591,
      "grad_norm": 1.296875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.0734,
      "mean_token_accuracy": 0.04591990895569324,
      "num_tokens": 368085.0,
      "step": 170
    },
    {
      "entropy": 9.283473300933839,
      "epoch": 0.01736800317586344,
      "grad_norm": 0.88671875,
      "learning_rate": 8.7e-05,
      "loss": 7.925,
      "mean_token_accuracy": 0.04852883592247963,
      "num_tokens": 378995.0,
      "step": 175
    },
    {
      "entropy": 9.047716808319091,
      "epoch": 0.017864231838030965,
      "grad_norm": 0.69921875,
      "learning_rate": 8.95e-05,
      "loss": 7.9793,
      "mean_token_accuracy": 0.0445900060236454,
      "num_tokens": 390727.0,
      "step": 180
    },
    {
      "entropy": 8.858506870269775,
      "epoch": 0.018360460500198492,
      "grad_norm": 0.71875,
      "learning_rate": 9.2e-05,
      "loss": 7.8485,
      "mean_token_accuracy": 0.04807438552379608,
      "num_tokens": 401875.0,
      "step": 185
    },
    {
      "entropy": 8.66808204650879,
      "epoch": 0.01885668916236602,
      "grad_norm": 0.671875,
      "learning_rate": 9.45e-05,
      "loss": 7.8301,
      "mean_token_accuracy": 0.05265191309154034,
      "num_tokens": 414959.0,
      "step": 190
    },
    {
      "entropy": 8.465768527984618,
      "epoch": 0.019352917824533546,
      "grad_norm": 0.69140625,
      "learning_rate": 9.7e-05,
      "loss": 7.7787,
      "mean_token_accuracy": 0.05121605955064297,
      "num_tokens": 425745.0,
      "step": 195
    },
    {
      "entropy": 8.350896739959717,
      "epoch": 0.019849146486701073,
      "grad_norm": 1.0,
      "learning_rate": 9.95e-05,
      "loss": 7.7229,
      "mean_token_accuracy": 0.04516980387270451,
      "num_tokens": 435149.0,
      "step": 200
    },
    {
      "entropy": 8.277827644348145,
      "epoch": 0.0203453751488686,
      "grad_norm": 0.859375,
      "learning_rate": 0.000102,
      "loss": 7.7528,
      "mean_token_accuracy": 0.04675165638327598,
      "num_tokens": 445706.0,
      "step": 205
    },
    {
      "entropy": 8.202262878417969,
      "epoch": 0.020841603811036124,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.7464,
      "mean_token_accuracy": 0.051094313710927965,
      "num_tokens": 455982.0,
      "step": 210
    },
    {
      "entropy": 8.231989574432372,
      "epoch": 0.02133783247320365,
      "grad_norm": 0.84375,
      "learning_rate": 0.000107,
      "loss": 7.7547,
      "mean_token_accuracy": 0.04971891045570374,
      "num_tokens": 465462.0,
      "step": 215
    },
    {
      "entropy": 8.276477718353272,
      "epoch": 0.021834061135371178,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0001095,
      "loss": 7.7201,
      "mean_token_accuracy": 0.049120894819498065,
      "num_tokens": 475770.0,
      "step": 220
    },
    {
      "entropy": 8.2308780670166,
      "epoch": 0.022330289797538705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000112,
      "loss": 7.6536,
      "mean_token_accuracy": 0.05627523325383663,
      "num_tokens": 486345.0,
      "step": 225
    },
    {
      "entropy": 8.243096160888673,
      "epoch": 0.022826518459706232,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0001145,
      "loss": 7.6637,
      "mean_token_accuracy": 0.057797220349311826,
      "num_tokens": 496588.0,
      "step": 230
    },
    {
      "entropy": 8.189409923553466,
      "epoch": 0.02332274712187376,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.6064,
      "mean_token_accuracy": 0.0653748583048582,
      "num_tokens": 506676.0,
      "step": 235
    },
    {
      "entropy": 8.106929397583007,
      "epoch": 0.023818975784041286,
      "grad_norm": 0.703125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.6016,
      "mean_token_accuracy": 0.06197558268904686,
      "num_tokens": 516866.0,
      "step": 240
    },
    {
      "entropy": 8.121264171600341,
      "epoch": 0.024315204446208813,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000122,
      "loss": 7.6847,
      "mean_token_accuracy": 0.0619472224265337,
      "num_tokens": 529349.0,
      "step": 245
    },
    {
      "entropy": 8.116961669921874,
      "epoch": 0.02481143310837634,
      "grad_norm": 0.8125,
      "learning_rate": 0.0001245,
      "loss": 7.6282,
      "mean_token_accuracy": 0.05949817337095738,
      "num_tokens": 541310.0,
      "step": 250
    },
    {
      "entropy": 8.116946792602539,
      "epoch": 0.025307661770543867,
      "grad_norm": 1.109375,
      "learning_rate": 0.000127,
      "loss": 7.6313,
      "mean_token_accuracy": 0.06268465593457222,
      "num_tokens": 551897.0,
      "step": 255
    },
    {
      "entropy": 8.077956581115723,
      "epoch": 0.025803890432711394,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001295,
      "loss": 7.6347,
      "mean_token_accuracy": 0.06445351354777813,
      "num_tokens": 564667.0,
      "step": 260
    },
    {
      "entropy": 8.109729480743407,
      "epoch": 0.02630011909487892,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000132,
      "loss": 7.51,
      "mean_token_accuracy": 0.07109251022338867,
      "num_tokens": 575338.0,
      "step": 265
    },
    {
      "entropy": 8.040228891372681,
      "epoch": 0.026796347757046448,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.6347,
      "mean_token_accuracy": 0.06142906881868839,
      "num_tokens": 586169.0,
      "step": 270
    },
    {
      "entropy": 8.13642110824585,
      "epoch": 0.027292576419213975,
      "grad_norm": 1.453125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.5917,
      "mean_token_accuracy": 0.06226538605988026,
      "num_tokens": 596864.0,
      "step": 275
    },
    {
      "entropy": 8.001047468185424,
      "epoch": 0.027788805081381502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001395,
      "loss": 7.545,
      "mean_token_accuracy": 0.06418961510062218,
      "num_tokens": 608894.0,
      "step": 280
    },
    {
      "entropy": 8.050166368484497,
      "epoch": 0.02828503374354903,
      "grad_norm": 1.0,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.5519,
      "mean_token_accuracy": 0.06657364331185818,
      "num_tokens": 619863.0,
      "step": 285
    },
    {
      "entropy": 8.061230039596557,
      "epoch": 0.028781262405716553,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001445,
      "loss": 7.5501,
      "mean_token_accuracy": 0.06628488451242447,
      "num_tokens": 632691.0,
      "step": 290
    },
    {
      "entropy": 8.056870412826537,
      "epoch": 0.02927749106788408,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000147,
      "loss": 7.4641,
      "mean_token_accuracy": 0.06920699514448643,
      "num_tokens": 643993.0,
      "step": 295
    },
    {
      "entropy": 8.001115942001343,
      "epoch": 0.029773719730051607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001495,
      "loss": 7.4315,
      "mean_token_accuracy": 0.06977621465921402,
      "num_tokens": 654275.0,
      "step": 300
    },
    {
      "entropy": 7.829830694198608,
      "epoch": 0.030269948392219134,
      "grad_norm": 1.0,
      "learning_rate": 0.000152,
      "loss": 7.4137,
      "mean_token_accuracy": 0.0679971806704998,
      "num_tokens": 664694.0,
      "step": 305
    },
    {
      "entropy": 8.01261920928955,
      "epoch": 0.03076617705438666,
      "grad_norm": 0.953125,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.4346,
      "mean_token_accuracy": 0.06760629676282406,
      "num_tokens": 674736.0,
      "step": 310
    },
    {
      "entropy": 7.828978490829468,
      "epoch": 0.03126240571655419,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000157,
      "loss": 7.3673,
      "mean_token_accuracy": 0.06551955416798591,
      "num_tokens": 687261.0,
      "step": 315
    },
    {
      "entropy": 7.922390127182007,
      "epoch": 0.031758634378721715,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0001595,
      "loss": 7.3618,
      "mean_token_accuracy": 0.07412862330675125,
      "num_tokens": 698904.0,
      "step": 320
    },
    {
      "entropy": 7.855106163024902,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.34375,
      "learning_rate": 0.000162,
      "loss": 7.3364,
      "mean_token_accuracy": 0.07321774028241634,
      "num_tokens": 709035.0,
      "step": 325
    },
    {
      "entropy": 7.789221477508545,
      "epoch": 0.03275109170305677,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.3969,
      "mean_token_accuracy": 0.07102149352431297,
      "num_tokens": 719565.0,
      "step": 330
    },
    {
      "entropy": 7.915437698364258,
      "epoch": 0.03324732036522429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.4084,
      "mean_token_accuracy": 0.07039739079773426,
      "num_tokens": 730329.0,
      "step": 335
    },
    {
      "entropy": 7.888462257385254,
      "epoch": 0.03374354902739182,
      "grad_norm": 1.0,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.4161,
      "mean_token_accuracy": 0.06667844317853451,
      "num_tokens": 741258.0,
      "step": 340
    },
    {
      "entropy": 7.846066999435425,
      "epoch": 0.034239777689559346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.3014,
      "mean_token_accuracy": 0.06842922903597355,
      "num_tokens": 752636.0,
      "step": 345
    },
    {
      "entropy": 7.859539461135864,
      "epoch": 0.03473600635172688,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.3813,
      "mean_token_accuracy": 0.06898528710007668,
      "num_tokens": 764113.0,
      "step": 350
    },
    {
      "entropy": 7.697615098953247,
      "epoch": 0.0352322350138944,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000177,
      "loss": 7.2791,
      "mean_token_accuracy": 0.07269814386963844,
      "num_tokens": 772872.0,
      "step": 355
    },
    {
      "entropy": 7.79153094291687,
      "epoch": 0.03572846367606193,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0001795,
      "loss": 7.2548,
      "mean_token_accuracy": 0.07947108298540115,
      "num_tokens": 782788.0,
      "step": 360
    },
    {
      "entropy": 7.759897947311401,
      "epoch": 0.036224692338229454,
      "grad_norm": 0.984375,
      "learning_rate": 0.000182,
      "loss": 7.3299,
      "mean_token_accuracy": 0.07359667010605335,
      "num_tokens": 793859.0,
      "step": 365
    },
    {
      "entropy": 7.739173507690429,
      "epoch": 0.036720921000396985,
      "grad_norm": 0.90625,
      "learning_rate": 0.0001845,
      "loss": 7.3104,
      "mean_token_accuracy": 0.07008818164467812,
      "num_tokens": 804906.0,
      "step": 370
    },
    {
      "entropy": 7.807758951187134,
      "epoch": 0.03721714966256451,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000187,
      "loss": 7.3094,
      "mean_token_accuracy": 0.07675720825791359,
      "num_tokens": 815520.0,
      "step": 375
    },
    {
      "entropy": 7.60875358581543,
      "epoch": 0.03771337832473204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0001895,
      "loss": 7.3141,
      "mean_token_accuracy": 0.07395646013319493,
      "num_tokens": 826395.0,
      "step": 380
    },
    {
      "entropy": 7.747971534729004,
      "epoch": 0.03820960698689956,
      "grad_norm": 1.171875,
      "learning_rate": 0.000192,
      "loss": 7.2688,
      "mean_token_accuracy": 0.07527201510965824,
      "num_tokens": 836627.0,
      "step": 385
    },
    {
      "entropy": 7.64195990562439,
      "epoch": 0.03870583564906709,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0001945,
      "loss": 7.3072,
      "mean_token_accuracy": 0.06740010790526867,
      "num_tokens": 848250.0,
      "step": 390
    },
    {
      "entropy": 7.7738360404968265,
      "epoch": 0.039202064311234616,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.2937,
      "mean_token_accuracy": 0.07260859198868275,
      "num_tokens": 859195.0,
      "step": 395
    },
    {
      "entropy": 7.725459146499634,
      "epoch": 0.03969829297340215,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.2597,
      "mean_token_accuracy": 0.078775167837739,
      "num_tokens": 871955.0,
      "step": 400
    },
    {
      "entropy": 7.691169404983521,
      "epoch": 0.04019452163556967,
      "grad_norm": 0.84375,
      "learning_rate": 0.000202,
      "loss": 7.2548,
      "mean_token_accuracy": 0.0749235574156046,
      "num_tokens": 882729.0,
      "step": 405
    },
    {
      "entropy": 7.6321714401245115,
      "epoch": 0.0406907502977372,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.1675,
      "mean_token_accuracy": 0.07198959216475487,
      "num_tokens": 893183.0,
      "step": 410
    },
    {
      "entropy": 7.637913227081299,
      "epoch": 0.041186978959904724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000207,
      "loss": 7.2852,
      "mean_token_accuracy": 0.07573982179164887,
      "num_tokens": 904988.0,
      "step": 415
    },
    {
      "entropy": 7.715639543533325,
      "epoch": 0.04168320762207225,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002095,
      "loss": 7.1686,
      "mean_token_accuracy": 0.07994935102760792,
      "num_tokens": 915192.0,
      "step": 420
    },
    {
      "entropy": 7.647213315963745,
      "epoch": 0.04217943628423978,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000212,
      "loss": 7.2004,
      "mean_token_accuracy": 0.07983001545071602,
      "num_tokens": 926494.0,
      "step": 425
    },
    {
      "entropy": 7.671146392822266,
      "epoch": 0.0426756649464073,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002145,
      "loss": 7.1945,
      "mean_token_accuracy": 0.07794273421168327,
      "num_tokens": 938376.0,
      "step": 430
    },
    {
      "entropy": 7.543527317047119,
      "epoch": 0.04317189360857483,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.1208,
      "mean_token_accuracy": 0.0846080757677555,
      "num_tokens": 948640.0,
      "step": 435
    },
    {
      "entropy": 7.636675453186035,
      "epoch": 0.043668122270742356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002195,
      "loss": 7.1961,
      "mean_token_accuracy": 0.07409511283040046,
      "num_tokens": 959759.0,
      "step": 440
    },
    {
      "entropy": 7.534220218658447,
      "epoch": 0.044164350932909886,
      "grad_norm": 1.1875,
      "learning_rate": 0.000222,
      "loss": 7.1323,
      "mean_token_accuracy": 0.08176640719175339,
      "num_tokens": 970137.0,
      "step": 445
    },
    {
      "entropy": 7.612626314163208,
      "epoch": 0.04466057959507741,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002245,
      "loss": 7.1693,
      "mean_token_accuracy": 0.076291673630476,
      "num_tokens": 979875.0,
      "step": 450
    },
    {
      "entropy": 7.50954155921936,
      "epoch": 0.04515680825724494,
      "grad_norm": 0.953125,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.1317,
      "mean_token_accuracy": 0.08605509772896766,
      "num_tokens": 990324.0,
      "step": 455
    },
    {
      "entropy": 7.6120024681091305,
      "epoch": 0.045653036919412464,
      "grad_norm": 0.90625,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.1461,
      "mean_token_accuracy": 0.08133600316941739,
      "num_tokens": 1000259.0,
      "step": 460
    },
    {
      "entropy": 7.572062015533447,
      "epoch": 0.046149265581579994,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.1323,
      "mean_token_accuracy": 0.08284945711493492,
      "num_tokens": 1012220.0,
      "step": 465
    },
    {
      "entropy": 7.614422273635864,
      "epoch": 0.04664549424374752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.1318,
      "mean_token_accuracy": 0.078894879296422,
      "num_tokens": 1022968.0,
      "step": 470
    },
    {
      "entropy": 7.54720253944397,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.21875,
      "learning_rate": 0.000237,
      "loss": 7.139,
      "mean_token_accuracy": 0.07864338010549546,
      "num_tokens": 1033415.0,
      "step": 475
    },
    {
      "entropy": 7.455529022216797,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002395,
      "loss": 7.0479,
      "mean_token_accuracy": 0.08645785674452781,
      "num_tokens": 1045384.0,
      "step": 480
    },
    {
      "entropy": 7.553596878051758,
      "epoch": 0.0481341802302501,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000242,
      "loss": 7.1197,
      "mean_token_accuracy": 0.08306692317128181,
      "num_tokens": 1054982.0,
      "step": 485
    },
    {
      "entropy": 7.476700925827027,
      "epoch": 0.048630408892417626,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002445,
      "loss": 7.0713,
      "mean_token_accuracy": 0.08579476028680802,
      "num_tokens": 1065244.0,
      "step": 490
    },
    {
      "entropy": 7.534849357604981,
      "epoch": 0.04912663755458515,
      "grad_norm": 0.953125,
      "learning_rate": 0.000247,
      "loss": 7.0722,
      "mean_token_accuracy": 0.08112104386091232,
      "num_tokens": 1075236.0,
      "step": 495
    },
    {
      "entropy": 7.512158679962158,
      "epoch": 0.04962286621675268,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002495,
      "loss": 7.0981,
      "mean_token_accuracy": 0.07874655723571777,
      "num_tokens": 1085397.0,
      "step": 500
    },
    {
      "entropy": 7.460465049743652,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.296875,
      "learning_rate": 0.000252,
      "loss": 7.097,
      "mean_token_accuracy": 0.08893987089395523,
      "num_tokens": 1095207.0,
      "step": 505
    },
    {
      "entropy": 7.566689109802246,
      "epoch": 0.050615323541087734,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0002545,
      "loss": 7.0936,
      "mean_token_accuracy": 0.07820081040263176,
      "num_tokens": 1105834.0,
      "step": 510
    },
    {
      "entropy": 7.45391845703125,
      "epoch": 0.05111155220325526,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000257,
      "loss": 7.1052,
      "mean_token_accuracy": 0.07973617538809777,
      "num_tokens": 1116105.0,
      "step": 515
    },
    {
      "entropy": 7.5173252582550045,
      "epoch": 0.05160778086542279,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0002595,
      "loss": 7.0979,
      "mean_token_accuracy": 0.0877754531800747,
      "num_tokens": 1125298.0,
      "step": 520
    },
    {
      "entropy": 7.461072683334351,
      "epoch": 0.05210400952759031,
      "grad_norm": 0.7421875,
      "learning_rate": 0.000262,
      "loss": 7.0785,
      "mean_token_accuracy": 0.0839509479701519,
      "num_tokens": 1136797.0,
      "step": 525
    },
    {
      "entropy": 7.396866083145142,
      "epoch": 0.05260023818975784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.0534,
      "mean_token_accuracy": 0.08347654044628143,
      "num_tokens": 1148674.0,
      "step": 530
    },
    {
      "entropy": 7.563003301620483,
      "epoch": 0.053096466851925365,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.1343,
      "mean_token_accuracy": 0.07938608936965466,
      "num_tokens": 1160329.0,
      "step": 535
    },
    {
      "entropy": 7.424808549880981,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.0971,
      "mean_token_accuracy": 0.08222321793437004,
      "num_tokens": 1170318.0,
      "step": 540
    },
    {
      "entropy": 7.452347183227539,
      "epoch": 0.05408892417626042,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.9431,
      "mean_token_accuracy": 0.09257944747805595,
      "num_tokens": 1181634.0,
      "step": 545
    },
    {
      "entropy": 7.393966245651245,
      "epoch": 0.05458515283842795,
      "grad_norm": 1.0,
      "learning_rate": 0.0002745,
      "loss": 7.0082,
      "mean_token_accuracy": 0.09105771631002427,
      "num_tokens": 1191829.0,
      "step": 550
    },
    {
      "entropy": 7.436964321136474,
      "epoch": 0.055081381500595473,
      "grad_norm": 0.984375,
      "learning_rate": 0.000277,
      "loss": 7.0071,
      "mean_token_accuracy": 0.08593062907457352,
      "num_tokens": 1202265.0,
      "step": 555
    },
    {
      "entropy": 7.300072574615479,
      "epoch": 0.055577610162763004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002795,
      "loss": 7.0075,
      "mean_token_accuracy": 0.08627107739448547,
      "num_tokens": 1214173.0,
      "step": 560
    },
    {
      "entropy": 7.399516344070435,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.921,
      "mean_token_accuracy": 0.08665959686040878,
      "num_tokens": 1224582.0,
      "step": 565
    },
    {
      "entropy": 7.463116979598999,
      "epoch": 0.05657006748709806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002845,
      "loss": 7.1136,
      "mean_token_accuracy": 0.0828106202185154,
      "num_tokens": 1235865.0,
      "step": 570
    },
    {
      "entropy": 7.282903718948364,
      "epoch": 0.05706629614926558,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000287,
      "loss": 6.9459,
      "mean_token_accuracy": 0.08898672014474869,
      "num_tokens": 1247194.0,
      "step": 575
    },
    {
      "entropy": 7.277571678161621,
      "epoch": 0.057562524811433105,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002895,
      "loss": 6.949,
      "mean_token_accuracy": 0.08503150045871735,
      "num_tokens": 1257981.0,
      "step": 580
    },
    {
      "entropy": 7.39735426902771,
      "epoch": 0.058058753473600636,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000292,
      "loss": 6.9111,
      "mean_token_accuracy": 0.09066925644874572,
      "num_tokens": 1268443.0,
      "step": 585
    },
    {
      "entropy": 7.440923500061035,
      "epoch": 0.05855498213576816,
      "grad_norm": 1.0,
      "learning_rate": 0.0002945,
      "loss": 7.0199,
      "mean_token_accuracy": 0.08563335537910462,
      "num_tokens": 1279134.0,
      "step": 590
    },
    {
      "entropy": 7.3426929950714115,
      "epoch": 0.05905121079793569,
      "grad_norm": 0.875,
      "learning_rate": 0.000297,
      "loss": 6.9811,
      "mean_token_accuracy": 0.07985961809754372,
      "num_tokens": 1289965.0,
      "step": 595
    },
    {
      "entropy": 7.358950996398926,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002995,
      "loss": 6.9003,
      "mean_token_accuracy": 0.08630152940750122,
      "num_tokens": 1300003.0,
      "step": 600
    },
    {
      "entropy": 7.291685724258423,
      "epoch": 0.060043668122270744,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000302,
      "loss": 6.9428,
      "mean_token_accuracy": 0.08731391057372093,
      "num_tokens": 1311990.0,
      "step": 605
    },
    {
      "entropy": 7.223467254638672,
      "epoch": 0.06053989678443827,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003045,
      "loss": 6.981,
      "mean_token_accuracy": 0.08844437226653099,
      "num_tokens": 1322963.0,
      "step": 610
    },
    {
      "entropy": 7.410216188430786,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000307,
      "loss": 6.9781,
      "mean_token_accuracy": 0.08575883433222771,
      "num_tokens": 1334580.0,
      "step": 615
    },
    {
      "entropy": 7.334858226776123,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003095,
      "loss": 6.9336,
      "mean_token_accuracy": 0.08547274991869927,
      "num_tokens": 1346545.0,
      "step": 620
    },
    {
      "entropy": 7.3943980693817135,
      "epoch": 0.06202858277094085,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000312,
      "loss": 6.919,
      "mean_token_accuracy": 0.08397532440721989,
      "num_tokens": 1358375.0,
      "step": 625
    },
    {
      "entropy": 7.2383945941925045,
      "epoch": 0.06252481143310838,
      "grad_norm": 1.0,
      "learning_rate": 0.0003145,
      "loss": 6.9856,
      "mean_token_accuracy": 0.08264455795288086,
      "num_tokens": 1368376.0,
      "step": 630
    },
    {
      "entropy": 7.319325971603393,
      "epoch": 0.0630210400952759,
      "grad_norm": 1.03125,
      "learning_rate": 0.000317,
      "loss": 6.864,
      "mean_token_accuracy": 0.08814716339111328,
      "num_tokens": 1377749.0,
      "step": 635
    },
    {
      "entropy": 7.229701232910156,
      "epoch": 0.06351726875744343,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003195,
      "loss": 6.914,
      "mean_token_accuracy": 0.08655236959457398,
      "num_tokens": 1389581.0,
      "step": 640
    },
    {
      "entropy": 7.250205373764038,
      "epoch": 0.06401349741961096,
      "grad_norm": 0.9375,
      "learning_rate": 0.000322,
      "loss": 6.8497,
      "mean_token_accuracy": 0.08738525807857514,
      "num_tokens": 1399772.0,
      "step": 645
    },
    {
      "entropy": 7.267027425765991,
      "epoch": 0.06450972608177849,
      "grad_norm": 1.40625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.9473,
      "mean_token_accuracy": 0.08941742517054081,
      "num_tokens": 1411158.0,
      "step": 650
    },
    {
      "entropy": 7.267431354522705,
      "epoch": 0.065005954743946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.8452,
      "mean_token_accuracy": 0.08903275355696678,
      "num_tokens": 1420651.0,
      "step": 655
    },
    {
      "entropy": 7.18158540725708,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.7959,
      "mean_token_accuracy": 0.08788748160004616,
      "num_tokens": 1431628.0,
      "step": 660
    },
    {
      "entropy": 7.198916244506836,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.8263,
      "mean_token_accuracy": 0.09162173792719841,
      "num_tokens": 1441850.0,
      "step": 665
    },
    {
      "entropy": 7.2991822242736815,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.903,
      "mean_token_accuracy": 0.09288545921444893,
      "num_tokens": 1452393.0,
      "step": 670
    },
    {
      "entropy": 7.235831642150879,
      "epoch": 0.06699086939261611,
      "grad_norm": 1.453125,
      "learning_rate": 0.000337,
      "loss": 6.8342,
      "mean_token_accuracy": 0.09165843799710274,
      "num_tokens": 1463109.0,
      "step": 675
    },
    {
      "entropy": 7.1810173988342285,
      "epoch": 0.06748709805478365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003395,
      "loss": 6.9279,
      "mean_token_accuracy": 0.0936855673789978,
      "num_tokens": 1475082.0,
      "step": 680
    },
    {
      "entropy": 7.174162340164185,
      "epoch": 0.06798332671695118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000342,
      "loss": 6.7829,
      "mean_token_accuracy": 0.09301829412579536,
      "num_tokens": 1485477.0,
      "step": 685
    },
    {
      "entropy": 7.1776305675506595,
      "epoch": 0.06847955537911869,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.7743,
      "mean_token_accuracy": 0.09528700262308121,
      "num_tokens": 1495422.0,
      "step": 690
    },
    {
      "entropy": 7.233342695236206,
      "epoch": 0.06897578404128622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000347,
      "loss": 6.799,
      "mean_token_accuracy": 0.08826558515429497,
      "num_tokens": 1506368.0,
      "step": 695
    },
    {
      "entropy": 7.151353263854981,
      "epoch": 0.06947201270345375,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003495,
      "loss": 6.801,
      "mean_token_accuracy": 0.09386853724718094,
      "num_tokens": 1517437.0,
      "step": 700
    },
    {
      "entropy": 7.06351146697998,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.203125,
      "learning_rate": 0.000352,
      "loss": 6.7397,
      "mean_token_accuracy": 0.09502186775207519,
      "num_tokens": 1527801.0,
      "step": 705
    },
    {
      "entropy": 7.185938215255737,
      "epoch": 0.0704644700277888,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003545,
      "loss": 6.8284,
      "mean_token_accuracy": 0.0896036982536316,
      "num_tokens": 1537621.0,
      "step": 710
    },
    {
      "entropy": 7.186826467514038,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.0625,
      "learning_rate": 0.000357,
      "loss": 6.8201,
      "mean_token_accuracy": 0.08602025657892227,
      "num_tokens": 1549123.0,
      "step": 715
    },
    {
      "entropy": 7.1724272727966305,
      "epoch": 0.07145692735212386,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003595,
      "loss": 6.8183,
      "mean_token_accuracy": 0.09465417116880417,
      "num_tokens": 1561167.0,
      "step": 720
    },
    {
      "entropy": 7.086180877685547,
      "epoch": 0.07195315601429139,
      "grad_norm": 1.140625,
      "learning_rate": 0.000362,
      "loss": 6.73,
      "mean_token_accuracy": 0.09641935527324677,
      "num_tokens": 1571552.0,
      "step": 725
    },
    {
      "entropy": 7.058502244949341,
      "epoch": 0.07244938467645891,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003645,
      "loss": 6.6992,
      "mean_token_accuracy": 0.09181095212697983,
      "num_tokens": 1582558.0,
      "step": 730
    },
    {
      "entropy": 6.986404323577881,
      "epoch": 0.07294561333862644,
      "grad_norm": 1.09375,
      "learning_rate": 0.000367,
      "loss": 6.6883,
      "mean_token_accuracy": 0.09569443389773369,
      "num_tokens": 1595259.0,
      "step": 735
    },
    {
      "entropy": 7.101809167861939,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003695,
      "loss": 6.715,
      "mean_token_accuracy": 0.09670865535736084,
      "num_tokens": 1605585.0,
      "step": 740
    },
    {
      "entropy": 7.043530368804932,
      "epoch": 0.07393807066296149,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000372,
      "loss": 6.6582,
      "mean_token_accuracy": 0.09511561095714569,
      "num_tokens": 1616529.0,
      "step": 745
    },
    {
      "entropy": 7.013380241394043,
      "epoch": 0.07443429932512902,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003745,
      "loss": 6.6465,
      "mean_token_accuracy": 0.09446082040667533,
      "num_tokens": 1626223.0,
      "step": 750
    },
    {
      "entropy": 7.02883677482605,
      "epoch": 0.07493052798729655,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000377,
      "loss": 6.7589,
      "mean_token_accuracy": 0.09522427693009376,
      "num_tokens": 1638173.0,
      "step": 755
    },
    {
      "entropy": 6.903225469589233,
      "epoch": 0.07542675664946408,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003795,
      "loss": 6.5802,
      "mean_token_accuracy": 0.0996096670627594,
      "num_tokens": 1648919.0,
      "step": 760
    },
    {
      "entropy": 7.012854671478271,
      "epoch": 0.0759229853116316,
      "grad_norm": 1.0625,
      "learning_rate": 0.000382,
      "loss": 6.821,
      "mean_token_accuracy": 0.08750703632831573,
      "num_tokens": 1660704.0,
      "step": 765
    },
    {
      "entropy": 7.085955572128296,
      "epoch": 0.07641921397379912,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003845,
      "loss": 6.7273,
      "mean_token_accuracy": 0.08895046189427376,
      "num_tokens": 1672080.0,
      "step": 770
    },
    {
      "entropy": 7.075372934341431,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.7014,
      "mean_token_accuracy": 0.09691619500517845,
      "num_tokens": 1681930.0,
      "step": 775
    },
    {
      "entropy": 6.990365028381348,
      "epoch": 0.07741167129813419,
      "grad_norm": 0.890625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.7129,
      "mean_token_accuracy": 0.095877256244421,
      "num_tokens": 1693075.0,
      "step": 780
    },
    {
      "entropy": 7.033229637145996,
      "epoch": 0.0779078999603017,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.6975,
      "mean_token_accuracy": 0.09433050006628037,
      "num_tokens": 1704163.0,
      "step": 785
    },
    {
      "entropy": 7.068866348266601,
      "epoch": 0.07840412862246923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.8117,
      "mean_token_accuracy": 0.09380641803145409,
      "num_tokens": 1714879.0,
      "step": 790
    },
    {
      "entropy": 7.057501363754272,
      "epoch": 0.07890035728463676,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.8275,
      "mean_token_accuracy": 0.09104160815477372,
      "num_tokens": 1725464.0,
      "step": 795
    },
    {
      "entropy": 7.169777536392212,
      "epoch": 0.0793965859468043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003995,
      "loss": 6.7748,
      "mean_token_accuracy": 0.09609344974160194,
      "num_tokens": 1736947.0,
      "step": 800
    },
    {
      "entropy": 6.911767673492432,
      "epoch": 0.07989281460897181,
      "grad_norm": 1.125,
      "learning_rate": 0.000402,
      "loss": 6.5898,
      "mean_token_accuracy": 0.10534465983510018,
      "num_tokens": 1746827.0,
      "step": 805
    },
    {
      "entropy": 7.104987621307373,
      "epoch": 0.08038904327113934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004045,
      "loss": 6.7954,
      "mean_token_accuracy": 0.09412779062986373,
      "num_tokens": 1757079.0,
      "step": 810
    },
    {
      "entropy": 6.993435382843018,
      "epoch": 0.08088527193330687,
      "grad_norm": 0.875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.7236,
      "mean_token_accuracy": 0.09052609503269196,
      "num_tokens": 1767684.0,
      "step": 815
    },
    {
      "entropy": 7.004999494552612,
      "epoch": 0.0813815005954744,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004095,
      "loss": 6.7172,
      "mean_token_accuracy": 0.0948351763188839,
      "num_tokens": 1778187.0,
      "step": 820
    },
    {
      "entropy": 6.968943452835083,
      "epoch": 0.08187772925764192,
      "grad_norm": 1.015625,
      "learning_rate": 0.000412,
      "loss": 6.6845,
      "mean_token_accuracy": 0.09864413291215897,
      "num_tokens": 1789139.0,
      "step": 825
    },
    {
      "entropy": 6.927632141113281,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004145,
      "loss": 6.5456,
      "mean_token_accuracy": 0.10136523321270943,
      "num_tokens": 1799328.0,
      "step": 830
    },
    {
      "entropy": 6.912218379974365,
      "epoch": 0.08287018658197698,
      "grad_norm": 0.890625,
      "learning_rate": 0.000417,
      "loss": 6.7194,
      "mean_token_accuracy": 0.09523818269371986,
      "num_tokens": 1811487.0,
      "step": 835
    },
    {
      "entropy": 6.958463907241821,
      "epoch": 0.0833664152441445,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004195,
      "loss": 6.6047,
      "mean_token_accuracy": 0.10164962559938431,
      "num_tokens": 1821923.0,
      "step": 840
    },
    {
      "entropy": 6.836940288543701,
      "epoch": 0.08386264390631203,
      "grad_norm": 0.984375,
      "learning_rate": 0.000422,
      "loss": 6.6066,
      "mean_token_accuracy": 0.10125545933842658,
      "num_tokens": 1833001.0,
      "step": 845
    },
    {
      "entropy": 6.989613962173462,
      "epoch": 0.08435887256847956,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004245,
      "loss": 6.6295,
      "mean_token_accuracy": 0.10127801448106766,
      "num_tokens": 1844473.0,
      "step": 850
    },
    {
      "entropy": 6.892078447341919,
      "epoch": 0.08485510123064709,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000427,
      "loss": 6.5524,
      "mean_token_accuracy": 0.10098758861422538,
      "num_tokens": 1855851.0,
      "step": 855
    },
    {
      "entropy": 6.790992307662964,
      "epoch": 0.0853513298928146,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004295,
      "loss": 6.5297,
      "mean_token_accuracy": 0.09879347607493401,
      "num_tokens": 1866712.0,
      "step": 860
    },
    {
      "entropy": 7.051131010055542,
      "epoch": 0.08584755855498213,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000432,
      "loss": 6.5771,
      "mean_token_accuracy": 0.09874777272343635,
      "num_tokens": 1876443.0,
      "step": 865
    },
    {
      "entropy": 6.778781223297119,
      "epoch": 0.08634378721714966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004345,
      "loss": 6.6377,
      "mean_token_accuracy": 0.10053054243326187,
      "num_tokens": 1886239.0,
      "step": 870
    },
    {
      "entropy": 6.9395850658416744,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000437,
      "loss": 6.5617,
      "mean_token_accuracy": 0.10510002821683884,
      "num_tokens": 1896457.0,
      "step": 875
    },
    {
      "entropy": 6.825030755996704,
      "epoch": 0.08733624454148471,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004395,
      "loss": 6.563,
      "mean_token_accuracy": 0.09209038838744163,
      "num_tokens": 1908068.0,
      "step": 880
    },
    {
      "entropy": 6.925831365585327,
      "epoch": 0.08783247320365224,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000442,
      "loss": 6.7105,
      "mean_token_accuracy": 0.09559524208307266,
      "num_tokens": 1920093.0,
      "step": 885
    },
    {
      "entropy": 6.803714609146118,
      "epoch": 0.08832870186581977,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004445,
      "loss": 6.5027,
      "mean_token_accuracy": 0.10239366963505744,
      "num_tokens": 1930401.0,
      "step": 890
    },
    {
      "entropy": 6.858418369293213,
      "epoch": 0.0888249305279873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000447,
      "loss": 6.6809,
      "mean_token_accuracy": 0.09742549583315849,
      "num_tokens": 1940163.0,
      "step": 895
    },
    {
      "entropy": 6.918571805953979,
      "epoch": 0.08932115919015482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.5271,
      "mean_token_accuracy": 0.09928687438368797,
      "num_tokens": 1952455.0,
      "step": 900
    },
    {
      "entropy": 6.713730764389038,
      "epoch": 0.08981738785232235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.6428,
      "mean_token_accuracy": 0.09646887555718422,
      "num_tokens": 1965079.0,
      "step": 905
    },
    {
      "entropy": 6.794367742538452,
      "epoch": 0.09031361651448988,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5174,
      "mean_token_accuracy": 0.10413466691970825,
      "num_tokens": 1976620.0,
      "step": 910
    },
    {
      "entropy": 6.848107433319091,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.5142,
      "mean_token_accuracy": 0.0995387688279152,
      "num_tokens": 1987626.0,
      "step": 915
    },
    {
      "entropy": 6.743514633178711,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.533,
      "mean_token_accuracy": 0.10342048928141594,
      "num_tokens": 1998250.0,
      "step": 920
    },
    {
      "entropy": 6.878083229064941,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000462,
      "loss": 6.5135,
      "mean_token_accuracy": 0.10976848751306534,
      "num_tokens": 2008098.0,
      "step": 925
    },
    {
      "entropy": 6.675069999694824,
      "epoch": 0.09229853116315999,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645,
      "loss": 6.4805,
      "mean_token_accuracy": 0.09911079183220864,
      "num_tokens": 2020125.0,
      "step": 930
    },
    {
      "entropy": 6.776659440994263,
      "epoch": 0.0927947598253275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000467,
      "loss": 6.6197,
      "mean_token_accuracy": 0.10722684860229492,
      "num_tokens": 2030791.0,
      "step": 935
    },
    {
      "entropy": 6.848675394058228,
      "epoch": 0.09329098848749504,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004695,
      "loss": 6.5898,
      "mean_token_accuracy": 0.10096785873174667,
      "num_tokens": 2042358.0,
      "step": 940
    },
    {
      "entropy": 6.729436016082763,
      "epoch": 0.09378721714966257,
      "grad_norm": 0.875,
      "learning_rate": 0.000472,
      "loss": 6.5408,
      "mean_token_accuracy": 0.10336418151855468,
      "num_tokens": 2054549.0,
      "step": 945
    },
    {
      "entropy": 6.836209058761597,
      "epoch": 0.0942834458118301,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745,
      "loss": 6.5251,
      "mean_token_accuracy": 0.10392895936965943,
      "num_tokens": 2064903.0,
      "step": 950
    },
    {
      "entropy": 6.76303129196167,
      "epoch": 0.09477967447399761,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000477,
      "loss": 6.4917,
      "mean_token_accuracy": 0.10068923830986024,
      "num_tokens": 2075795.0,
      "step": 955
    },
    {
      "entropy": 6.75622067451477,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004795,
      "loss": 6.5309,
      "mean_token_accuracy": 0.10283064991235732,
      "num_tokens": 2087249.0,
      "step": 960
    },
    {
      "entropy": 6.796994209289551,
      "epoch": 0.09577213179833267,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000482,
      "loss": 6.4935,
      "mean_token_accuracy": 0.10899099111557006,
      "num_tokens": 2097527.0,
      "step": 965
    },
    {
      "entropy": 6.761102533340454,
      "epoch": 0.0962683604605002,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004845,
      "loss": 6.5146,
      "mean_token_accuracy": 0.10290635898709297,
      "num_tokens": 2108174.0,
      "step": 970
    },
    {
      "entropy": 6.649438142776489,
      "epoch": 0.09676458912266772,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000487,
      "loss": 6.5184,
      "mean_token_accuracy": 0.100522430986166,
      "num_tokens": 2120355.0,
      "step": 975
    },
    {
      "entropy": 6.7309898853302,
      "epoch": 0.09726081778483525,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004895,
      "loss": 6.3974,
      "mean_token_accuracy": 0.10792523175477982,
      "num_tokens": 2130861.0,
      "step": 980
    },
    {
      "entropy": 6.701409101486206,
      "epoch": 0.09775704644700278,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000492,
      "loss": 6.561,
      "mean_token_accuracy": 0.10112565234303475,
      "num_tokens": 2142162.0,
      "step": 985
    },
    {
      "entropy": 6.733495473861694,
      "epoch": 0.0982532751091703,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945,
      "loss": 6.4792,
      "mean_token_accuracy": 0.10216469764709472,
      "num_tokens": 2153374.0,
      "step": 990
    },
    {
      "entropy": 6.732190895080566,
      "epoch": 0.09874950377133783,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497,
      "loss": 6.4011,
      "mean_token_accuracy": 0.10262768343091011,
      "num_tokens": 2162941.0,
      "step": 995
    },
    {
      "entropy": 6.757260656356811,
      "epoch": 0.09924573243350536,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995,
      "loss": 6.5,
      "mean_token_accuracy": 0.09760607853531837,
      "num_tokens": 2172864.0,
      "step": 1000
    },
    {
      "entropy": 6.697972249984741,
      "epoch": 0.09974196109567289,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.5121,
      "mean_token_accuracy": 0.10802309364080429,
      "num_tokens": 2184219.0,
      "step": 1005
    },
    {
      "entropy": 6.768112850189209,
      "epoch": 0.1002381897578404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.4823,
      "mean_token_accuracy": 0.09585310667753219,
      "num_tokens": 2194788.0,
      "step": 1010
    },
    {
      "entropy": 6.702127504348755,
      "epoch": 0.10073441842000794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.4988,
      "mean_token_accuracy": 0.09752344265580178,
      "num_tokens": 2205493.0,
      "step": 1015
    },
    {
      "entropy": 6.653623962402344,
      "epoch": 0.10123064708217547,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.4934,
      "mean_token_accuracy": 0.10524687170982361,
      "num_tokens": 2215354.0,
      "step": 1020
    },
    {
      "entropy": 6.642564010620117,
      "epoch": 0.101726875744343,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.4358,
      "mean_token_accuracy": 0.1029872551560402,
      "num_tokens": 2225781.0,
      "step": 1025
    },
    {
      "entropy": 6.660511827468872,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.5018,
      "mean_token_accuracy": 0.10526410788297653,
      "num_tokens": 2236008.0,
      "step": 1030
    },
    {
      "entropy": 6.666350555419922,
      "epoch": 0.10271933306867805,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.4144,
      "mean_token_accuracy": 0.10993591621518135,
      "num_tokens": 2246998.0,
      "step": 1035
    },
    {
      "entropy": 6.685410451889038,
      "epoch": 0.10321556173084558,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.537,
      "mean_token_accuracy": 0.09969183057546616,
      "num_tokens": 2257451.0,
      "step": 1040
    },
    {
      "entropy": 6.716910266876221,
      "epoch": 0.1037117903930131,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.515,
      "mean_token_accuracy": 0.10016707330942154,
      "num_tokens": 2268151.0,
      "step": 1045
    },
    {
      "entropy": 6.784289979934693,
      "epoch": 0.10420801905518062,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.5271,
      "mean_token_accuracy": 0.09749956205487251,
      "num_tokens": 2278924.0,
      "step": 1050
    },
    {
      "entropy": 6.590696668624878,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.4375,
      "mean_token_accuracy": 0.10931600704789161,
      "num_tokens": 2289229.0,
      "step": 1055
    },
    {
      "entropy": 6.65490140914917,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499999611554265,
      "loss": 6.4772,
      "mean_token_accuracy": 0.0997220940887928,
      "num_tokens": 2300203.0,
      "step": 1060
    },
    {
      "entropy": 6.689665460586548,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.4325,
      "mean_token_accuracy": 0.10597841739654541,
      "num_tokens": 2311991.0,
      "step": 1065
    },
    {
      "entropy": 6.5814300060272215,
      "epoch": 0.10619293370385073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.4177,
      "mean_token_accuracy": 0.10108351483941078,
      "num_tokens": 2322110.0,
      "step": 1070
    },
    {
      "entropy": 6.7269233703613285,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.4664,
      "mean_token_accuracy": 0.10983085185289383,
      "num_tokens": 2332439.0,
      "step": 1075
    },
    {
      "entropy": 6.517318058013916,
      "epoch": 0.10718539102818579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.3221,
      "mean_token_accuracy": 0.11178816556930542,
      "num_tokens": 2342339.0,
      "step": 1080
    },
    {
      "entropy": 6.633081340789795,
      "epoch": 0.10768161969035331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.4596,
      "mean_token_accuracy": 0.10636994913220406,
      "num_tokens": 2353211.0,
      "step": 1085
    },
    {
      "entropy": 6.709958791732788,
      "epoch": 0.10817784835252084,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499999116093784,
      "loss": 6.4886,
      "mean_token_accuracy": 0.10582839027047157,
      "num_tokens": 2364709.0,
      "step": 1090
    },
    {
      "entropy": 6.638761806488037,
      "epoch": 0.10867407701468837,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.4311,
      "mean_token_accuracy": 0.10091341882944108,
      "num_tokens": 2375095.0,
      "step": 1095
    },
    {
      "entropy": 6.641486692428589,
      "epoch": 0.1091703056768559,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.4043,
      "mean_token_accuracy": 0.10428427532315254,
      "num_tokens": 2384206.0,
      "step": 1100
    },
    {
      "entropy": 6.592662334442139,
      "epoch": 0.10966653433902342,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.3251,
      "mean_token_accuracy": 0.11068035364151001,
      "num_tokens": 2395139.0,
      "step": 1105
    },
    {
      "entropy": 6.577283239364624,
      "epoch": 0.11016276300119095,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.4501,
      "mean_token_accuracy": 0.09892478436231614,
      "num_tokens": 2405652.0,
      "step": 1110
    },
    {
      "entropy": 6.707149410247803,
      "epoch": 0.11065899166335848,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.4528,
      "mean_token_accuracy": 0.10405765473842621,
      "num_tokens": 2416386.0,
      "step": 1115
    },
    {
      "entropy": 6.549600219726562,
      "epoch": 0.11115522032552601,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.471,
      "mean_token_accuracy": 0.10263076052069664,
      "num_tokens": 2427807.0,
      "step": 1120
    },
    {
      "entropy": 6.584462308883667,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.3742,
      "mean_token_accuracy": 0.1078455038368702,
      "num_tokens": 2439481.0,
      "step": 1125
    },
    {
      "entropy": 6.644891214370728,
      "epoch": 0.11214767764986106,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.422,
      "mean_token_accuracy": 0.10707498490810394,
      "num_tokens": 2449857.0,
      "step": 1130
    },
    {
      "entropy": 6.55607008934021,
      "epoch": 0.11264390631202859,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.3453,
      "mean_token_accuracy": 0.10785013660788537,
      "num_tokens": 2461569.0,
      "step": 1135
    },
    {
      "entropy": 6.635892677307129,
      "epoch": 0.11314013497419612,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.4311,
      "mean_token_accuracy": 0.10901110246777534,
      "num_tokens": 2473255.0,
      "step": 1140
    },
    {
      "entropy": 6.633499383926392,
      "epoch": 0.11363636363636363,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.3833,
      "mean_token_accuracy": 0.1040490709245205,
      "num_tokens": 2484473.0,
      "step": 1145
    },
    {
      "entropy": 6.566697025299073,
      "epoch": 0.11413259229853116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.4046,
      "mean_token_accuracy": 0.10690351948142052,
      "num_tokens": 2495652.0,
      "step": 1150
    },
    {
      "entropy": 6.65309100151062,
      "epoch": 0.1146288209606987,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499997353529548,
      "loss": 6.4341,
      "mean_token_accuracy": 0.10881170257925987,
      "num_tokens": 2507137.0,
      "step": 1155
    },
    {
      "entropy": 6.535683012008667,
      "epoch": 0.11512504962286621,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.3607,
      "mean_token_accuracy": 0.10740326344966888,
      "num_tokens": 2517896.0,
      "step": 1160
    },
    {
      "entropy": 6.523780012130738,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.3629,
      "mean_token_accuracy": 0.10846158191561699,
      "num_tokens": 2529976.0,
      "step": 1165
    },
    {
      "entropy": 6.653525257110596,
      "epoch": 0.11611750694720127,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.4176,
      "mean_token_accuracy": 0.10163490399718285,
      "num_tokens": 2540872.0,
      "step": 1170
    },
    {
      "entropy": 6.6223382472991945,
      "epoch": 0.1166137356093688,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.4314,
      "mean_token_accuracy": 0.1041704148054123,
      "num_tokens": 2551178.0,
      "step": 1175
    },
    {
      "entropy": 6.540678930282593,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.3356,
      "mean_token_accuracy": 0.10806481763720513,
      "num_tokens": 2561472.0,
      "step": 1180
    },
    {
      "entropy": 6.559564065933228,
      "epoch": 0.11760619293370385,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.3357,
      "mean_token_accuracy": 0.10722169801592826,
      "num_tokens": 2572717.0,
      "step": 1185
    },
    {
      "entropy": 6.526041793823242,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.2463,
      "mean_token_accuracy": 0.11155131533741951,
      "num_tokens": 2582991.0,
      "step": 1190
    },
    {
      "entropy": 6.437657451629638,
      "epoch": 0.11859865025803891,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.3384,
      "mean_token_accuracy": 0.11157690584659577,
      "num_tokens": 2594872.0,
      "step": 1195
    },
    {
      "entropy": 6.519007015228271,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.3501,
      "mean_token_accuracy": 0.11119264140725135,
      "num_tokens": 2605991.0,
      "step": 1200
    },
    {
      "entropy": 6.5526268005371096,
      "epoch": 0.11959110758237396,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.3742,
      "mean_token_accuracy": 0.10888314321637153,
      "num_tokens": 2617613.0,
      "step": 1205
    },
    {
      "entropy": 6.564966106414795,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.4382,
      "mean_token_accuracy": 0.10179812833666801,
      "num_tokens": 2627724.0,
      "step": 1210
    },
    {
      "entropy": 6.583070945739746,
      "epoch": 0.12058356490670902,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499994889629802,
      "loss": 6.3525,
      "mean_token_accuracy": 0.10510100647807122,
      "num_tokens": 2638944.0,
      "step": 1215
    },
    {
      "entropy": 6.452510929107666,
      "epoch": 0.12107979356887653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499994648038667,
      "loss": 6.2019,
      "mean_token_accuracy": 0.1171508140861988,
      "num_tokens": 2649132.0,
      "step": 1220
    },
    {
      "entropy": 6.549446058273316,
      "epoch": 0.12157602223104406,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.3345,
      "mean_token_accuracy": 0.10771893933415413,
      "num_tokens": 2660482.0,
      "step": 1225
    },
    {
      "entropy": 6.59168848991394,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.4071,
      "mean_token_accuracy": 0.10588582754135131,
      "num_tokens": 2672953.0,
      "step": 1230
    },
    {
      "entropy": 6.458992004394531,
      "epoch": 0.12256847955537911,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.3076,
      "mean_token_accuracy": 0.11501472815871239,
      "num_tokens": 2684958.0,
      "step": 1235
    },
    {
      "entropy": 6.5847851753234865,
      "epoch": 0.12306470821754664,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.3103,
      "mean_token_accuracy": 0.10793442577123642,
      "num_tokens": 2695375.0,
      "step": 1240
    },
    {
      "entropy": 6.559399366378784,
      "epoch": 0.12356093687971417,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.4781,
      "mean_token_accuracy": 0.10172695145010949,
      "num_tokens": 2707762.0,
      "step": 1245
    },
    {
      "entropy": 6.5289863586425785,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.309,
      "mean_token_accuracy": 0.10860389918088913,
      "num_tokens": 2718237.0,
      "step": 1250
    },
    {
      "entropy": 6.479602432250976,
      "epoch": 0.12455339420404922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.278,
      "mean_token_accuracy": 0.11147248297929764,
      "num_tokens": 2728746.0,
      "step": 1255
    },
    {
      "entropy": 6.495798921585083,
      "epoch": 0.12504962286621676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.281,
      "mean_token_accuracy": 0.10627977550029755,
      "num_tokens": 2739030.0,
      "step": 1260
    },
    {
      "entropy": 6.411447954177857,
      "epoch": 0.12554585152838427,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.2412,
      "mean_token_accuracy": 0.11322782710194587,
      "num_tokens": 2750233.0,
      "step": 1265
    },
    {
      "entropy": 6.518495082855225,
      "epoch": 0.1260420801905518,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.3145,
      "mean_token_accuracy": 0.10560797899961472,
      "num_tokens": 2761298.0,
      "step": 1270
    },
    {
      "entropy": 6.510967969894409,
      "epoch": 0.12653830885271933,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.3645,
      "mean_token_accuracy": 0.10636256784200668,
      "num_tokens": 2771493.0,
      "step": 1275
    },
    {
      "entropy": 6.477094602584839,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.271,
      "mean_token_accuracy": 0.10662685185670853,
      "num_tokens": 2782969.0,
      "step": 1280
    },
    {
      "entropy": 6.5462244033813475,
      "epoch": 0.1275307661770544,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.2984,
      "mean_token_accuracy": 0.11486872658133507,
      "num_tokens": 2793355.0,
      "step": 1285
    },
    {
      "entropy": 6.421256589889526,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.875,
      "learning_rate": 0.000499990679930598,
      "loss": 6.2118,
      "mean_token_accuracy": 0.11601667776703835,
      "num_tokens": 2804473.0,
      "step": 1290
    },
    {
      "entropy": 6.47119288444519,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.2858,
      "mean_token_accuracy": 0.10927127152681351,
      "num_tokens": 2814899.0,
      "step": 1295
    },
    {
      "entropy": 6.552784252166748,
      "epoch": 0.12901945216355698,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.3085,
      "mean_token_accuracy": 0.11181130483746529,
      "num_tokens": 2825592.0,
      "step": 1300
    },
    {
      "entropy": 6.395250940322876,
      "epoch": 0.12951568082572448,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.3107,
      "mean_token_accuracy": 0.11031345725059509,
      "num_tokens": 2837444.0,
      "step": 1305
    },
    {
      "entropy": 6.5371500015258786,
      "epoch": 0.130011909487892,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.3052,
      "mean_token_accuracy": 0.10789528265595436,
      "num_tokens": 2849256.0,
      "step": 1310
    },
    {
      "entropy": 6.516609334945679,
      "epoch": 0.13050813815005954,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.3508,
      "mean_token_accuracy": 0.10912716761231422,
      "num_tokens": 2860606.0,
      "step": 1315
    },
    {
      "entropy": 6.436788082122803,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.2797,
      "mean_token_accuracy": 0.11875139251351356,
      "num_tokens": 2871825.0,
      "step": 1320
    },
    {
      "entropy": 6.539361381530762,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499988285797938,
      "loss": 6.3158,
      "mean_token_accuracy": 0.11384041458368302,
      "num_tokens": 2883408.0,
      "step": 1325
    },
    {
      "entropy": 6.459595441818237,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499987921462018,
      "loss": 6.2928,
      "mean_token_accuracy": 0.11346368417143822,
      "num_tokens": 2895786.0,
      "step": 1330
    },
    {
      "entropy": 6.4756751537323,
      "epoch": 0.13249305279872967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.2804,
      "mean_token_accuracy": 0.11255847290158272,
      "num_tokens": 2905273.0,
      "step": 1335
    },
    {
      "entropy": 6.56948299407959,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.3865,
      "mean_token_accuracy": 0.105442725867033,
      "num_tokens": 2917546.0,
      "step": 1340
    },
    {
      "entropy": 6.5599915981292725,
      "epoch": 0.1334855101230647,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.3222,
      "mean_token_accuracy": 0.11331170350313187,
      "num_tokens": 2928715.0,
      "step": 1345
    },
    {
      "entropy": 6.435283613204956,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.2668,
      "mean_token_accuracy": 0.1143470786511898,
      "num_tokens": 2940260.0,
      "step": 1350
    },
    {
      "entropy": 6.487202262878418,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.2601,
      "mean_token_accuracy": 0.11385833024978638,
      "num_tokens": 2951543.0,
      "step": 1355
    },
    {
      "entropy": 6.443022680282593,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.306,
      "mean_token_accuracy": 0.11349669769406319,
      "num_tokens": 2962504.0,
      "step": 1360
    },
    {
      "entropy": 6.441706657409668,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.2202,
      "mean_token_accuracy": 0.11742114797234535,
      "num_tokens": 2973784.0,
      "step": 1365
    },
    {
      "entropy": 6.491312503814697,
      "epoch": 0.13596665343390235,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.3388,
      "mean_token_accuracy": 0.10710015147924423,
      "num_tokens": 2985281.0,
      "step": 1370
    },
    {
      "entropy": 6.395635318756104,
      "epoch": 0.13646288209606988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.1979,
      "mean_token_accuracy": 0.11877716705203056,
      "num_tokens": 2996090.0,
      "step": 1375
    },
    {
      "entropy": 6.510016632080078,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.2469,
      "mean_token_accuracy": 0.11208310946822167,
      "num_tokens": 3006235.0,
      "step": 1380
    },
    {
      "entropy": 6.3472301959991455,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.1965,
      "mean_token_accuracy": 0.11313682422041893,
      "num_tokens": 3016481.0,
      "step": 1385
    },
    {
      "entropy": 6.527986335754394,
      "epoch": 0.13795156808257245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.2562,
      "mean_token_accuracy": 0.11551145687699318,
      "num_tokens": 3026382.0,
      "step": 1390
    },
    {
      "entropy": 6.418882274627686,
      "epoch": 0.13844779674473998,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.2683,
      "mean_token_accuracy": 0.10792074427008629,
      "num_tokens": 3036637.0,
      "step": 1395
    },
    {
      "entropy": 6.469444751739502,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.3248,
      "mean_token_accuracy": 0.10550551563501358,
      "num_tokens": 3047115.0,
      "step": 1400
    },
    {
      "entropy": 6.574054908752442,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.3377,
      "mean_token_accuracy": 0.11256245598196983,
      "num_tokens": 3057499.0,
      "step": 1405
    },
    {
      "entropy": 6.4168126583099365,
      "epoch": 0.13993648273124257,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.2538,
      "mean_token_accuracy": 0.11392025426030158,
      "num_tokens": 3068503.0,
      "step": 1410
    },
    {
      "entropy": 6.42028923034668,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.2314,
      "mean_token_accuracy": 0.11181280761957169,
      "num_tokens": 3079220.0,
      "step": 1415
    },
    {
      "entropy": 6.359934759140015,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.2197,
      "mean_token_accuracy": 0.11571786925196648,
      "num_tokens": 3091621.0,
      "step": 1420
    },
    {
      "entropy": 6.446488618850708,
      "epoch": 0.14142516871774513,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499979939039694,
      "loss": 6.2005,
      "mean_token_accuracy": 0.11424326300621032,
      "num_tokens": 3101378.0,
      "step": 1425
    },
    {
      "entropy": 6.348664093017578,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.2166,
      "mean_token_accuracy": 0.11557507440447808,
      "num_tokens": 3112996.0,
      "step": 1430
    },
    {
      "entropy": 6.414705753326416,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.2043,
      "mean_token_accuracy": 0.10700563341379166,
      "num_tokens": 3124262.0,
      "step": 1435
    },
    {
      "entropy": 6.345302104949951,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.2497,
      "mean_token_accuracy": 0.10843175053596496,
      "num_tokens": 3135785.0,
      "step": 1440
    },
    {
      "entropy": 6.416088008880616,
      "epoch": 0.14341008336641525,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.0994,
      "mean_token_accuracy": 0.12602320015430452,
      "num_tokens": 3146841.0,
      "step": 1445
    },
    {
      "entropy": 6.290076351165771,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.2572,
      "mean_token_accuracy": 0.10866076126694679,
      "num_tokens": 3158931.0,
      "step": 1450
    },
    {
      "entropy": 6.446962547302246,
      "epoch": 0.1444025406907503,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.1983,
      "mean_token_accuracy": 0.11449775323271752,
      "num_tokens": 3169320.0,
      "step": 1455
    },
    {
      "entropy": 6.4680766582489015,
      "epoch": 0.14489876935291782,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.2332,
      "mean_token_accuracy": 0.11214599311351776,
      "num_tokens": 3179163.0,
      "step": 1460
    },
    {
      "entropy": 6.34765396118164,
      "epoch": 0.14539499801508535,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.1782,
      "mean_token_accuracy": 0.11042089611291886,
      "num_tokens": 3187987.0,
      "step": 1465
    },
    {
      "entropy": 6.372549295425415,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.1683,
      "mean_token_accuracy": 0.11804649159312249,
      "num_tokens": 3198052.0,
      "step": 1470
    },
    {
      "entropy": 6.2816180229187015,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499974928802568,
      "loss": 6.1512,
      "mean_token_accuracy": 0.11864635944366456,
      "num_tokens": 3209019.0,
      "step": 1475
    },
    {
      "entropy": 6.298905229568481,
      "epoch": 0.14688368400158794,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.1034,
      "mean_token_accuracy": 0.12014764696359634,
      "num_tokens": 3219130.0,
      "step": 1480
    },
    {
      "entropy": 6.288952350616455,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499973859808013,
      "loss": 6.1357,
      "mean_token_accuracy": 0.11747602671384812,
      "num_tokens": 3230882.0,
      "step": 1485
    },
    {
      "entropy": 6.3970338821411135,
      "epoch": 0.14787614132592297,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.2601,
      "mean_token_accuracy": 0.11354640573263168,
      "num_tokens": 3241752.0,
      "step": 1490
    },
    {
      "entropy": 6.496769285202026,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.2828,
      "mean_token_accuracy": 0.1073771022260189,
      "num_tokens": 3252291.0,
      "step": 1495
    },
    {
      "entropy": 6.41784029006958,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.1967,
      "mean_token_accuracy": 0.12008021101355552,
      "num_tokens": 3263525.0,
      "step": 1500
    },
    {
      "entropy": 6.366907787322998,
      "epoch": 0.14936482731242556,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.1538,
      "mean_token_accuracy": 0.11907590329647064,
      "num_tokens": 3274164.0,
      "step": 1505
    },
    {
      "entropy": 6.2786328315734865,
      "epoch": 0.1498610559745931,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999710896916921,
      "loss": 6.1281,
      "mean_token_accuracy": 0.120063366740942,
      "num_tokens": 3286244.0,
      "step": 1510
    },
    {
      "entropy": 6.408924198150634,
      "epoch": 0.15035728463676062,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.2339,
      "mean_token_accuracy": 0.10920965149998665,
      "num_tokens": 3296569.0,
      "step": 1515
    },
    {
      "entropy": 6.31035418510437,
      "epoch": 0.15085351329892815,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.2497,
      "mean_token_accuracy": 0.11545195132493973,
      "num_tokens": 3307259.0,
      "step": 1520
    },
    {
      "entropy": 6.3992188453674315,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999693606762103,
      "loss": 6.0531,
      "mean_token_accuracy": 0.12170021310448646,
      "num_tokens": 3317473.0,
      "step": 1525
    },
    {
      "entropy": 6.274422454833984,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.1779,
      "mean_token_accuracy": 0.11477752849459648,
      "num_tokens": 3328783.0,
      "step": 1530
    },
    {
      "entropy": 6.309112119674682,
      "epoch": 0.15234219928543072,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.2142,
      "mean_token_accuracy": 0.11577230393886566,
      "num_tokens": 3338308.0,
      "step": 1535
    },
    {
      "entropy": 6.423832416534424,
      "epoch": 0.15283842794759825,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499967581451931,
      "loss": 6.1213,
      "mean_token_accuracy": 0.11633878573775291,
      "num_tokens": 3348802.0,
      "step": 1540
    },
    {
      "entropy": 6.263744926452636,
      "epoch": 0.15333465660976578,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499966977219729,
      "loss": 6.1365,
      "mean_token_accuracy": 0.11869108527898789,
      "num_tokens": 3358781.0,
      "step": 1545
    },
    {
      "entropy": 6.3492378234863285,
      "epoch": 0.1538308852719333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.0753,
      "mean_token_accuracy": 0.12006449326872826,
      "num_tokens": 3369360.0,
      "step": 1550
    },
    {
      "entropy": 6.379684591293335,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1884,
      "mean_token_accuracy": 0.11728774011135101,
      "num_tokens": 3380111.0,
      "step": 1555
    },
    {
      "entropy": 6.255392026901245,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.1373,
      "mean_token_accuracy": 0.11838044002652168,
      "num_tokens": 3391932.0,
      "step": 1560
    },
    {
      "entropy": 6.3688195705413815,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.1855,
      "mean_token_accuracy": 0.11579569801688194,
      "num_tokens": 3402192.0,
      "step": 1565
    },
    {
      "entropy": 6.375439691543579,
      "epoch": 0.1558157999206034,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999638723785791,
      "loss": 6.0869,
      "mean_token_accuracy": 0.1146264374256134,
      "num_tokens": 3412382.0,
      "step": 1570
    },
    {
      "entropy": 6.255414152145386,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499963234674429,
      "loss": 6.1256,
      "mean_token_accuracy": 0.11626426354050637,
      "num_tokens": 3422471.0,
      "step": 1575
    },
    {
      "entropy": 6.325457906723022,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.1156,
      "mean_token_accuracy": 0.12189654409885406,
      "num_tokens": 3432348.0,
      "step": 1580
    },
    {
      "entropy": 6.290401220321655,
      "epoch": 0.157304485907106,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.1179,
      "mean_token_accuracy": 0.1161232702434063,
      "num_tokens": 3442549.0,
      "step": 1585
    },
    {
      "entropy": 6.2806614398956295,
      "epoch": 0.15780071456927353,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.1954,
      "mean_token_accuracy": 0.11219574809074402,
      "num_tokens": 3452317.0,
      "step": 1590
    },
    {
      "entropy": 6.3340983390808105,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.1293,
      "mean_token_accuracy": 0.1172614224255085,
      "num_tokens": 3463432.0,
      "step": 1595
    },
    {
      "entropy": 6.395657205581665,
      "epoch": 0.1587931718936086,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1889,
      "mean_token_accuracy": 0.11410781070590019,
      "num_tokens": 3473800.0,
      "step": 1600
    },
    {
      "entropy": 6.311441802978516,
      "epoch": 0.1592894005557761,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.171,
      "mean_token_accuracy": 0.1144875705242157,
      "num_tokens": 3484803.0,
      "step": 1605
    },
    {
      "entropy": 6.3286027908325195,
      "epoch": 0.15978562921794362,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.1047,
      "mean_token_accuracy": 0.1186643548309803,
      "num_tokens": 3494687.0,
      "step": 1610
    },
    {
      "entropy": 6.32429256439209,
      "epoch": 0.16028185788011115,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999579322128608,
      "loss": 6.1522,
      "mean_token_accuracy": 0.11586391776800156,
      "num_tokens": 3505586.0,
      "step": 1615
    },
    {
      "entropy": 6.186865186691284,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999572443018654,
      "loss": 5.9894,
      "mean_token_accuracy": 0.13321231454610824,
      "num_tokens": 3516702.0,
      "step": 1620
    },
    {
      "entropy": 6.295269012451172,
      "epoch": 0.1612743152044462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.1263,
      "mean_token_accuracy": 0.11919939890503883,
      "num_tokens": 3527240.0,
      "step": 1625
    },
    {
      "entropy": 6.314893007278442,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499955851744527,
      "loss": 6.2405,
      "mean_token_accuracy": 0.11684711053967475,
      "num_tokens": 3539368.0,
      "step": 1630
    },
    {
      "entropy": 6.3168638229370115,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.1061,
      "mean_token_accuracy": 0.11994773149490356,
      "num_tokens": 3551128.0,
      "step": 1635
    },
    {
      "entropy": 6.26086835861206,
      "epoch": 0.1627630011909488,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.1133,
      "mean_token_accuracy": 0.11799252703785897,
      "num_tokens": 3562385.0,
      "step": 1640
    },
    {
      "entropy": 6.35860800743103,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.107,
      "mean_token_accuracy": 0.11943186968564987,
      "num_tokens": 3572737.0,
      "step": 1645
    },
    {
      "entropy": 6.24028697013855,
      "epoch": 0.16375545851528384,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.1151,
      "mean_token_accuracy": 0.11283019334077835,
      "num_tokens": 3583872.0,
      "step": 1650
    },
    {
      "entropy": 6.313495445251465,
      "epoch": 0.16425168717745137,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.1013,
      "mean_token_accuracy": 0.11968362778425216,
      "num_tokens": 3595012.0,
      "step": 1655
    },
    {
      "entropy": 6.366147470474243,
      "epoch": 0.1647479158396189,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499951540190987,
      "loss": 6.209,
      "mean_token_accuracy": 0.11140498891472816,
      "num_tokens": 3605733.0,
      "step": 1660
    },
    {
      "entropy": 6.24420485496521,
      "epoch": 0.16524414450178643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.1366,
      "mean_token_accuracy": 0.1271058052778244,
      "num_tokens": 3615997.0,
      "step": 1665
    },
    {
      "entropy": 6.272862815856934,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.1016,
      "mean_token_accuracy": 0.12369741126894951,
      "num_tokens": 3628742.0,
      "step": 1670
    },
    {
      "entropy": 6.227112102508545,
      "epoch": 0.1662366018261215,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.1043,
      "mean_token_accuracy": 0.12184740528464318,
      "num_tokens": 3639702.0,
      "step": 1675
    },
    {
      "entropy": 6.369050741195679,
      "epoch": 0.166732830488289,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.0985,
      "mean_token_accuracy": 0.11733277067542076,
      "num_tokens": 3650414.0,
      "step": 1680
    },
    {
      "entropy": 6.190679216384888,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999477938259256,
      "loss": 6.0631,
      "mean_token_accuracy": 0.11306010484695435,
      "num_tokens": 3660739.0,
      "step": 1685
    },
    {
      "entropy": 6.306669664382935,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.1382,
      "mean_token_accuracy": 0.12038198262453079,
      "num_tokens": 3673109.0,
      "step": 1690
    },
    {
      "entropy": 6.293005657196045,
      "epoch": 0.16822151647479158,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.1175,
      "mean_token_accuracy": 0.1181846633553505,
      "num_tokens": 3683168.0,
      "step": 1695
    },
    {
      "entropy": 6.264427709579468,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499945479068093,
      "loss": 6.1092,
      "mean_token_accuracy": 0.11897704154253005,
      "num_tokens": 3694309.0,
      "step": 1700
    },
    {
      "entropy": 6.261802101135254,
      "epoch": 0.16921397379912664,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.1845,
      "mean_token_accuracy": 0.11810654923319816,
      "num_tokens": 3706165.0,
      "step": 1705
    },
    {
      "entropy": 6.351796007156372,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999439080053571,
      "loss": 6.1929,
      "mean_token_accuracy": 0.11733902394771575,
      "num_tokens": 3717818.0,
      "step": 1710
    },
    {
      "entropy": 6.290229082107544,
      "epoch": 0.1702064311234617,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999431141067928,
      "loss": 6.0696,
      "mean_token_accuracy": 0.11890798285603524,
      "num_tokens": 3727837.0,
      "step": 1715
    },
    {
      "entropy": 6.205497550964355,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999423146301242,
      "loss": 6.0576,
      "mean_token_accuracy": 0.11845887154340744,
      "num_tokens": 3738827.0,
      "step": 1720
    },
    {
      "entropy": 6.330425024032593,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999415095753709,
      "loss": 6.0668,
      "mean_token_accuracy": 0.11846604123711586,
      "num_tokens": 3749859.0,
      "step": 1725
    },
    {
      "entropy": 6.18771653175354,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999406989425527,
      "loss": 6.0777,
      "mean_token_accuracy": 0.11917969733476638,
      "num_tokens": 3760999.0,
      "step": 1730
    },
    {
      "entropy": 6.328619718551636,
      "epoch": 0.1721913457721318,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.1057,
      "mean_token_accuracy": 0.11780516654253007,
      "num_tokens": 3771541.0,
      "step": 1735
    },
    {
      "entropy": 6.212802362442017,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499939060942803,
      "loss": 6.0552,
      "mean_token_accuracy": 0.12180277779698372,
      "num_tokens": 3781503.0,
      "step": 1740
    },
    {
      "entropy": 6.240700149536133,
      "epoch": 0.17318380309646686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.127,
      "mean_token_accuracy": 0.11963073760271073,
      "num_tokens": 3791278.0,
      "step": 1745
    },
    {
      "entropy": 6.374361801147461,
      "epoch": 0.1736800317586344,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.1482,
      "mean_token_accuracy": 0.12124232724308967,
      "num_tokens": 3802995.0,
      "step": 1750
    },
    {
      "entropy": 6.22139801979065,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999365621081999,
      "loss": 6.1014,
      "mean_token_accuracy": 0.12158005386590957,
      "num_tokens": 3813912.0,
      "step": 1755
    },
    {
      "entropy": 6.329367971420288,
      "epoch": 0.17467248908296942,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999357180074204,
      "loss": 6.0405,
      "mean_token_accuracy": 0.11887364089488983,
      "num_tokens": 3823513.0,
      "step": 1760
    },
    {
      "entropy": 6.203952169418335,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999348683287199,
      "loss": 6.0567,
      "mean_token_accuracy": 0.12055648267269134,
      "num_tokens": 3835602.0,
      "step": 1765
    },
    {
      "entropy": 6.282654809951782,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999340130721192,
      "loss": 6.1153,
      "mean_token_accuracy": 0.11746885254979134,
      "num_tokens": 3847031.0,
      "step": 1770
    },
    {
      "entropy": 6.1639931201934814,
      "epoch": 0.17616117506947201,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999331522376396,
      "loss": 5.9808,
      "mean_token_accuracy": 0.11802588552236556,
      "num_tokens": 3857714.0,
      "step": 1775
    },
    {
      "entropy": 6.25999174118042,
      "epoch": 0.17665740373163955,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999322858253027,
      "loss": 6.0745,
      "mean_token_accuracy": 0.12271984964609146,
      "num_tokens": 3868818.0,
      "step": 1780
    },
    {
      "entropy": 6.28449068069458,
      "epoch": 0.17715363239380708,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.1107,
      "mean_token_accuracy": 0.12038928046822547,
      "num_tokens": 3878802.0,
      "step": 1785
    },
    {
      "entropy": 6.277543830871582,
      "epoch": 0.1776498610559746,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.1207,
      "mean_token_accuracy": 0.11644105166196823,
      "num_tokens": 3890045.0,
      "step": 1790
    },
    {
      "entropy": 6.266772079467773,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999296531213624,
      "loss": 6.0013,
      "mean_token_accuracy": 0.12227872386574745,
      "num_tokens": 3901429.0,
      "step": 1795
    },
    {
      "entropy": 6.185096502304077,
      "epoch": 0.17864231838030964,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9674,
      "mean_token_accuracy": 0.12167818248271942,
      "num_tokens": 3912201.0,
      "step": 1800
    },
    {
      "entropy": 6.186278486251831,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999278700965124,
      "loss": 6.0153,
      "mean_token_accuracy": 0.12317613437771797,
      "num_tokens": 3922338.0,
      "step": 1805
    },
    {
      "entropy": 6.213421297073364,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999269702174866,
      "loss": 6.0516,
      "mean_token_accuracy": 0.12139523699879647,
      "num_tokens": 3934469.0,
      "step": 1810
    },
    {
      "entropy": 6.172092056274414,
      "epoch": 0.18013100436681223,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999260647607565,
      "loss": 6.0251,
      "mean_token_accuracy": 0.1218095563352108,
      "num_tokens": 3946136.0,
      "step": 1815
    },
    {
      "entropy": 6.243712377548218,
      "epoch": 0.18062723302897976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999251537263448,
      "loss": 6.0815,
      "mean_token_accuracy": 0.12056513875722885,
      "num_tokens": 3956976.0,
      "step": 1820
    },
    {
      "entropy": 6.264207172393799,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.0832,
      "mean_token_accuracy": 0.12240420579910279,
      "num_tokens": 3967929.0,
      "step": 1825
    },
    {
      "entropy": 6.104647302627564,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004999233149245665,
      "loss": 6.03,
      "mean_token_accuracy": 0.127287620306015,
      "num_tokens": 3979585.0,
      "step": 1830
    },
    {
      "entropy": 6.303787088394165,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999223871572456,
      "loss": 6.0066,
      "mean_token_accuracy": 0.12537923753261565,
      "num_tokens": 3990097.0,
      "step": 1835
    },
    {
      "entropy": 6.271418666839599,
      "epoch": 0.18261214767764986,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.0754,
      "mean_token_accuracy": 0.1204076036810875,
      "num_tokens": 3999864.0,
      "step": 1840
    },
    {
      "entropy": 6.181163644790649,
      "epoch": 0.18310837633981739,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.0935,
      "mean_token_accuracy": 0.11834281980991364,
      "num_tokens": 4012986.0,
      "step": 1845
    },
    {
      "entropy": 6.230453109741211,
      "epoch": 0.18360460500198492,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999195703898325,
      "loss": 6.0812,
      "mean_token_accuracy": 0.12549262940883638,
      "num_tokens": 4024231.0,
      "step": 1850
    },
    {
      "entropy": 6.205415105819702,
      "epoch": 0.18410083366415245,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004999186203122889,
      "loss": 6.0739,
      "mean_token_accuracy": 0.11859336569905281,
      "num_tokens": 4035709.0,
      "step": 1855
    },
    {
      "entropy": 6.167673015594483,
      "epoch": 0.18459706232631998,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999176646572482,
      "loss": 5.9531,
      "mean_token_accuracy": 0.12172793224453926,
      "num_tokens": 4046504.0,
      "step": 1860
    },
    {
      "entropy": 6.215722465515137,
      "epoch": 0.1850932909884875,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499916703424734,
      "loss": 6.0105,
      "mean_token_accuracy": 0.11848408207297326,
      "num_tokens": 4056864.0,
      "step": 1865
    },
    {
      "entropy": 6.213238525390625,
      "epoch": 0.185589519650655,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004999157366147702,
      "loss": 6.0176,
      "mean_token_accuracy": 0.12398831099271775,
      "num_tokens": 4067682.0,
      "step": 1870
    },
    {
      "entropy": 6.203938627243042,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999147642273809,
      "loss": 6.026,
      "mean_token_accuracy": 0.12088289335370064,
      "num_tokens": 4078623.0,
      "step": 1875
    },
    {
      "entropy": 6.248812532424926,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049991378626259,
      "loss": 6.1217,
      "mean_token_accuracy": 0.1188036285340786,
      "num_tokens": 4090479.0,
      "step": 1880
    },
    {
      "entropy": 6.2731225967407225,
      "epoch": 0.1870782056371576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999128027204219,
      "loss": 6.0342,
      "mean_token_accuracy": 0.12075977995991707,
      "num_tokens": 4100379.0,
      "step": 1885
    },
    {
      "entropy": 6.173857021331787,
      "epoch": 0.18757443429932513,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499911813600901,
      "loss": 6.1629,
      "mean_token_accuracy": 0.11313794627785682,
      "num_tokens": 4111738.0,
      "step": 1890
    },
    {
      "entropy": 6.290290641784668,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.1278,
      "mean_token_accuracy": 0.11461761593818665,
      "num_tokens": 4123602.0,
      "step": 1895
    },
    {
      "entropy": 6.206590223312378,
      "epoch": 0.1885668916236602,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999098186298986,
      "loss": 6.0744,
      "mean_token_accuracy": 0.12139438912272453,
      "num_tokens": 4133558.0,
      "step": 1900
    },
    {
      "entropy": 6.151382112503052,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999088127784667,
      "loss": 5.9884,
      "mean_token_accuracy": 0.12704532966017723,
      "num_tokens": 4143901.0,
      "step": 1905
    },
    {
      "entropy": 6.201519012451172,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499907801349781,
      "loss": 6.034,
      "mean_token_accuracy": 0.11795807406306266,
      "num_tokens": 4154180.0,
      "step": 1910
    },
    {
      "entropy": 6.204054355621338,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999067843438664,
      "loss": 5.9706,
      "mean_token_accuracy": 0.12664248719811438,
      "num_tokens": 4163540.0,
      "step": 1915
    },
    {
      "entropy": 6.17950029373169,
      "epoch": 0.1905518062723303,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999057617607482,
      "loss": 6.0436,
      "mean_token_accuracy": 0.12264975383877755,
      "num_tokens": 4173723.0,
      "step": 1920
    },
    {
      "entropy": 6.263732385635376,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999047336004517,
      "loss": 6.0341,
      "mean_token_accuracy": 0.12559134289622306,
      "num_tokens": 4184771.0,
      "step": 1925
    },
    {
      "entropy": 6.172854852676392,
      "epoch": 0.19154426359666535,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999036998630024,
      "loss": 6.0439,
      "mean_token_accuracy": 0.11994816809892654,
      "num_tokens": 4195791.0,
      "step": 1930
    },
    {
      "entropy": 6.213911008834839,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.7421875,
      "learning_rate": 0.000499902660548426,
      "loss": 6.0617,
      "mean_token_accuracy": 0.11922118812799454,
      "num_tokens": 4207393.0,
      "step": 1935
    },
    {
      "entropy": 6.174426698684693,
      "epoch": 0.1925367209210004,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004999016156567482,
      "loss": 5.9961,
      "mean_token_accuracy": 0.12561587989330292,
      "num_tokens": 4218278.0,
      "step": 1940
    },
    {
      "entropy": 6.258877611160278,
      "epoch": 0.1930329495831679,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499900565187995,
      "loss": 6.1393,
      "mean_token_accuracy": 0.11696879714727401,
      "num_tokens": 4229448.0,
      "step": 1945
    },
    {
      "entropy": 6.302958202362061,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998995091421924,
      "loss": 6.0653,
      "mean_token_accuracy": 0.11950641125440598,
      "num_tokens": 4240750.0,
      "step": 1950
    },
    {
      "entropy": 6.1788359642028805,
      "epoch": 0.19402540690750297,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998984475193665,
      "loss": 6.0601,
      "mean_token_accuracy": 0.12437045946717262,
      "num_tokens": 4252086.0,
      "step": 1955
    },
    {
      "entropy": 6.237712335586548,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998973803195439,
      "loss": 6.0034,
      "mean_token_accuracy": 0.12056957632303238,
      "num_tokens": 4262756.0,
      "step": 1960
    },
    {
      "entropy": 6.126477241516113,
      "epoch": 0.19501786423183803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998963075427508,
      "loss": 6.061,
      "mean_token_accuracy": 0.12428061813116073,
      "num_tokens": 4272965.0,
      "step": 1965
    },
    {
      "entropy": 6.328072643280029,
      "epoch": 0.19551409289400556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998952291890138,
      "loss": 6.1214,
      "mean_token_accuracy": 0.11712106689810753,
      "num_tokens": 4284465.0,
      "step": 1970
    },
    {
      "entropy": 6.1758839130401615,
      "epoch": 0.1960103215561731,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998941452583599,
      "loss": 6.0375,
      "mean_token_accuracy": 0.12105096429586411,
      "num_tokens": 4295266.0,
      "step": 1975
    },
    {
      "entropy": 6.3052088737487795,
      "epoch": 0.1965065502183406,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998930557508156,
      "loss": 6.0844,
      "mean_token_accuracy": 0.11636104881763458,
      "num_tokens": 4306477.0,
      "step": 1980
    },
    {
      "entropy": 6.259469699859619,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.71875,
      "learning_rate": 0.0004998919606664082,
      "loss": 6.113,
      "mean_token_accuracy": 0.11556319296360015,
      "num_tokens": 4319150.0,
      "step": 1985
    },
    {
      "entropy": 6.230728054046631,
      "epoch": 0.19749900754267566,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998908600051648,
      "loss": 6.0739,
      "mean_token_accuracy": 0.12062246054410934,
      "num_tokens": 4330679.0,
      "step": 1990
    },
    {
      "entropy": 6.29167947769165,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998897537671126,
      "loss": 6.0125,
      "mean_token_accuracy": 0.12165434509515763,
      "num_tokens": 4340597.0,
      "step": 1995
    },
    {
      "entropy": 6.019966983795166,
      "epoch": 0.19849146486701072,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998886419522792,
      "loss": 6.0044,
      "mean_token_accuracy": 0.12512857243418693,
      "num_tokens": 4351814.0,
      "step": 2000
    },
    {
      "entropy": 6.3719175338745115,
      "epoch": 0.19898769352917825,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004998875245606919,
      "loss": 6.1068,
      "mean_token_accuracy": 0.12338865548372269,
      "num_tokens": 4363930.0,
      "step": 2005
    },
    {
      "entropy": 6.162980699539185,
      "epoch": 0.19948392219134578,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998864015923788,
      "loss": 6.0042,
      "mean_token_accuracy": 0.12410440891981125,
      "num_tokens": 4376127.0,
      "step": 2010
    },
    {
      "entropy": 6.127772045135498,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998852730473674,
      "loss": 5.9963,
      "mean_token_accuracy": 0.12394299507141113,
      "num_tokens": 4386095.0,
      "step": 2015
    },
    {
      "entropy": 6.152992010116577,
      "epoch": 0.2004763795156808,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998841389256859,
      "loss": 5.9298,
      "mean_token_accuracy": 0.12322643324732781,
      "num_tokens": 4397441.0,
      "step": 2020
    },
    {
      "entropy": 6.104267597198486,
      "epoch": 0.20097260817784834,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998829992273622,
      "loss": 6.029,
      "mean_token_accuracy": 0.1249389573931694,
      "num_tokens": 4408295.0,
      "step": 2025
    },
    {
      "entropy": 6.137422323226929,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998818539524248,
      "loss": 6.0172,
      "mean_token_accuracy": 0.12392294108867645,
      "num_tokens": 4419307.0,
      "step": 2030
    },
    {
      "entropy": 6.212697267532349,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.75390625,
      "learning_rate": 0.000499880703100902,
      "loss": 6.0055,
      "mean_token_accuracy": 0.1231863833963871,
      "num_tokens": 4431908.0,
      "step": 2035
    },
    {
      "entropy": 6.264675903320312,
      "epoch": 0.20246129416435094,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004998795466728223,
      "loss": 6.0427,
      "mean_token_accuracy": 0.12163790315389633,
      "num_tokens": 4443417.0,
      "step": 2040
    },
    {
      "entropy": 6.152185726165771,
      "epoch": 0.20295752282651847,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.0887,
      "mean_token_accuracy": 0.11930633261799813,
      "num_tokens": 4453224.0,
      "step": 2045
    },
    {
      "entropy": 6.206977796554566,
      "epoch": 0.203453751488686,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.9259,
      "mean_token_accuracy": 0.128037216514349,
      "num_tokens": 4463231.0,
      "step": 2050
    },
    {
      "entropy": 6.146184635162354,
      "epoch": 0.20394998015085353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998760439295295,
      "loss": 5.9802,
      "mean_token_accuracy": 0.11942070052027702,
      "num_tokens": 4474397.0,
      "step": 2055
    },
    {
      "entropy": 6.1672406673431395,
      "epoch": 0.20444620881302103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998748651955105,
      "loss": 5.991,
      "mean_token_accuracy": 0.12958767488598824,
      "num_tokens": 4485623.0,
      "step": 2060
    },
    {
      "entropy": 6.16445894241333,
      "epoch": 0.20494243747518856,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998736808850795,
      "loss": 5.9764,
      "mean_token_accuracy": 0.12821295857429504,
      "num_tokens": 4495247.0,
      "step": 2065
    },
    {
      "entropy": 6.162666988372803,
      "epoch": 0.2054386661373561,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004998724909982656,
      "loss": 6.0394,
      "mean_token_accuracy": 0.11950158551335335,
      "num_tokens": 4506490.0,
      "step": 2070
    },
    {
      "entropy": 6.283525276184082,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998712955350986,
      "loss": 6.054,
      "mean_token_accuracy": 0.12060755118727684,
      "num_tokens": 4517511.0,
      "step": 2075
    },
    {
      "entropy": 6.1165454387664795,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998700944956081,
      "loss": 6.0188,
      "mean_token_accuracy": 0.12796789407730103,
      "num_tokens": 4528655.0,
      "step": 2080
    },
    {
      "entropy": 6.175433397293091,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998688878798237,
      "loss": 6.0368,
      "mean_token_accuracy": 0.12539120018482208,
      "num_tokens": 4539369.0,
      "step": 2085
    },
    {
      "entropy": 6.179052114486694,
      "epoch": 0.2074235807860262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998676756877753,
      "loss": 5.9377,
      "mean_token_accuracy": 0.12160812318325043,
      "num_tokens": 4548747.0,
      "step": 2090
    },
    {
      "entropy": 6.033137226104737,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.9297,
      "mean_token_accuracy": 0.12320928424596786,
      "num_tokens": 4561283.0,
      "step": 2095
    },
    {
      "entropy": 6.199184131622315,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004998652345750077,
      "loss": 6.0623,
      "mean_token_accuracy": 0.12004454135894775,
      "num_tokens": 4571929.0,
      "step": 2100
    },
    {
      "entropy": 6.221105670928955,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998640056543487,
      "loss": 6.0467,
      "mean_token_accuracy": 0.1145884521305561,
      "num_tokens": 4583105.0,
      "step": 2105
    },
    {
      "entropy": 6.202408409118652,
      "epoch": 0.2094084954346963,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004998627711575469,
      "loss": 6.1025,
      "mean_token_accuracy": 0.12161830887198448,
      "num_tokens": 4595360.0,
      "step": 2110
    },
    {
      "entropy": 6.137826633453369,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000499861531084633,
      "loss": 6.0667,
      "mean_token_accuracy": 0.1217377468943596,
      "num_tokens": 4607621.0,
      "step": 2115
    },
    {
      "entropy": 6.210973501205444,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.9206,
      "mean_token_accuracy": 0.1290735885500908,
      "num_tokens": 4618286.0,
      "step": 2120
    },
    {
      "entropy": 6.100604152679443,
      "epoch": 0.2108971814211989,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.9972,
      "mean_token_accuracy": 0.12515969797968865,
      "num_tokens": 4628345.0,
      "step": 2125
    },
    {
      "entropy": 6.3336169719696045,
      "epoch": 0.21139341008336643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998577774095262,
      "loss": 6.0492,
      "mean_token_accuracy": 0.1257066994905472,
      "num_tokens": 4639074.0,
      "step": 2130
    },
    {
      "entropy": 6.104618978500366,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004998565150324724,
      "loss": 6.0812,
      "mean_token_accuracy": 0.12450103238224983,
      "num_tokens": 4652698.0,
      "step": 2135
    },
    {
      "entropy": 6.192484760284424,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.9189,
      "mean_token_accuracy": 0.130835435539484,
      "num_tokens": 4663924.0,
      "step": 2140
    },
    {
      "entropy": 6.1426764011383055,
      "epoch": 0.212882096069869,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998539735505251,
      "loss": 6.0329,
      "mean_token_accuracy": 0.12281454429030418,
      "num_tokens": 4673821.0,
      "step": 2145
    },
    {
      "entropy": 6.187332391738892,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.9649,
      "mean_token_accuracy": 0.12759994044899942,
      "num_tokens": 4686646.0,
      "step": 2150
    },
    {
      "entropy": 6.159171628952026,
      "epoch": 0.21387455339420405,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998514097650018,
      "loss": 5.9464,
      "mean_token_accuracy": 0.1264292851090431,
      "num_tokens": 4698118.0,
      "step": 2155
    },
    {
      "entropy": 6.125618743896484,
      "epoch": 0.21437078205637158,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.975,
      "mean_token_accuracy": 0.1183625340461731,
      "num_tokens": 4708869.0,
      "step": 2160
    },
    {
      "entropy": 6.130851745605469,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998488236761569,
      "loss": 5.9915,
      "mean_token_accuracy": 0.1173173077404499,
      "num_tokens": 4720172.0,
      "step": 2165
    },
    {
      "entropy": 6.137715911865234,
      "epoch": 0.21536323938070662,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998475222680688,
      "loss": 6.0444,
      "mean_token_accuracy": 0.1223752222955227,
      "num_tokens": 4731886.0,
      "step": 2170
    },
    {
      "entropy": 6.143771076202393,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.9246,
      "mean_token_accuracy": 0.1209682933986187,
      "num_tokens": 4742381.0,
      "step": 2175
    },
    {
      "entropy": 6.140761852264404,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000499844902724723,
      "loss": 5.8892,
      "mean_token_accuracy": 0.1302737794816494,
      "num_tokens": 4752851.0,
      "step": 2180
    },
    {
      "entropy": 6.099326419830322,
      "epoch": 0.2168519253672092,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.9608,
      "mean_token_accuracy": 0.12656577080488204,
      "num_tokens": 4764271.0,
      "step": 2185
    },
    {
      "entropy": 6.10595817565918,
      "epoch": 0.21734815402937674,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499842260878701,
      "loss": 6.0187,
      "mean_token_accuracy": 0.12466859817504883,
      "num_tokens": 4776216.0,
      "step": 2190
    },
    {
      "entropy": 6.157292985916138,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998409315922683,
      "loss": 5.9608,
      "mean_token_accuracy": 0.12152473330497741,
      "num_tokens": 4786666.0,
      "step": 2195
    },
    {
      "entropy": 6.1741681575775145,
      "epoch": 0.2183406113537118,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998395967302649,
      "loss": 6.0336,
      "mean_token_accuracy": 0.12166730538010598,
      "num_tokens": 4798062.0,
      "step": 2200
    },
    {
      "entropy": 6.1284833431243895,
      "epoch": 0.21883684001587933,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499838256292724,
      "loss": 6.0083,
      "mean_token_accuracy": 0.12412882819771767,
      "num_tokens": 4808837.0,
      "step": 2205
    },
    {
      "entropy": 6.108082437515259,
      "epoch": 0.21933306867804683,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.9248,
      "mean_token_accuracy": 0.12848058566451073,
      "num_tokens": 4819894.0,
      "step": 2210
    },
    {
      "entropy": 6.1002240657806395,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.9195,
      "mean_token_accuracy": 0.12593771144747734,
      "num_tokens": 4830976.0,
      "step": 2215
    },
    {
      "entropy": 6.242991495132446,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998342015272093,
      "loss": 6.0939,
      "mean_token_accuracy": 0.12325356379151345,
      "num_tokens": 4844292.0,
      "step": 2220
    },
    {
      "entropy": 6.131158256530762,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004998328387878521,
      "loss": 5.8928,
      "mean_token_accuracy": 0.13142480850219726,
      "num_tokens": 4855284.0,
      "step": 2225
    },
    {
      "entropy": 6.130004787445069,
      "epoch": 0.22131798332671695,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998314704731249,
      "loss": 6.0136,
      "mean_token_accuracy": 0.1262281470000744,
      "num_tokens": 4867801.0,
      "step": 2230
    },
    {
      "entropy": 6.122326231002807,
      "epoch": 0.22181421198888449,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.8111,
      "mean_token_accuracy": 0.131843601167202,
      "num_tokens": 4877392.0,
      "step": 2235
    },
    {
      "entropy": 6.032444047927856,
      "epoch": 0.22231044065105202,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.9276,
      "mean_token_accuracy": 0.12481720820069313,
      "num_tokens": 4888772.0,
      "step": 2240
    },
    {
      "entropy": 6.186656522750854,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.974,
      "mean_token_accuracy": 0.12461790218949317,
      "num_tokens": 4899739.0,
      "step": 2245
    },
    {
      "entropy": 6.121433925628662,
      "epoch": 0.22330289797538705,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499825941461198,
      "loss": 5.9563,
      "mean_token_accuracy": 0.12550890147686006,
      "num_tokens": 4910455.0,
      "step": 2250
    },
    {
      "entropy": 6.1388201236724855,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.8957,
      "mean_token_accuracy": 0.12686022520065307,
      "num_tokens": 4921156.0,
      "step": 2255
    },
    {
      "entropy": 6.107001113891601,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.915,
      "mean_token_accuracy": 0.12454593926668167,
      "num_tokens": 4931771.0,
      "step": 2260
    },
    {
      "entropy": 6.079628324508667,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.9101,
      "mean_token_accuracy": 0.12066908702254295,
      "num_tokens": 4942790.0,
      "step": 2265
    },
    {
      "entropy": 6.068901777267456,
      "epoch": 0.22528781262405717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.9191,
      "mean_token_accuracy": 0.12233595103025437,
      "num_tokens": 4952752.0,
      "step": 2270
    },
    {
      "entropy": 6.073587369918823,
      "epoch": 0.2257840412862247,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.9571,
      "mean_token_accuracy": 0.12296144515275956,
      "num_tokens": 4963659.0,
      "step": 2275
    },
    {
      "entropy": 6.169452142715454,
      "epoch": 0.22628026994839223,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.9291,
      "mean_token_accuracy": 0.1246725894510746,
      "num_tokens": 4976121.0,
      "step": 2280
    },
    {
      "entropy": 6.072994184494019,
      "epoch": 0.22677649861055973,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.9597,
      "mean_token_accuracy": 0.12557747438549996,
      "num_tokens": 4986515.0,
      "step": 2285
    },
    {
      "entropy": 6.053368473052979,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.8354,
      "mean_token_accuracy": 0.12849313914775848,
      "num_tokens": 4998231.0,
      "step": 2290
    },
    {
      "entropy": 6.0678062915802,
      "epoch": 0.2277689559348948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.8649,
      "mean_token_accuracy": 0.12834521606564522,
      "num_tokens": 5010309.0,
      "step": 2295
    },
    {
      "entropy": 6.064406776428223,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9625,
      "mean_token_accuracy": 0.13037839084863662,
      "num_tokens": 5020864.0,
      "step": 2300
    },
    {
      "entropy": 6.049251317977905,
      "epoch": 0.22876141325922986,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9612,
      "mean_token_accuracy": 0.1286421723663807,
      "num_tokens": 5032224.0,
      "step": 2305
    },
    {
      "entropy": 6.169411468505859,
      "epoch": 0.2292576419213974,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.9116,
      "mean_token_accuracy": 0.13263707458972931,
      "num_tokens": 5044318.0,
      "step": 2310
    },
    {
      "entropy": 6.110779809951782,
      "epoch": 0.22975387058356492,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998073561245801,
      "loss": 6.0024,
      "mean_token_accuracy": 0.12361429855227471,
      "num_tokens": 5055270.0,
      "step": 2315
    },
    {
      "entropy": 6.167765951156616,
      "epoch": 0.23025009924573242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.9674,
      "mean_token_accuracy": 0.12938261702656745,
      "num_tokens": 5064771.0,
      "step": 2320
    },
    {
      "entropy": 6.092255401611328,
      "epoch": 0.23074632790789995,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.9604,
      "mean_token_accuracy": 0.12339024245738983,
      "num_tokens": 5075509.0,
      "step": 2325
    },
    {
      "entropy": 6.085866403579712,
      "epoch": 0.23124255657006748,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499802933404163,
      "loss": 5.9046,
      "mean_token_accuracy": 0.12324860319495201,
      "num_tokens": 5086270.0,
      "step": 2330
    },
    {
      "entropy": 6.066864776611328,
      "epoch": 0.231738785232235,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.856,
      "mean_token_accuracy": 0.13291038572788239,
      "num_tokens": 5096695.0,
      "step": 2335
    },
    {
      "entropy": 6.0964189052581785,
      "epoch": 0.23223501389440254,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.9362,
      "mean_token_accuracy": 0.12667595222592354,
      "num_tokens": 5107775.0,
      "step": 2340
    },
    {
      "entropy": 6.1279243469238285,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.9071,
      "mean_token_accuracy": 0.12755124419927596,
      "num_tokens": 5118215.0,
      "step": 2345
    },
    {
      "entropy": 6.1014261722564695,
      "epoch": 0.2332274712187376,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.9279,
      "mean_token_accuracy": 0.1297263316810131,
      "num_tokens": 5129509.0,
      "step": 2350
    },
    {
      "entropy": 6.018870449066162,
      "epoch": 0.23372369988090513,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.8869,
      "mean_token_accuracy": 0.12231870293617249,
      "num_tokens": 5140689.0,
      "step": 2355
    },
    {
      "entropy": 6.0995978832244875,
      "epoch": 0.23421992854307264,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.8627,
      "mean_token_accuracy": 0.12557717710733413,
      "num_tokens": 5151060.0,
      "step": 2360
    },
    {
      "entropy": 6.008219385147095,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499792418612928,
      "loss": 5.8457,
      "mean_token_accuracy": 0.12821394056081772,
      "num_tokens": 5161481.0,
      "step": 2365
    },
    {
      "entropy": 6.113403987884522,
      "epoch": 0.2352123858674077,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.9664,
      "mean_token_accuracy": 0.12153320610523224,
      "num_tokens": 5171961.0,
      "step": 2370
    },
    {
      "entropy": 6.208441591262817,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004997893642169904,
      "loss": 6.0081,
      "mean_token_accuracy": 0.11836620941758155,
      "num_tokens": 5184158.0,
      "step": 2375
    },
    {
      "entropy": 6.094049882888794,
      "epoch": 0.23620484319174276,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997878286575648,
      "loss": 6.0239,
      "mean_token_accuracy": 0.1250876732170582,
      "num_tokens": 5195942.0,
      "step": 2380
    },
    {
      "entropy": 6.102467155456543,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.8999,
      "mean_token_accuracy": 0.12667744383215904,
      "num_tokens": 5208063.0,
      "step": 2385
    },
    {
      "entropy": 6.109834671020508,
      "epoch": 0.23719730051607782,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9515,
      "mean_token_accuracy": 0.1283447615802288,
      "num_tokens": 5219424.0,
      "step": 2390
    },
    {
      "entropy": 6.169302701950073,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.95,
      "mean_token_accuracy": 0.12057619765400887,
      "num_tokens": 5230351.0,
      "step": 2395
    },
    {
      "entropy": 6.002688694000244,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.8321,
      "mean_token_accuracy": 0.1295178197324276,
      "num_tokens": 5242906.0,
      "step": 2400
    },
    {
      "entropy": 6.046356296539306,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997800672473969,
      "loss": 5.8767,
      "mean_token_accuracy": 0.13203179612755775,
      "num_tokens": 5253689.0,
      "step": 2405
    },
    {
      "entropy": 6.11405086517334,
      "epoch": 0.2391822151647479,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.9668,
      "mean_token_accuracy": 0.127695781737566,
      "num_tokens": 5265896.0,
      "step": 2410
    },
    {
      "entropy": 6.092328882217407,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.891,
      "mean_token_accuracy": 0.1248737022280693,
      "num_tokens": 5276800.0,
      "step": 2415
    },
    {
      "entropy": 6.026219177246094,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997753435122522,
      "loss": 5.8682,
      "mean_token_accuracy": 0.13026652336120606,
      "num_tokens": 5289113.0,
      "step": 2420
    },
    {
      "entropy": 6.040958499908447,
      "epoch": 0.2406709011512505,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049977375778593,
      "loss": 5.8421,
      "mean_token_accuracy": 0.12867625132203103,
      "num_tokens": 5298825.0,
      "step": 2425
    },
    {
      "entropy": 5.986614465713501,
      "epoch": 0.24116712981341804,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.8609,
      "mean_token_accuracy": 0.13291131407022477,
      "num_tokens": 5310376.0,
      "step": 2430
    },
    {
      "entropy": 6.01220850944519,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.869,
      "mean_token_accuracy": 0.1286493256688118,
      "num_tokens": 5320483.0,
      "step": 2435
    },
    {
      "entropy": 6.052859735488892,
      "epoch": 0.24215958713775307,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.7784,
      "mean_token_accuracy": 0.13417652249336243,
      "num_tokens": 5330995.0,
      "step": 2440
    },
    {
      "entropy": 6.047038984298706,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.906,
      "mean_token_accuracy": 0.13337108343839646,
      "num_tokens": 5341208.0,
      "step": 2445
    },
    {
      "entropy": 6.080028533935547,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.9776,
      "mean_token_accuracy": 0.1305466502904892,
      "num_tokens": 5351769.0,
      "step": 2450
    },
    {
      "entropy": 6.106999635696411,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.7579,
      "mean_token_accuracy": 0.13589820861816407,
      "num_tokens": 5361033.0,
      "step": 2455
    },
    {
      "entropy": 5.878340005874634,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499762501634559,
      "loss": 5.8218,
      "mean_token_accuracy": 0.12982406690716744,
      "num_tokens": 5372381.0,
      "step": 2460
    },
    {
      "entropy": 6.031815814971924,
      "epoch": 0.24464073044859072,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.887,
      "mean_token_accuracy": 0.1292889140546322,
      "num_tokens": 5383101.0,
      "step": 2465
    },
    {
      "entropy": 6.09915075302124,
      "epoch": 0.24513695911075822,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.9121,
      "mean_token_accuracy": 0.12618087828159333,
      "num_tokens": 5393661.0,
      "step": 2470
    },
    {
      "entropy": 6.133698320388794,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000499757593964499,
      "loss": 5.9332,
      "mean_token_accuracy": 0.13135130926966668,
      "num_tokens": 5405377.0,
      "step": 2475
    },
    {
      "entropy": 5.966562271118164,
      "epoch": 0.24612941643509328,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.8388,
      "mean_token_accuracy": 0.12627245411276816,
      "num_tokens": 5417172.0,
      "step": 2480
    },
    {
      "entropy": 6.075893211364746,
      "epoch": 0.24662564509726081,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.9689,
      "mean_token_accuracy": 0.1163121998310089,
      "num_tokens": 5428053.0,
      "step": 2485
    },
    {
      "entropy": 6.147674179077148,
      "epoch": 0.24712187375942835,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.8906,
      "mean_token_accuracy": 0.12604017555713654,
      "num_tokens": 5438282.0,
      "step": 2490
    },
    {
      "entropy": 5.98391547203064,
      "epoch": 0.24761810242159588,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8671,
      "mean_token_accuracy": 0.12558303028345108,
      "num_tokens": 5448940.0,
      "step": 2495
    },
    {
      "entropy": 5.965530967712402,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1320234015583992,
      "num_tokens": 5458471.0,
      "step": 2500
    },
    {
      "entropy": 6.013654708862305,
      "epoch": 0.24861055974593094,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.8341,
      "mean_token_accuracy": 0.13364496156573297,
      "num_tokens": 5467979.0,
      "step": 2505
    },
    {
      "entropy": 6.064499998092652,
      "epoch": 0.24910678840809844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.9608,
      "mean_token_accuracy": 0.12101689428091049,
      "num_tokens": 5479852.0,
      "step": 2510
    },
    {
      "entropy": 6.007748031616211,
      "epoch": 0.24960301707026597,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.8138,
      "mean_token_accuracy": 0.12828044667840005,
      "num_tokens": 5491827.0,
      "step": 2515
    },
    {
      "entropy": 6.006802654266357,
      "epoch": 0.25009924573243353,
      "grad_norm": 0.875,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.8029,
      "mean_token_accuracy": 0.1329714372754097,
      "num_tokens": 5502206.0,
      "step": 2520
    },
    {
      "entropy": 5.989188814163208,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.8227,
      "mean_token_accuracy": 0.12857313305139542,
      "num_tokens": 5514352.0,
      "step": 2525
    },
    {
      "entropy": 6.100388622283935,
      "epoch": 0.25109170305676853,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.8341,
      "mean_token_accuracy": 0.1370926581323147,
      "num_tokens": 5524413.0,
      "step": 2530
    },
    {
      "entropy": 5.976933336257934,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499737461680026,
      "loss": 5.8412,
      "mean_token_accuracy": 0.12043567895889282,
      "num_tokens": 5533692.0,
      "step": 2535
    },
    {
      "entropy": 6.050266551971435,
      "epoch": 0.2520841603811036,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.9739,
      "mean_token_accuracy": 0.12689899504184723,
      "num_tokens": 5545955.0,
      "step": 2540
    },
    {
      "entropy": 6.0521016120910645,
      "epoch": 0.2525803890432711,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.8999,
      "mean_token_accuracy": 0.1276570178568363,
      "num_tokens": 5557667.0,
      "step": 2545
    },
    {
      "entropy": 5.9338500022888185,
      "epoch": 0.25307661770543866,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499732303213466,
      "loss": 5.8107,
      "mean_token_accuracy": 0.1274030514061451,
      "num_tokens": 5567635.0,
      "step": 2550
    },
    {
      "entropy": 5.971767520904541,
      "epoch": 0.2535728463676062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.8188,
      "mean_token_accuracy": 0.13125893771648406,
      "num_tokens": 5578584.0,
      "step": 2555
    },
    {
      "entropy": 6.077448606491089,
      "epoch": 0.2540690750297737,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499728836371304,
      "loss": 5.924,
      "mean_token_accuracy": 0.13032637014985085,
      "num_tokens": 5589679.0,
      "step": 2560
    },
    {
      "entropy": 6.066747713088989,
      "epoch": 0.25456530369194125,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.8927,
      "mean_token_accuracy": 0.131753259152174,
      "num_tokens": 5601550.0,
      "step": 2565
    },
    {
      "entropy": 6.0625098705291744,
      "epoch": 0.2550615323541088,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.9181,
      "mean_token_accuracy": 0.12465985491871834,
      "num_tokens": 5611777.0,
      "step": 2570
    },
    {
      "entropy": 6.093583059310913,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.9347,
      "mean_token_accuracy": 0.1278335139155388,
      "num_tokens": 5623792.0,
      "step": 2575
    },
    {
      "entropy": 5.975198221206665,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.7337,
      "mean_token_accuracy": 0.13377925157546997,
      "num_tokens": 5634072.0,
      "step": 2580
    },
    {
      "entropy": 6.06852536201477,
      "epoch": 0.25655021834061137,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.9925,
      "mean_token_accuracy": 0.12501400709152222,
      "num_tokens": 5645046.0,
      "step": 2585
    },
    {
      "entropy": 5.986842298507691,
      "epoch": 0.2570464470027789,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.8603,
      "mean_token_accuracy": 0.1285901851952076,
      "num_tokens": 5655156.0,
      "step": 2590
    },
    {
      "entropy": 6.025115919113159,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.8286,
      "mean_token_accuracy": 0.12798064202070236,
      "num_tokens": 5665621.0,
      "step": 2595
    },
    {
      "entropy": 6.023578310012818,
      "epoch": 0.25803890432711396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.8843,
      "mean_token_accuracy": 0.12540870010852814,
      "num_tokens": 5677072.0,
      "step": 2600
    },
    {
      "entropy": 6.093571472167969,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.834,
      "mean_token_accuracy": 0.13478555008769036,
      "num_tokens": 5687926.0,
      "step": 2605
    },
    {
      "entropy": 5.910076093673706,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.8391,
      "mean_token_accuracy": 0.1295539267361164,
      "num_tokens": 5698923.0,
      "step": 2610
    },
    {
      "entropy": 6.104871034622192,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.9432,
      "mean_token_accuracy": 0.12401391565799713,
      "num_tokens": 5709879.0,
      "step": 2615
    },
    {
      "entropy": 6.0787454605102536,
      "epoch": 0.260023818975784,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049970756721609,
      "loss": 5.8857,
      "mean_token_accuracy": 0.12929130122065544,
      "num_tokens": 5720698.0,
      "step": 2620
    },
    {
      "entropy": 5.974215459823609,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499705758565719,
      "loss": 5.8317,
      "mean_token_accuracy": 0.1328559137880802,
      "num_tokens": 5732890.0,
      "step": 2625
    },
    {
      "entropy": 6.003826332092285,
      "epoch": 0.2610162763001191,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997039443431292,
      "loss": 5.8339,
      "mean_token_accuracy": 0.1282294899225235,
      "num_tokens": 5743741.0,
      "step": 2630
    },
    {
      "entropy": 5.945276737213135,
      "epoch": 0.2615125049622866,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.745,
      "mean_token_accuracy": 0.1344357542693615,
      "num_tokens": 5754730.0,
      "step": 2635
    },
    {
      "entropy": 5.975873947143555,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499700299181474,
      "loss": 5.7816,
      "mean_token_accuracy": 0.13547359257936478,
      "num_tokens": 5764705.0,
      "step": 2640
    },
    {
      "entropy": 6.053955745697022,
      "epoch": 0.2625049622866217,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.9062,
      "mean_token_accuracy": 0.12410091534256935,
      "num_tokens": 5776803.0,
      "step": 2645
    },
    {
      "entropy": 6.090683031082153,
      "epoch": 0.2630011909487892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499696631731486,
      "loss": 5.9541,
      "mean_token_accuracy": 0.12392831742763519,
      "num_tokens": 5788203.0,
      "step": 2650
    },
    {
      "entropy": 6.06274151802063,
      "epoch": 0.26349741961095674,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.8677,
      "mean_token_accuracy": 0.12673098966479301,
      "num_tokens": 5800476.0,
      "step": 2655
    },
    {
      "entropy": 5.944343852996826,
      "epoch": 0.26399364827312427,
      "grad_norm": 0.875,
      "learning_rate": 0.000499692941993529,
      "loss": 5.922,
      "mean_token_accuracy": 0.12605136781930923,
      "num_tokens": 5811652.0,
      "step": 2660
    },
    {
      "entropy": 6.055945110321045,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.8237,
      "mean_token_accuracy": 0.1357803925871849,
      "num_tokens": 5823036.0,
      "step": 2665
    },
    {
      "entropy": 5.979827785491944,
      "epoch": 0.26498610559745933,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499689229967969,
      "loss": 5.7742,
      "mean_token_accuracy": 0.13261008635163307,
      "num_tokens": 5833220.0,
      "step": 2670
    },
    {
      "entropy": 5.952842378616333,
      "epoch": 0.26548233425962686,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.8511,
      "mean_token_accuracy": 0.12471630349755287,
      "num_tokens": 5843777.0,
      "step": 2675
    },
    {
      "entropy": 6.013588809967041,
      "epoch": 0.26597856292179434,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.8734,
      "mean_token_accuracy": 0.12791570276021957,
      "num_tokens": 5854942.0,
      "step": 2680
    },
    {
      "entropy": 6.092654275894165,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.8779,
      "mean_token_accuracy": 0.1297077603638172,
      "num_tokens": 5866536.0,
      "step": 2685
    },
    {
      "entropy": 5.936705446243286,
      "epoch": 0.2669710202461294,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499681739055515,
      "loss": 5.7586,
      "mean_token_accuracy": 0.13553930893540383,
      "num_tokens": 5876653.0,
      "step": 2690
    },
    {
      "entropy": 6.047937488555908,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.8022,
      "mean_token_accuracy": 0.13020532354712486,
      "num_tokens": 5887092.0,
      "step": 2695
    },
    {
      "entropy": 6.006089544296264,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499677960169364,
      "loss": 5.9163,
      "mean_token_accuracy": 0.12631798684597015,
      "num_tokens": 5898224.0,
      "step": 2700
    },
    {
      "entropy": 6.0130256652832035,
      "epoch": 0.268459706232632,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.7482,
      "mean_token_accuracy": 0.13969457894563675,
      "num_tokens": 5909235.0,
      "step": 2705
    },
    {
      "entropy": 6.01822190284729,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499674158997096,
      "loss": 5.9196,
      "mean_token_accuracy": 0.13010889217257499,
      "num_tokens": 5921299.0,
      "step": 2710
    },
    {
      "entropy": 6.007842779159546,
      "epoch": 0.26945216355696705,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.8747,
      "mean_token_accuracy": 0.12728465348482132,
      "num_tokens": 5933249.0,
      "step": 2715
    },
    {
      "entropy": 6.0062947273254395,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.9094,
      "mean_token_accuracy": 0.1252631276845932,
      "num_tokens": 5945942.0,
      "step": 2720
    },
    {
      "entropy": 5.968673038482666,
      "epoch": 0.2704446208813021,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049966841545305,
      "loss": 5.8196,
      "mean_token_accuracy": 0.13000037670135497,
      "num_tokens": 5956315.0,
      "step": 2725
    },
    {
      "entropy": 6.0654479503631595,
      "epoch": 0.27094084954346964,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.8891,
      "mean_token_accuracy": 0.12286564558744431,
      "num_tokens": 5966566.0,
      "step": 2730
    },
    {
      "entropy": 5.943171072006225,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.7425,
      "mean_token_accuracy": 0.13684265613555907,
      "num_tokens": 5977137.0,
      "step": 2735
    },
    {
      "entropy": 6.021245813369751,
      "epoch": 0.2719333068678047,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.8674,
      "mean_token_accuracy": 0.1314192496240139,
      "num_tokens": 5987653.0,
      "step": 2740
    },
    {
      "entropy": 6.0384129047393795,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.9187,
      "mean_token_accuracy": 0.12729719430208206,
      "num_tokens": 5998530.0,
      "step": 2745
    },
    {
      "entropy": 5.968395137786866,
      "epoch": 0.27292576419213976,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.7641,
      "mean_token_accuracy": 0.13758938238024712,
      "num_tokens": 6009550.0,
      "step": 2750
    },
    {
      "entropy": 6.017692184448242,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.9465,
      "mean_token_accuracy": 0.12807289212942125,
      "num_tokens": 6021522.0,
      "step": 2755
    },
    {
      "entropy": 6.015359878540039,
      "epoch": 0.27391822151647477,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.8181,
      "mean_token_accuracy": 0.12846075966954232,
      "num_tokens": 6033993.0,
      "step": 2760
    },
    {
      "entropy": 6.101863288879395,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.9794,
      "mean_token_accuracy": 0.1278089925646782,
      "num_tokens": 6044804.0,
      "step": 2765
    },
    {
      "entropy": 5.9826210021972654,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499650883976301,
      "loss": 5.8315,
      "mean_token_accuracy": 0.12525355890393258,
      "num_tokens": 6054878.0,
      "step": 2770
    },
    {
      "entropy": 6.0613480567932125,
      "epoch": 0.27540690750297736,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.83,
      "mean_token_accuracy": 0.12792847976088523,
      "num_tokens": 6065492.0,
      "step": 2775
    },
    {
      "entropy": 6.012570095062256,
      "epoch": 0.2759031361651449,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.8623,
      "mean_token_accuracy": 0.12609369233250617,
      "num_tokens": 6077360.0,
      "step": 2780
    },
    {
      "entropy": 5.981580781936645,
      "epoch": 0.2763993648273124,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.8343,
      "mean_token_accuracy": 0.12407092973589898,
      "num_tokens": 6087530.0,
      "step": 2785
    },
    {
      "entropy": 6.0558830261230465,
      "epoch": 0.27689559348947995,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.8194,
      "mean_token_accuracy": 0.13177746906876564,
      "num_tokens": 6098064.0,
      "step": 2790
    },
    {
      "entropy": 5.905488109588623,
      "epoch": 0.2773918221516475,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.7419,
      "mean_token_accuracy": 0.1355844110250473,
      "num_tokens": 6108446.0,
      "step": 2795
    },
    {
      "entropy": 5.991134786605835,
      "epoch": 0.277888050813815,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.7935,
      "mean_token_accuracy": 0.12804083600640298,
      "num_tokens": 6118923.0,
      "step": 2800
    },
    {
      "entropy": 5.9639771461486815,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004996369364134507,
      "loss": 5.6892,
      "mean_token_accuracy": 0.13721818029880523,
      "num_tokens": 6130583.0,
      "step": 2805
    },
    {
      "entropy": 5.992746829986572,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499634921622027,
      "loss": 5.9108,
      "mean_token_accuracy": 0.1296452686190605,
      "num_tokens": 6142351.0,
      "step": 2810
    },
    {
      "entropy": 5.940733814239502,
      "epoch": 0.2793767368003176,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.7348,
      "mean_token_accuracy": 0.1339046314358711,
      "num_tokens": 6152964.0,
      "step": 2815
    },
    {
      "entropy": 5.968123912811279,
      "epoch": 0.27987296546248513,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.8149,
      "mean_token_accuracy": 0.13025947213172911,
      "num_tokens": 6163522.0,
      "step": 2820
    },
    {
      "entropy": 6.044860601425171,
      "epoch": 0.28036919412465267,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.8309,
      "mean_token_accuracy": 0.12943392023444175,
      "num_tokens": 6174060.0,
      "step": 2825
    },
    {
      "entropy": 5.940817499160767,
      "epoch": 0.28086542278682014,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.7561,
      "mean_token_accuracy": 0.13433966636657715,
      "num_tokens": 6183850.0,
      "step": 2830
    },
    {
      "entropy": 6.010298871994019,
      "epoch": 0.28136165144898767,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.822,
      "mean_token_accuracy": 0.12921777442097665,
      "num_tokens": 6195073.0,
      "step": 2835
    },
    {
      "entropy": 6.014202022552491,
      "epoch": 0.2818578801111552,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.8334,
      "mean_token_accuracy": 0.13027386069297792,
      "num_tokens": 6206658.0,
      "step": 2840
    },
    {
      "entropy": 5.891300058364868,
      "epoch": 0.28235410877332273,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00049962066211194,
      "loss": 5.7269,
      "mean_token_accuracy": 0.1362061731517315,
      "num_tokens": 6217183.0,
      "step": 2845
    },
    {
      "entropy": 5.912215232849121,
      "epoch": 0.28285033743549026,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.7702,
      "mean_token_accuracy": 0.1330777660012245,
      "num_tokens": 6227575.0,
      "step": 2850
    },
    {
      "entropy": 5.9731401920318605,
      "epoch": 0.2833465660976578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.7632,
      "mean_token_accuracy": 0.13205480054020882,
      "num_tokens": 6239698.0,
      "step": 2855
    },
    {
      "entropy": 5.963889408111572,
      "epoch": 0.2838427947598253,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.7796,
      "mean_token_accuracy": 0.13623137325048446,
      "num_tokens": 6249730.0,
      "step": 2860
    },
    {
      "entropy": 5.922169256210327,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.7807,
      "mean_token_accuracy": 0.1335216723382473,
      "num_tokens": 6259493.0,
      "step": 2865
    },
    {
      "entropy": 5.947891187667847,
      "epoch": 0.2848352520841604,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.7801,
      "mean_token_accuracy": 0.13493156805634499,
      "num_tokens": 6269568.0,
      "step": 2870
    },
    {
      "entropy": 5.986701917648316,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.8276,
      "mean_token_accuracy": 0.13422205969691275,
      "num_tokens": 6280028.0,
      "step": 2875
    },
    {
      "entropy": 5.926356792449951,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.7768,
      "mean_token_accuracy": 0.13312428295612336,
      "num_tokens": 6290321.0,
      "step": 2880
    },
    {
      "entropy": 5.971750688552857,
      "epoch": 0.286323938070663,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.7778,
      "mean_token_accuracy": 0.13141178712248802,
      "num_tokens": 6300729.0,
      "step": 2885
    },
    {
      "entropy": 5.991149187088013,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.8172,
      "mean_token_accuracy": 0.12256841883063316,
      "num_tokens": 6311051.0,
      "step": 2890
    },
    {
      "entropy": 5.88832540512085,
      "epoch": 0.28731639539499804,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000499599817928441,
      "loss": 5.7324,
      "mean_token_accuracy": 0.14212445318698883,
      "num_tokens": 6322201.0,
      "step": 2895
    },
    {
      "entropy": 5.9889750480651855,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.8283,
      "mean_token_accuracy": 0.13292579874396324,
      "num_tokens": 6332372.0,
      "step": 2900
    },
    {
      "entropy": 6.032767820358276,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.9456,
      "mean_token_accuracy": 0.1312311202287674,
      "num_tokens": 6344070.0,
      "step": 2905
    },
    {
      "entropy": 5.872374725341797,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.7614,
      "mean_token_accuracy": 0.13059834465384484,
      "num_tokens": 6355118.0,
      "step": 2910
    },
    {
      "entropy": 6.010786342620849,
      "epoch": 0.2893013100436681,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.8164,
      "mean_token_accuracy": 0.13058557361364365,
      "num_tokens": 6366722.0,
      "step": 2915
    },
    {
      "entropy": 5.9085742950439455,
      "epoch": 0.28979753870583563,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.763,
      "mean_token_accuracy": 0.13042425736784935,
      "num_tokens": 6376190.0,
      "step": 2920
    },
    {
      "entropy": 6.031855201721191,
      "epoch": 0.29029376736800316,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.8846,
      "mean_token_accuracy": 0.12776149213314056,
      "num_tokens": 6385883.0,
      "step": 2925
    },
    {
      "entropy": 5.9083640575408936,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.7467,
      "mean_token_accuracy": 0.13742488473653794,
      "num_tokens": 6397762.0,
      "step": 2930
    },
    {
      "entropy": 5.953339385986328,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.8544,
      "mean_token_accuracy": 0.1268588684499264,
      "num_tokens": 6408972.0,
      "step": 2935
    },
    {
      "entropy": 5.9890693664550785,
      "epoch": 0.29178245335450576,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499580581961813,
      "loss": 5.7297,
      "mean_token_accuracy": 0.12740637883543968,
      "num_tokens": 6418998.0,
      "step": 2940
    },
    {
      "entropy": 5.9924187660217285,
      "epoch": 0.2922786820166733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.8176,
      "mean_token_accuracy": 0.133935584127903,
      "num_tokens": 6430046.0,
      "step": 2945
    },
    {
      "entropy": 6.004527235031128,
      "epoch": 0.2927749106788408,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.7752,
      "mean_token_accuracy": 0.13516283929347991,
      "num_tokens": 6441351.0,
      "step": 2950
    },
    {
      "entropy": 5.874765920639038,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.8047,
      "mean_token_accuracy": 0.13173421323299409,
      "num_tokens": 6453259.0,
      "step": 2955
    },
    {
      "entropy": 5.952497243881226,
      "epoch": 0.2937673680031759,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.7901,
      "mean_token_accuracy": 0.1302712343633175,
      "num_tokens": 6464819.0,
      "step": 2960
    },
    {
      "entropy": 6.05925498008728,
      "epoch": 0.2942635966653434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.8809,
      "mean_token_accuracy": 0.13085756078362465,
      "num_tokens": 6477315.0,
      "step": 2965
    },
    {
      "entropy": 5.986642456054687,
      "epoch": 0.29475982532751094,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.7895,
      "mean_token_accuracy": 0.1315566822886467,
      "num_tokens": 6488461.0,
      "step": 2970
    },
    {
      "entropy": 5.9029899597167965,
      "epoch": 0.29525605398967847,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.7658,
      "mean_token_accuracy": 0.13758422061800957,
      "num_tokens": 6499486.0,
      "step": 2975
    },
    {
      "entropy": 6.06349925994873,
      "epoch": 0.29575228265184594,
      "grad_norm": 0.8125,
      "learning_rate": 0.000499563104623865,
      "loss": 5.8495,
      "mean_token_accuracy": 0.13247794285416603,
      "num_tokens": 6511337.0,
      "step": 2980
    },
    {
      "entropy": 5.87068018913269,
      "epoch": 0.2962485113140135,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.7496,
      "mean_token_accuracy": 0.13712706938385963,
      "num_tokens": 6521066.0,
      "step": 2985
    },
    {
      "entropy": 5.923775768280029,
      "epoch": 0.296744739976181,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.7578,
      "mean_token_accuracy": 0.13636653870344162,
      "num_tokens": 6531267.0,
      "step": 2990
    },
    {
      "entropy": 5.9300535202026365,
      "epoch": 0.29724096863834853,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.7948,
      "mean_token_accuracy": 0.12985561564564704,
      "num_tokens": 6541891.0,
      "step": 2995
    },
    {
      "entropy": 5.960822010040284,
      "epoch": 0.29773719730051607,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.7554,
      "mean_token_accuracy": 0.1335455968976021,
      "num_tokens": 6551974.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.791192339332746,
      "eval_loss": 5.854363441467285,
      "eval_mean_token_accuracy": 0.13373962398746797,
      "eval_num_tokens": 6551974.0,
      "eval_runtime": 20.5635,
      "eval_samples_per_second": 1637.315,
      "eval_steps_per_second": 204.683,
      "step": 3000
    },
    {
      "entropy": 5.964084672927856,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499552000303751,
      "loss": 5.7652,
      "mean_token_accuracy": 0.12887574657797812,
      "num_tokens": 6563296.0,
      "step": 3005
    },
    {
      "entropy": 5.991165637969971,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00049954976273429,
      "loss": 5.7989,
      "mean_token_accuracy": 0.13300191164016723,
      "num_tokens": 6573699.0,
      "step": 3010
    },
    {
      "entropy": 5.947231817245483,
      "epoch": 0.29922588328701866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.7995,
      "mean_token_accuracy": 0.1347610831260681,
      "num_tokens": 6585111.0,
      "step": 3015
    },
    {
      "entropy": 5.9104045867919925,
      "epoch": 0.2997221119491862,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.7754,
      "mean_token_accuracy": 0.1333615891635418,
      "num_tokens": 6596102.0,
      "step": 3020
    },
    {
      "entropy": 5.943351411819458,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499543016616028,
      "loss": 5.7518,
      "mean_token_accuracy": 0.13800903409719467,
      "num_tokens": 6605989.0,
      "step": 3025
    },
    {
      "entropy": 5.972800874710083,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.7637,
      "mean_token_accuracy": 0.13150317296385766,
      "num_tokens": 6615771.0,
      "step": 3030
    },
    {
      "entropy": 5.944656038284302,
      "epoch": 0.3012107979356888,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.7159,
      "mean_token_accuracy": 0.13868248835206032,
      "num_tokens": 6626707.0,
      "step": 3035
    },
    {
      "entropy": 5.9109409809112545,
      "epoch": 0.3017070265978563,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.7556,
      "mean_token_accuracy": 0.13460727781057358,
      "num_tokens": 6637539.0,
      "step": 3040
    },
    {
      "entropy": 5.881481599807739,
      "epoch": 0.30220325526002384,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.8225,
      "mean_token_accuracy": 0.1278907537460327,
      "num_tokens": 6648372.0,
      "step": 3045
    },
    {
      "entropy": 5.964266204833985,
      "epoch": 0.30269948392219137,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.8262,
      "mean_token_accuracy": 0.12617874816060065,
      "num_tokens": 6658732.0,
      "step": 3050
    },
    {
      "entropy": 5.966398048400879,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.7029,
      "mean_token_accuracy": 0.13636866360902786,
      "num_tokens": 6670108.0,
      "step": 3055
    },
    {
      "entropy": 5.956323051452637,
      "epoch": 0.3036919412465264,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.8268,
      "mean_token_accuracy": 0.12821760326623916,
      "num_tokens": 6681337.0,
      "step": 3060
    },
    {
      "entropy": 5.938261938095093,
      "epoch": 0.3041881699086939,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.7428,
      "mean_token_accuracy": 0.13437946289777755,
      "num_tokens": 6691221.0,
      "step": 3065
    },
    {
      "entropy": 6.030601119995117,
      "epoch": 0.30468439857086144,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995224775854454,
      "loss": 5.8923,
      "mean_token_accuracy": 0.12948382943868636,
      "num_tokens": 6703363.0,
      "step": 3070
    },
    {
      "entropy": 5.943697214126587,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.7937,
      "mean_token_accuracy": 0.13257720544934273,
      "num_tokens": 6715526.0,
      "step": 3075
    },
    {
      "entropy": 5.942606639862061,
      "epoch": 0.3056768558951965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.7779,
      "mean_token_accuracy": 0.1298063114285469,
      "num_tokens": 6725607.0,
      "step": 3080
    },
    {
      "entropy": 5.915257167816162,
      "epoch": 0.30617308455736403,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.729,
      "mean_token_accuracy": 0.13447875827550887,
      "num_tokens": 6735762.0,
      "step": 3085
    },
    {
      "entropy": 5.9864640712738035,
      "epoch": 0.30666931321953156,
      "grad_norm": 0.875,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.8398,
      "mean_token_accuracy": 0.13669491484761237,
      "num_tokens": 6746334.0,
      "step": 3090
    },
    {
      "entropy": 5.963445377349854,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.7931,
      "mean_token_accuracy": 0.13090421184897422,
      "num_tokens": 6756082.0,
      "step": 3095
    },
    {
      "entropy": 5.940825796127319,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.835,
      "mean_token_accuracy": 0.1257372260093689,
      "num_tokens": 6765723.0,
      "step": 3100
    },
    {
      "entropy": 5.963672399520874,
      "epoch": 0.30815799920603415,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.8091,
      "mean_token_accuracy": 0.12877568751573562,
      "num_tokens": 6778567.0,
      "step": 3105
    },
    {
      "entropy": 6.004975605010986,
      "epoch": 0.3086542278682017,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.8442,
      "mean_token_accuracy": 0.13627742007374763,
      "num_tokens": 6789721.0,
      "step": 3110
    },
    {
      "entropy": 5.938079833984375,
      "epoch": 0.3091504565303692,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.7403,
      "mean_token_accuracy": 0.13720087036490441,
      "num_tokens": 6800606.0,
      "step": 3115
    },
    {
      "entropy": 5.884873104095459,
      "epoch": 0.30964668519253674,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.7948,
      "mean_token_accuracy": 0.1325895868241787,
      "num_tokens": 6813068.0,
      "step": 3120
    },
    {
      "entropy": 5.913717317581177,
      "epoch": 0.31014291385470427,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.7394,
      "mean_token_accuracy": 0.13386166617274284,
      "num_tokens": 6824502.0,
      "step": 3125
    },
    {
      "entropy": 5.92196798324585,
      "epoch": 0.31063914251687175,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.772,
      "mean_token_accuracy": 0.13242748379707336,
      "num_tokens": 6835434.0,
      "step": 3130
    },
    {
      "entropy": 5.92221474647522,
      "epoch": 0.3111353711790393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.822,
      "mean_token_accuracy": 0.1294836826622486,
      "num_tokens": 6846754.0,
      "step": 3135
    },
    {
      "entropy": 5.976176834106445,
      "epoch": 0.3116315998412068,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.748,
      "mean_token_accuracy": 0.1322815276682377,
      "num_tokens": 6858050.0,
      "step": 3140
    },
    {
      "entropy": 5.923260879516602,
      "epoch": 0.31212782850337434,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499487243707501,
      "loss": 5.826,
      "mean_token_accuracy": 0.13392639383673668,
      "num_tokens": 6869940.0,
      "step": 3145
    },
    {
      "entropy": 5.91810302734375,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.764,
      "mean_token_accuracy": 0.13164957985281944,
      "num_tokens": 6882161.0,
      "step": 3150
    },
    {
      "entropy": 6.03770079612732,
      "epoch": 0.3131202858277094,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.7941,
      "mean_token_accuracy": 0.13264409452676773,
      "num_tokens": 6893236.0,
      "step": 3155
    },
    {
      "entropy": 5.910415363311768,
      "epoch": 0.31361651448987693,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.7816,
      "mean_token_accuracy": 0.1374106653034687,
      "num_tokens": 6904934.0,
      "step": 3160
    },
    {
      "entropy": 5.9100690364837645,
      "epoch": 0.31411274315204446,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.7923,
      "mean_token_accuracy": 0.13251397609710694,
      "num_tokens": 6917537.0,
      "step": 3165
    },
    {
      "entropy": 5.993865346908569,
      "epoch": 0.314608971814212,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.838,
      "mean_token_accuracy": 0.13407180681824685,
      "num_tokens": 6928354.0,
      "step": 3170
    },
    {
      "entropy": 5.871166944503784,
      "epoch": 0.3151052004763795,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499472799436847,
      "loss": 5.7249,
      "mean_token_accuracy": 0.1391881600022316,
      "num_tokens": 6939553.0,
      "step": 3175
    },
    {
      "entropy": 5.898001718521118,
      "epoch": 0.31560142913854705,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.7286,
      "mean_token_accuracy": 0.13632505759596825,
      "num_tokens": 6950233.0,
      "step": 3180
    },
    {
      "entropy": 5.92787389755249,
      "epoch": 0.3160976578007146,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.7486,
      "mean_token_accuracy": 0.1380390614271164,
      "num_tokens": 6960716.0,
      "step": 3185
    },
    {
      "entropy": 5.9155830383300785,
      "epoch": 0.3165938864628821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.7098,
      "mean_token_accuracy": 0.13736789897084237,
      "num_tokens": 6970391.0,
      "step": 3190
    },
    {
      "entropy": 5.89755482673645,
      "epoch": 0.31709011512504964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.7107,
      "mean_token_accuracy": 0.13658948317170144,
      "num_tokens": 6981067.0,
      "step": 3195
    },
    {
      "entropy": 5.932046556472779,
      "epoch": 0.3175863437872172,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.7895,
      "mean_token_accuracy": 0.13591778874397278,
      "num_tokens": 6992000.0,
      "step": 3200
    },
    {
      "entropy": 5.925393962860108,
      "epoch": 0.3180825724493847,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.742,
      "mean_token_accuracy": 0.13471074998378754,
      "num_tokens": 7001791.0,
      "step": 3205
    },
    {
      "entropy": 5.935115385055542,
      "epoch": 0.3185788011115522,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.7841,
      "mean_token_accuracy": 0.13461261466145516,
      "num_tokens": 7011635.0,
      "step": 3210
    },
    {
      "entropy": 5.932962036132812,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.7975,
      "mean_token_accuracy": 0.13155610263347625,
      "num_tokens": 7023011.0,
      "step": 3215
    },
    {
      "entropy": 5.874264287948608,
      "epoch": 0.31957125843588724,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499450757300155,
      "loss": 5.723,
      "mean_token_accuracy": 0.13323820382356644,
      "num_tokens": 7033756.0,
      "step": 3220
    },
    {
      "entropy": 5.907634115219116,
      "epoch": 0.32006748709805477,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.7964,
      "mean_token_accuracy": 0.12840933203697205,
      "num_tokens": 7045038.0,
      "step": 3225
    },
    {
      "entropy": 5.982826614379883,
      "epoch": 0.3205637157602223,
      "grad_norm": 0.78125,
      "learning_rate": 0.000499445797821144,
      "loss": 5.7881,
      "mean_token_accuracy": 0.1322704665362835,
      "num_tokens": 7057181.0,
      "step": 3230
    },
    {
      "entropy": 5.922992849349976,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.8914,
      "mean_token_accuracy": 0.12643930315971375,
      "num_tokens": 7069441.0,
      "step": 3235
    },
    {
      "entropy": 5.9271252155303955,
      "epoch": 0.32155617308455736,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.7686,
      "mean_token_accuracy": 0.13725759983062744,
      "num_tokens": 7080017.0,
      "step": 3240
    },
    {
      "entropy": 5.934252309799194,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004994383168594907,
      "loss": 5.786,
      "mean_token_accuracy": 0.13711378425359727,
      "num_tokens": 7091034.0,
      "step": 3245
    },
    {
      "entropy": 5.9433753967285154,
      "epoch": 0.3225486304088924,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.7248,
      "mean_token_accuracy": 0.13346370235085486,
      "num_tokens": 7101227.0,
      "step": 3250
    },
    {
      "entropy": 5.873114490509034,
      "epoch": 0.32304485907105995,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.7232,
      "mean_token_accuracy": 0.1347164236009121,
      "num_tokens": 7112056.0,
      "step": 3255
    },
    {
      "entropy": 5.872904634475708,
      "epoch": 0.3235410877332275,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.7155,
      "mean_token_accuracy": 0.13577598035335542,
      "num_tokens": 7121895.0,
      "step": 3260
    },
    {
      "entropy": 5.920404434204102,
      "epoch": 0.324037316395395,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.6352,
      "mean_token_accuracy": 0.1344951681792736,
      "num_tokens": 7132726.0,
      "step": 3265
    },
    {
      "entropy": 5.896666431427002,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.7184,
      "mean_token_accuracy": 0.13523210138082503,
      "num_tokens": 7143394.0,
      "step": 3270
    },
    {
      "entropy": 5.85381326675415,
      "epoch": 0.3250297737197301,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.7219,
      "mean_token_accuracy": 0.13861265704035758,
      "num_tokens": 7154416.0,
      "step": 3275
    },
    {
      "entropy": 5.932457399368286,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.8036,
      "mean_token_accuracy": 0.13475491479039192,
      "num_tokens": 7164876.0,
      "step": 3280
    },
    {
      "entropy": 5.93551778793335,
      "epoch": 0.3260222310440651,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.6914,
      "mean_token_accuracy": 0.1398068495094776,
      "num_tokens": 7176053.0,
      "step": 3285
    },
    {
      "entropy": 5.8336615562438965,
      "epoch": 0.3265184597062326,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.8264,
      "mean_token_accuracy": 0.12871640026569367,
      "num_tokens": 7188180.0,
      "step": 3290
    },
    {
      "entropy": 6.0211320400238035,
      "epoch": 0.32701468836840014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499413018566984,
      "loss": 5.8583,
      "mean_token_accuracy": 0.13153597712516785,
      "num_tokens": 7198012.0,
      "step": 3295
    },
    {
      "entropy": 5.8717607975006105,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.6924,
      "mean_token_accuracy": 0.13936394453048706,
      "num_tokens": 7209465.0,
      "step": 3300
    },
    {
      "entropy": 5.9295085906982425,
      "epoch": 0.3280071456927352,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.7548,
      "mean_token_accuracy": 0.12718145027756692,
      "num_tokens": 7220546.0,
      "step": 3305
    },
    {
      "entropy": 5.893476819992065,
      "epoch": 0.32850337435490273,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.7882,
      "mean_token_accuracy": 0.13318324387073516,
      "num_tokens": 7232870.0,
      "step": 3310
    },
    {
      "entropy": 5.851389217376709,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.6709,
      "mean_token_accuracy": 0.14008632600307463,
      "num_tokens": 7242052.0,
      "step": 3315
    },
    {
      "entropy": 5.8930370807647705,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.7238,
      "mean_token_accuracy": 0.13118343874812127,
      "num_tokens": 7253060.0,
      "step": 3320
    },
    {
      "entropy": 5.979122447967529,
      "epoch": 0.3299920603414053,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499397572469416,
      "loss": 5.8563,
      "mean_token_accuracy": 0.1265694111585617,
      "num_tokens": 7264034.0,
      "step": 3325
    },
    {
      "entropy": 5.982485628128051,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.7816,
      "mean_token_accuracy": 0.13368444442749022,
      "num_tokens": 7275865.0,
      "step": 3330
    },
    {
      "entropy": 5.857488298416138,
      "epoch": 0.3309845176657404,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.6519,
      "mean_token_accuracy": 0.13691321164369583,
      "num_tokens": 7286652.0,
      "step": 3335
    },
    {
      "entropy": 5.8878926753997805,
      "epoch": 0.3314807463279079,
      "grad_norm": 0.73828125,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.8143,
      "mean_token_accuracy": 0.13285368159413338,
      "num_tokens": 7298032.0,
      "step": 3340
    },
    {
      "entropy": 5.902023363113403,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.7352,
      "mean_token_accuracy": 0.13993850946426392,
      "num_tokens": 7310883.0,
      "step": 3345
    },
    {
      "entropy": 5.913423824310303,
      "epoch": 0.332473203652243,
      "grad_norm": 0.875,
      "learning_rate": 0.000499384547695016,
      "loss": 5.7407,
      "mean_token_accuracy": 0.13315809816122054,
      "num_tokens": 7320934.0,
      "step": 3350
    },
    {
      "entropy": 5.918759393692016,
      "epoch": 0.3329694323144105,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499381926047117,
      "loss": 5.7737,
      "mean_token_accuracy": 0.1331883378326893,
      "num_tokens": 7331137.0,
      "step": 3355
    },
    {
      "entropy": 5.946372222900391,
      "epoch": 0.333465660976578,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049937929883503,
      "loss": 5.791,
      "mean_token_accuracy": 0.13182156682014465,
      "num_tokens": 7341697.0,
      "step": 3360
    },
    {
      "entropy": 5.8814027309417725,
      "epoch": 0.3339618896387455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.7106,
      "mean_token_accuracy": 0.13688041716814042,
      "num_tokens": 7352952.0,
      "step": 3365
    },
    {
      "entropy": 5.900327491760254,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993740277185515,
      "loss": 5.699,
      "mean_token_accuracy": 0.13415437415242196,
      "num_tokens": 7362904.0,
      "step": 3370
    },
    {
      "entropy": 5.974015378952027,
      "epoch": 0.3349543469630806,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499371383814291,
      "loss": 5.7572,
      "mean_token_accuracy": 0.13363182172179222,
      "num_tokens": 7372457.0,
      "step": 3375
    },
    {
      "entropy": 5.879163360595703,
      "epoch": 0.3354505756252481,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.7942,
      "mean_token_accuracy": 0.13216106370091438,
      "num_tokens": 7383732.0,
      "step": 3380
    },
    {
      "entropy": 5.91751651763916,
      "epoch": 0.33594680428741563,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.6585,
      "mean_token_accuracy": 0.1440861202776432,
      "num_tokens": 7395985.0,
      "step": 3385
    },
    {
      "entropy": 5.824702167510987,
      "epoch": 0.33644303294958317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.6436,
      "mean_token_accuracy": 0.13681414276361464,
      "num_tokens": 7405508.0,
      "step": 3390
    },
    {
      "entropy": 5.87082462310791,
      "epoch": 0.3369392616117507,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.7574,
      "mean_token_accuracy": 0.1357421875,
      "num_tokens": 7416369.0,
      "step": 3395
    },
    {
      "entropy": 5.933330392837524,
      "epoch": 0.3374354902739182,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.8055,
      "mean_token_accuracy": 0.13527968525886536,
      "num_tokens": 7427357.0,
      "step": 3400
    },
    {
      "entropy": 6.003190374374389,
      "epoch": 0.33793171893608576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.694,
      "mean_token_accuracy": 0.13455158174037934,
      "num_tokens": 7438019.0,
      "step": 3405
    },
    {
      "entropy": 5.908734464645386,
      "epoch": 0.3384279475982533,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.776,
      "mean_token_accuracy": 0.13500426560640336,
      "num_tokens": 7449063.0,
      "step": 3410
    },
    {
      "entropy": 5.862820768356324,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.6714,
      "mean_token_accuracy": 0.14003854468464852,
      "num_tokens": 7459731.0,
      "step": 3415
    },
    {
      "entropy": 5.975004434585571,
      "epoch": 0.33942040492258835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.6974,
      "mean_token_accuracy": 0.13529627472162248,
      "num_tokens": 7470763.0,
      "step": 3420
    },
    {
      "entropy": 5.805208539962768,
      "epoch": 0.3399166335847559,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.6798,
      "mean_token_accuracy": 0.13901263251900672,
      "num_tokens": 7482281.0,
      "step": 3425
    },
    {
      "entropy": 5.8485414505004885,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.7581,
      "mean_token_accuracy": 0.12874139919877053,
      "num_tokens": 7494622.0,
      "step": 3430
    },
    {
      "entropy": 5.909423160552978,
      "epoch": 0.3409090909090909,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.6825,
      "mean_token_accuracy": 0.14152807071805001,
      "num_tokens": 7505554.0,
      "step": 3435
    },
    {
      "entropy": 5.9753546714782715,
      "epoch": 0.3414053195712584,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.7596,
      "mean_token_accuracy": 0.14017617255449294,
      "num_tokens": 7517308.0,
      "step": 3440
    },
    {
      "entropy": 5.897326946258545,
      "epoch": 0.34190154823342594,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.8295,
      "mean_token_accuracy": 0.13342875316739083,
      "num_tokens": 7527937.0,
      "step": 3445
    },
    {
      "entropy": 5.904563474655151,
      "epoch": 0.3423977768955935,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.7335,
      "mean_token_accuracy": 0.13155163526535035,
      "num_tokens": 7538987.0,
      "step": 3450
    },
    {
      "entropy": 5.830388879776001,
      "epoch": 0.342894005557761,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.6903,
      "mean_token_accuracy": 0.13643986284732817,
      "num_tokens": 7550526.0,
      "step": 3455
    },
    {
      "entropy": 5.86571478843689,
      "epoch": 0.34339023421992854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.7248,
      "mean_token_accuracy": 0.14177241623401643,
      "num_tokens": 7561705.0,
      "step": 3460
    },
    {
      "entropy": 5.942563199996949,
      "epoch": 0.34388646288209607,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.787,
      "mean_token_accuracy": 0.12768477946519852,
      "num_tokens": 7573789.0,
      "step": 3465
    },
    {
      "entropy": 5.850177812576294,
      "epoch": 0.3443826915442636,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.6431,
      "mean_token_accuracy": 0.13789672404527664,
      "num_tokens": 7585516.0,
      "step": 3470
    },
    {
      "entropy": 5.949050569534302,
      "epoch": 0.3448789202064311,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499317337392652,
      "loss": 5.7788,
      "mean_token_accuracy": 0.1318759500980377,
      "num_tokens": 7596324.0,
      "step": 3475
    },
    {
      "entropy": 5.931936025619507,
      "epoch": 0.34537514886859866,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.7976,
      "mean_token_accuracy": 0.1297684609889984,
      "num_tokens": 7608103.0,
      "step": 3480
    },
    {
      "entropy": 5.882292366027832,
      "epoch": 0.3458713775307662,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.6749,
      "mean_token_accuracy": 0.13691046610474586,
      "num_tokens": 7619442.0,
      "step": 3485
    },
    {
      "entropy": 5.89331841468811,
      "epoch": 0.3463676061929337,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.7491,
      "mean_token_accuracy": 0.13759796395897866,
      "num_tokens": 7630066.0,
      "step": 3490
    },
    {
      "entropy": 5.807013273239136,
      "epoch": 0.34686383485510125,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1418273150920868,
      "num_tokens": 7641209.0,
      "step": 3495
    },
    {
      "entropy": 5.771910238265991,
      "epoch": 0.3473600635172688,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.6943,
      "mean_token_accuracy": 0.13641476407647132,
      "num_tokens": 7652957.0,
      "step": 3500
    },
    {
      "entropy": 5.886564254760742,
      "epoch": 0.3478562921794363,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499300689559793,
      "loss": 5.6951,
      "mean_token_accuracy": 0.13493456095457076,
      "num_tokens": 7664538.0,
      "step": 3505
    },
    {
      "entropy": 5.894466733932495,
      "epoch": 0.3483525208416038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.7154,
      "mean_token_accuracy": 0.13852872252464293,
      "num_tokens": 7674596.0,
      "step": 3510
    },
    {
      "entropy": 5.832696151733399,
      "epoch": 0.3488487495037713,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.6923,
      "mean_token_accuracy": 0.13129882737994195,
      "num_tokens": 7685496.0,
      "step": 3515
    },
    {
      "entropy": 5.940644359588623,
      "epoch": 0.34934497816593885,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499292290553087,
      "loss": 5.7235,
      "mean_token_accuracy": 0.1340453654527664,
      "num_tokens": 7696202.0,
      "step": 3520
    },
    {
      "entropy": 5.822767925262451,
      "epoch": 0.3498412068281064,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.6428,
      "mean_token_accuracy": 0.1384785458445549,
      "num_tokens": 7706570.0,
      "step": 3525
    },
    {
      "entropy": 5.836594915390014,
      "epoch": 0.3503374354902739,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049928666340534,
      "loss": 5.6493,
      "mean_token_accuracy": 0.14206060022115707,
      "num_tokens": 7716903.0,
      "step": 3530
    },
    {
      "entropy": 5.900732946395874,
      "epoch": 0.35083366415244144,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.7135,
      "mean_token_accuracy": 0.13520229384303092,
      "num_tokens": 7727059.0,
      "step": 3535
    },
    {
      "entropy": 5.932958459854126,
      "epoch": 0.35132989281460897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.8455,
      "mean_token_accuracy": 0.1343632586300373,
      "num_tokens": 7739089.0,
      "step": 3540
    },
    {
      "entropy": 5.956815576553344,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.75,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.7426,
      "mean_token_accuracy": 0.13446101769804955,
      "num_tokens": 7750953.0,
      "step": 3545
    },
    {
      "entropy": 5.853573226928711,
      "epoch": 0.35232235013894403,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.6991,
      "mean_token_accuracy": 0.13454861119389533,
      "num_tokens": 7761535.0,
      "step": 3550
    },
    {
      "entropy": 5.976637363433838,
      "epoch": 0.35281857880111156,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.8349,
      "mean_token_accuracy": 0.12968518286943437,
      "num_tokens": 7773497.0,
      "step": 3555
    },
    {
      "entropy": 5.816864728927612,
      "epoch": 0.3533148074632791,
      "grad_norm": 0.7421875,
      "learning_rate": 0.00049926964848052,
      "loss": 5.7317,
      "mean_token_accuracy": 0.13604558780789375,
      "num_tokens": 7785764.0,
      "step": 3560
    },
    {
      "entropy": 5.912026214599609,
      "epoch": 0.3538110361254466,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.6429,
      "mean_token_accuracy": 0.13882051557302474,
      "num_tokens": 7797231.0,
      "step": 3565
    },
    {
      "entropy": 5.8796247959136965,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.7008,
      "mean_token_accuracy": 0.13108594492077827,
      "num_tokens": 7809218.0,
      "step": 3570
    },
    {
      "entropy": 5.869585752487183,
      "epoch": 0.3548034934497817,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499261065938205,
      "loss": 5.7741,
      "mean_token_accuracy": 0.13558753281831742,
      "num_tokens": 7820418.0,
      "step": 3575
    },
    {
      "entropy": 5.845438528060913,
      "epoch": 0.3552997221119492,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.6412,
      "mean_token_accuracy": 0.14093869253993035,
      "num_tokens": 7831034.0,
      "step": 3580
    },
    {
      "entropy": 5.7888490676879885,
      "epoch": 0.3557959507741167,
      "grad_norm": 0.796875,
      "learning_rate": 0.000499255316437254,
      "loss": 5.6844,
      "mean_token_accuracy": 0.13652455508708955,
      "num_tokens": 7843703.0,
      "step": 3585
    },
    {
      "entropy": 5.893563890457154,
      "epoch": 0.3562921794362842,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.6934,
      "mean_token_accuracy": 0.13805954903364182,
      "num_tokens": 7855806.0,
      "step": 3590
    },
    {
      "entropy": 5.932254695892334,
      "epoch": 0.35678840809845175,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.8068,
      "mean_token_accuracy": 0.13560713455080986,
      "num_tokens": 7868638.0,
      "step": 3595
    },
    {
      "entropy": 5.898502588272095,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.6497,
      "mean_token_accuracy": 0.13955551013350487,
      "num_tokens": 7880269.0,
      "step": 3600
    },
    {
      "entropy": 5.7724402904510494,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.6167,
      "mean_token_accuracy": 0.13869745060801505,
      "num_tokens": 7889870.0,
      "step": 3605
    },
    {
      "entropy": 5.935569477081299,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.8055,
      "mean_token_accuracy": 0.13309841975569725,
      "num_tokens": 7902025.0,
      "step": 3610
    },
    {
      "entropy": 5.7609467029571535,
      "epoch": 0.35877332274712187,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.591,
      "mean_token_accuracy": 0.13937846422195435,
      "num_tokens": 7912994.0,
      "step": 3615
    },
    {
      "entropy": 5.793561744689941,
      "epoch": 0.3592695514092894,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.676,
      "mean_token_accuracy": 0.14364308044314383,
      "num_tokens": 7924184.0,
      "step": 3620
    },
    {
      "entropy": 5.845511150360108,
      "epoch": 0.35976578007145693,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.6462,
      "mean_token_accuracy": 0.1384884387254715,
      "num_tokens": 7935394.0,
      "step": 3625
    },
    {
      "entropy": 5.895503234863281,
      "epoch": 0.36026200873362446,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.7331,
      "mean_token_accuracy": 0.1407325714826584,
      "num_tokens": 7946699.0,
      "step": 3630
    },
    {
      "entropy": 5.831504869461059,
      "epoch": 0.360758237395792,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.6792,
      "mean_token_accuracy": 0.14085863083600997,
      "num_tokens": 7957362.0,
      "step": 3635
    },
    {
      "entropy": 5.8577423095703125,
      "epoch": 0.3612544660579595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.6682,
      "mean_token_accuracy": 0.13818608224391937,
      "num_tokens": 7967316.0,
      "step": 3640
    },
    {
      "entropy": 5.804681730270386,
      "epoch": 0.36175069472012705,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000499220352323582,
      "loss": 5.6261,
      "mean_token_accuracy": 0.14074852019548417,
      "num_tokens": 7978163.0,
      "step": 3645
    },
    {
      "entropy": 5.837596464157104,
      "epoch": 0.3622469233822946,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.6426,
      "mean_token_accuracy": 0.13854158371686937,
      "num_tokens": 7988441.0,
      "step": 3650
    },
    {
      "entropy": 5.840954446792603,
      "epoch": 0.3627431520444621,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.6358,
      "mean_token_accuracy": 0.13985022827982901,
      "num_tokens": 7999779.0,
      "step": 3655
    },
    {
      "entropy": 5.845841884613037,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.7877,
      "mean_token_accuracy": 0.13552388399839402,
      "num_tokens": 8010921.0,
      "step": 3660
    },
    {
      "entropy": 5.902638053894043,
      "epoch": 0.3637356093687971,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.7677,
      "mean_token_accuracy": 0.1313785932958126,
      "num_tokens": 8020918.0,
      "step": 3665
    },
    {
      "entropy": 5.907895040512085,
      "epoch": 0.36423183803096465,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.7895,
      "mean_token_accuracy": 0.13416459634900094,
      "num_tokens": 8032238.0,
      "step": 3670
    },
    {
      "entropy": 5.841483306884766,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.6308,
      "mean_token_accuracy": 0.14457927644252777,
      "num_tokens": 8043288.0,
      "step": 3675
    },
    {
      "entropy": 5.761933994293213,
      "epoch": 0.3652242953552997,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.6094,
      "mean_token_accuracy": 0.14315860122442245,
      "num_tokens": 8053356.0,
      "step": 3680
    },
    {
      "entropy": 5.871804332733154,
      "epoch": 0.36572052401746724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.6204,
      "mean_token_accuracy": 0.14282752126455306,
      "num_tokens": 8062641.0,
      "step": 3685
    },
    {
      "entropy": 5.896347284317017,
      "epoch": 0.36621675267963477,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.7859,
      "mean_token_accuracy": 0.13109915107488632,
      "num_tokens": 8074232.0,
      "step": 3690
    },
    {
      "entropy": 5.893853664398193,
      "epoch": 0.3667129813418023,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.6442,
      "mean_token_accuracy": 0.14742069989442824,
      "num_tokens": 8085276.0,
      "step": 3695
    },
    {
      "entropy": 5.847226762771607,
      "epoch": 0.36720921000396983,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.6847,
      "mean_token_accuracy": 0.1360927253961563,
      "num_tokens": 8095769.0,
      "step": 3700
    },
    {
      "entropy": 5.8576277732849125,
      "epoch": 0.36770543866613736,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000499184587543998,
      "loss": 5.6272,
      "mean_token_accuracy": 0.13834724947810173,
      "num_tokens": 8106932.0,
      "step": 3705
    },
    {
      "entropy": 5.8724089622497555,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.7313,
      "mean_token_accuracy": 0.14015434086322784,
      "num_tokens": 8118307.0,
      "step": 3710
    },
    {
      "entropy": 5.817163324356079,
      "epoch": 0.3686978959904724,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.664,
      "mean_token_accuracy": 0.13858182430267335,
      "num_tokens": 8129940.0,
      "step": 3715
    },
    {
      "entropy": 5.904032135009766,
      "epoch": 0.36919412465263995,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.7354,
      "mean_token_accuracy": 0.13806151151657103,
      "num_tokens": 8141362.0,
      "step": 3720
    },
    {
      "entropy": 5.8136218070983885,
      "epoch": 0.3696903533148075,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.6559,
      "mean_token_accuracy": 0.13952167332172394,
      "num_tokens": 8151795.0,
      "step": 3725
    },
    {
      "entropy": 5.833437395095825,
      "epoch": 0.370186581976975,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.6835,
      "mean_token_accuracy": 0.13554125279188156,
      "num_tokens": 8162490.0,
      "step": 3730
    },
    {
      "entropy": 5.8535802364349365,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.6551,
      "mean_token_accuracy": 0.13590577319264413,
      "num_tokens": 8173546.0,
      "step": 3735
    },
    {
      "entropy": 5.877345323562622,
      "epoch": 0.37117903930131,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.7166,
      "mean_token_accuracy": 0.13483782708644867,
      "num_tokens": 8184079.0,
      "step": 3740
    },
    {
      "entropy": 5.866656112670898,
      "epoch": 0.37167526796347755,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.7507,
      "mean_token_accuracy": 0.13449111357331275,
      "num_tokens": 8195603.0,
      "step": 3745
    },
    {
      "entropy": 5.88632869720459,
      "epoch": 0.3721714966256451,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.685,
      "mean_token_accuracy": 0.13895501494407653,
      "num_tokens": 8206605.0,
      "step": 3750
    },
    {
      "entropy": 5.855900430679322,
      "epoch": 0.3726677252878126,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.7432,
      "mean_token_accuracy": 0.1339529387652874,
      "num_tokens": 8216950.0,
      "step": 3755
    },
    {
      "entropy": 5.8636352062225345,
      "epoch": 0.37316395394998014,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.7446,
      "mean_token_accuracy": 0.1312355987727642,
      "num_tokens": 8227272.0,
      "step": 3760
    },
    {
      "entropy": 5.842629051208496,
      "epoch": 0.3736601826121477,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.7095,
      "mean_token_accuracy": 0.14670437723398208,
      "num_tokens": 8238062.0,
      "step": 3765
    },
    {
      "entropy": 5.841026878356933,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.6875,
      "mean_token_accuracy": 0.1373550184071064,
      "num_tokens": 8248851.0,
      "step": 3770
    },
    {
      "entropy": 5.925320720672607,
      "epoch": 0.37465263993648273,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.7624,
      "mean_token_accuracy": 0.13524728193879126,
      "num_tokens": 8258535.0,
      "step": 3775
    },
    {
      "entropy": 5.830336427688598,
      "epoch": 0.37514886859865026,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.6791,
      "mean_token_accuracy": 0.13500638157129288,
      "num_tokens": 8271666.0,
      "step": 3780
    },
    {
      "entropy": 5.8889875411987305,
      "epoch": 0.3756450972608178,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.7611,
      "mean_token_accuracy": 0.13629817590117455,
      "num_tokens": 8281869.0,
      "step": 3785
    },
    {
      "entropy": 5.881752634048462,
      "epoch": 0.3761413259229853,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.7306,
      "mean_token_accuracy": 0.13635860234498978,
      "num_tokens": 8292699.0,
      "step": 3790
    },
    {
      "entropy": 5.924288940429688,
      "epoch": 0.37663755458515286,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.6501,
      "mean_token_accuracy": 0.14117683693766594,
      "num_tokens": 8302832.0,
      "step": 3795
    },
    {
      "entropy": 5.900856876373291,
      "epoch": 0.3771337832473204,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.7329,
      "mean_token_accuracy": 0.1359848089516163,
      "num_tokens": 8314721.0,
      "step": 3800
    },
    {
      "entropy": 5.7943638324737545,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.6135,
      "mean_token_accuracy": 0.13641808331012725,
      "num_tokens": 8325140.0,
      "step": 3805
    },
    {
      "entropy": 5.847335720062256,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.6786,
      "mean_token_accuracy": 0.13666619807481767,
      "num_tokens": 8336509.0,
      "step": 3810
    },
    {
      "entropy": 5.816822910308838,
      "epoch": 0.3786224692338229,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.6791,
      "mean_token_accuracy": 0.1448786400258541,
      "num_tokens": 8346999.0,
      "step": 3815
    },
    {
      "entropy": 5.804215717315674,
      "epoch": 0.37911869789599045,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.6065,
      "mean_token_accuracy": 0.14424843713641167,
      "num_tokens": 8358697.0,
      "step": 3820
    },
    {
      "entropy": 5.9240092754364015,
      "epoch": 0.379614926558158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.738,
      "mean_token_accuracy": 0.1382055662572384,
      "num_tokens": 8370365.0,
      "step": 3825
    },
    {
      "entropy": 5.840770530700683,
      "epoch": 0.3801111552203255,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.6484,
      "mean_token_accuracy": 0.13382984697818756,
      "num_tokens": 8381137.0,
      "step": 3830
    },
    {
      "entropy": 5.827268743515015,
      "epoch": 0.38060738388249304,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.7709,
      "mean_token_accuracy": 0.1432642236351967,
      "num_tokens": 8393299.0,
      "step": 3835
    },
    {
      "entropy": 5.86376986503601,
      "epoch": 0.3811036125446606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.7218,
      "mean_token_accuracy": 0.13187216296792031,
      "num_tokens": 8404564.0,
      "step": 3840
    },
    {
      "entropy": 5.913081359863281,
      "epoch": 0.3815998412068281,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.6714,
      "mean_token_accuracy": 0.14094971716403962,
      "num_tokens": 8413850.0,
      "step": 3845
    },
    {
      "entropy": 5.74772458076477,
      "epoch": 0.38209606986899564,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.6503,
      "mean_token_accuracy": 0.13791525065898896,
      "num_tokens": 8425226.0,
      "step": 3850
    },
    {
      "entropy": 5.919849824905396,
      "epoch": 0.38259229853116317,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.7484,
      "mean_token_accuracy": 0.1349697604775429,
      "num_tokens": 8435714.0,
      "step": 3855
    },
    {
      "entropy": 5.940012073516845,
      "epoch": 0.3830885271933307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.6681,
      "mean_token_accuracy": 0.14216530025005342,
      "num_tokens": 8446691.0,
      "step": 3860
    },
    {
      "entropy": 5.782909250259399,
      "epoch": 0.3835847558554982,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.7334,
      "mean_token_accuracy": 0.13524289354681968,
      "num_tokens": 8457376.0,
      "step": 3865
    },
    {
      "entropy": 5.810849523544311,
      "epoch": 0.38408098451766576,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.6733,
      "mean_token_accuracy": 0.13647696673870086,
      "num_tokens": 8469456.0,
      "step": 3870
    },
    {
      "entropy": 5.964277076721191,
      "epoch": 0.3845772131798333,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.7324,
      "mean_token_accuracy": 0.13757467567920684,
      "num_tokens": 8481142.0,
      "step": 3875
    },
    {
      "entropy": 5.8390617847442625,
      "epoch": 0.3850734418420008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.801,
      "mean_token_accuracy": 0.1280309207737446,
      "num_tokens": 8492900.0,
      "step": 3880
    },
    {
      "entropy": 5.85485520362854,
      "epoch": 0.3855696705041683,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1367727555334568,
      "num_tokens": 8503323.0,
      "step": 3885
    },
    {
      "entropy": 5.8329376697540285,
      "epoch": 0.3860658991663358,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499069273881985,
      "loss": 5.7132,
      "mean_token_accuracy": 0.1381821908056736,
      "num_tokens": 8514713.0,
      "step": 3890
    },
    {
      "entropy": 5.872621297836304,
      "epoch": 0.38656212782850335,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.5762,
      "mean_token_accuracy": 0.1428079128265381,
      "num_tokens": 8524907.0,
      "step": 3895
    },
    {
      "entropy": 5.7864421844482425,
      "epoch": 0.3870583564906709,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.5972,
      "mean_token_accuracy": 0.1435305170714855,
      "num_tokens": 8535769.0,
      "step": 3900
    },
    {
      "entropy": 5.8502112865448,
      "epoch": 0.3875545851528384,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.7434,
      "mean_token_accuracy": 0.1362551860511303,
      "num_tokens": 8547754.0,
      "step": 3905
    },
    {
      "entropy": 5.826516389846802,
      "epoch": 0.38805081381500595,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.6598,
      "mean_token_accuracy": 0.13282526582479476,
      "num_tokens": 8558896.0,
      "step": 3910
    },
    {
      "entropy": 5.792256498336792,
      "epoch": 0.3885470424771735,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.5748,
      "mean_token_accuracy": 0.14356540516018867,
      "num_tokens": 8568721.0,
      "step": 3915
    },
    {
      "entropy": 5.798781538009644,
      "epoch": 0.389043271139341,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.637,
      "mean_token_accuracy": 0.142483389377594,
      "num_tokens": 8579199.0,
      "step": 3920
    },
    {
      "entropy": 5.873562908172607,
      "epoch": 0.38953949980150854,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.7326,
      "mean_token_accuracy": 0.134691421687603,
      "num_tokens": 8590362.0,
      "step": 3925
    },
    {
      "entropy": 5.875561618804932,
      "epoch": 0.39003572846367607,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.6901,
      "mean_token_accuracy": 0.13582185059785842,
      "num_tokens": 8600603.0,
      "step": 3930
    },
    {
      "entropy": 5.833240270614624,
      "epoch": 0.3905319571258436,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.67,
      "mean_token_accuracy": 0.13707878515124322,
      "num_tokens": 8611436.0,
      "step": 3935
    },
    {
      "entropy": 5.858435773849488,
      "epoch": 0.39102818578801113,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.7585,
      "mean_token_accuracy": 0.1360964857041836,
      "num_tokens": 8624152.0,
      "step": 3940
    },
    {
      "entropy": 5.823321914672851,
      "epoch": 0.39152441445017866,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499033524363663,
      "loss": 5.577,
      "mean_token_accuracy": 0.14084477797150613,
      "num_tokens": 8634762.0,
      "step": 3945
    },
    {
      "entropy": 5.818802070617676,
      "epoch": 0.3920206431123462,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.5812,
      "mean_token_accuracy": 0.1392807349562645,
      "num_tokens": 8646204.0,
      "step": 3950
    },
    {
      "entropy": 5.74844651222229,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.5979,
      "mean_token_accuracy": 0.1378256045281887,
      "num_tokens": 8656662.0,
      "step": 3955
    },
    {
      "entropy": 5.8304401397705075,
      "epoch": 0.3930131004366812,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.6249,
      "mean_token_accuracy": 0.1406857840716839,
      "num_tokens": 8668180.0,
      "step": 3960
    },
    {
      "entropy": 5.843784189224243,
      "epoch": 0.3935093290988487,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.6518,
      "mean_token_accuracy": 0.13405353501439093,
      "num_tokens": 8679522.0,
      "step": 3965
    },
    {
      "entropy": 5.778100538253784,
      "epoch": 0.39400555776101626,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.738,
      "mean_token_accuracy": 0.13023702427744865,
      "num_tokens": 8691451.0,
      "step": 3970
    },
    {
      "entropy": 5.80493130683899,
      "epoch": 0.3945017864231838,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.6063,
      "mean_token_accuracy": 0.13593686521053314,
      "num_tokens": 8702437.0,
      "step": 3975
    },
    {
      "entropy": 5.812478160858154,
      "epoch": 0.3949980150853513,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.6323,
      "mean_token_accuracy": 0.13778943568468094,
      "num_tokens": 8714612.0,
      "step": 3980
    },
    {
      "entropy": 5.792282056808472,
      "epoch": 0.39549424374751885,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.6067,
      "mean_token_accuracy": 0.1396419808268547,
      "num_tokens": 8725065.0,
      "step": 3985
    },
    {
      "entropy": 5.761460590362549,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.6062,
      "mean_token_accuracy": 0.14276854917407036,
      "num_tokens": 8736142.0,
      "step": 3990
    },
    {
      "entropy": 5.891938066482544,
      "epoch": 0.3964867010718539,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.7289,
      "mean_token_accuracy": 0.13371312618255615,
      "num_tokens": 8746708.0,
      "step": 3995
    },
    {
      "entropy": 5.889298772811889,
      "epoch": 0.39698292973402144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.6268,
      "mean_token_accuracy": 0.1416555315256119,
      "num_tokens": 8756756.0,
      "step": 4000
    },
    {
      "entropy": 5.7893242835998535,
      "epoch": 0.39747915839618897,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.6757,
      "mean_token_accuracy": 0.14008952751755716,
      "num_tokens": 8767151.0,
      "step": 4005
    },
    {
      "entropy": 5.82853102684021,
      "epoch": 0.3979753870583565,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.7618,
      "mean_token_accuracy": 0.13732962384819986,
      "num_tokens": 8778169.0,
      "step": 4010
    },
    {
      "entropy": 5.875279140472412,
      "epoch": 0.39847161572052403,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000498987052770255,
      "loss": 5.6361,
      "mean_token_accuracy": 0.13776904642581939,
      "num_tokens": 8788981.0,
      "step": 4015
    },
    {
      "entropy": 5.830997705459595,
      "epoch": 0.39896784438269156,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498983691710063,
      "loss": 5.6457,
      "mean_token_accuracy": 0.14300358146429062,
      "num_tokens": 8799190.0,
      "step": 4020
    },
    {
      "entropy": 5.838423442840576,
      "epoch": 0.3994640730448591,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000498980325095558,
      "loss": 5.6262,
      "mean_token_accuracy": 0.14072234630584718,
      "num_tokens": 8810030.0,
      "step": 4025
    },
    {
      "entropy": 5.744981288909912,
      "epoch": 0.3999603017070266,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.6429,
      "mean_token_accuracy": 0.13680833503603934,
      "num_tokens": 8820715.0,
      "step": 4030
    },
    {
      "entropy": 5.834112310409546,
      "epoch": 0.4004565303691941,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.6514,
      "mean_token_accuracy": 0.1345742605626583,
      "num_tokens": 8830484.0,
      "step": 4035
    },
    {
      "entropy": 5.819972562789917,
      "epoch": 0.4009527590313616,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.6524,
      "mean_token_accuracy": 0.1366778366267681,
      "num_tokens": 8840732.0,
      "step": 4040
    },
    {
      "entropy": 5.842714500427246,
      "epoch": 0.40144898769352916,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.6617,
      "mean_token_accuracy": 0.13634128496050835,
      "num_tokens": 8853334.0,
      "step": 4045
    },
    {
      "entropy": 5.844744253158569,
      "epoch": 0.4019452163556967,
      "grad_norm": 1.0,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.6661,
      "mean_token_accuracy": 0.13867569118738174,
      "num_tokens": 8864129.0,
      "step": 4050
    },
    {
      "entropy": 5.7935614585876465,
      "epoch": 0.4024414450178642,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.6391,
      "mean_token_accuracy": 0.14020620808005332,
      "num_tokens": 8875353.0,
      "step": 4055
    },
    {
      "entropy": 5.8473481178283695,
      "epoch": 0.40293767368003175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.6895,
      "mean_token_accuracy": 0.14006996676325797,
      "num_tokens": 8886933.0,
      "step": 4060
    },
    {
      "entropy": 5.863424253463745,
      "epoch": 0.4034339023421993,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.6463,
      "mean_token_accuracy": 0.14140839651226997,
      "num_tokens": 8898465.0,
      "step": 4065
    },
    {
      "entropy": 5.7452678203582765,
      "epoch": 0.4039301310043668,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.5909,
      "mean_token_accuracy": 0.14476610869169235,
      "num_tokens": 8909072.0,
      "step": 4070
    },
    {
      "entropy": 5.798135614395141,
      "epoch": 0.40442635966653434,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.676,
      "mean_token_accuracy": 0.1369663007557392,
      "num_tokens": 8918841.0,
      "step": 4075
    },
    {
      "entropy": 5.896200180053711,
      "epoch": 0.40492258832870187,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.7004,
      "mean_token_accuracy": 0.1366220660507679,
      "num_tokens": 8929718.0,
      "step": 4080
    },
    {
      "entropy": 5.751429414749145,
      "epoch": 0.4054188169908694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.5918,
      "mean_token_accuracy": 0.14551517590880395,
      "num_tokens": 8938737.0,
      "step": 4085
    },
    {
      "entropy": 5.850674438476562,
      "epoch": 0.40591504565303693,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.7246,
      "mean_token_accuracy": 0.14013531878590585,
      "num_tokens": 8949121.0,
      "step": 4090
    },
    {
      "entropy": 5.796566200256348,
      "epoch": 0.40641127431520446,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498932609336636,
      "loss": 5.5745,
      "mean_token_accuracy": 0.1416654735803604,
      "num_tokens": 8958585.0,
      "step": 4095
    },
    {
      "entropy": 5.800935840606689,
      "epoch": 0.406907502977372,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.61,
      "mean_token_accuracy": 0.13719116896390915,
      "num_tokens": 8968528.0,
      "step": 4100
    },
    {
      "entropy": 5.820908927917481,
      "epoch": 0.4074037316395395,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.6978,
      "mean_token_accuracy": 0.14038729965686797,
      "num_tokens": 8980870.0,
      "step": 4105
    },
    {
      "entropy": 5.830300569534302,
      "epoch": 0.40789996030170705,
      "grad_norm": 0.828125,
      "learning_rate": 0.00049892224292053,
      "loss": 5.6414,
      "mean_token_accuracy": 0.14067394584417342,
      "num_tokens": 8992147.0,
      "step": 4110
    },
    {
      "entropy": 5.808791971206665,
      "epoch": 0.40839618896387453,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.6292,
      "mean_token_accuracy": 0.1384664535522461,
      "num_tokens": 9003768.0,
      "step": 4115
    },
    {
      "entropy": 5.8365038394927975,
      "epoch": 0.40889241762604206,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498915304212371,
      "loss": 5.7334,
      "mean_token_accuracy": 0.13195103034377098,
      "num_tokens": 9017259.0,
      "step": 4120
    },
    {
      "entropy": 5.794817113876343,
      "epoch": 0.4093886462882096,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.6284,
      "mean_token_accuracy": 0.14268521964550018,
      "num_tokens": 9028696.0,
      "step": 4125
    },
    {
      "entropy": 5.810578775405884,
      "epoch": 0.4098848749503771,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.6533,
      "mean_token_accuracy": 0.1401242844760418,
      "num_tokens": 9040231.0,
      "step": 4130
    },
    {
      "entropy": 5.73698468208313,
      "epoch": 0.41038110361254465,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.5151,
      "mean_token_accuracy": 0.14004018604755403,
      "num_tokens": 9050254.0,
      "step": 4135
    },
    {
      "entropy": 5.789291048049927,
      "epoch": 0.4108773322747122,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.662,
      "mean_token_accuracy": 0.13600490391254424,
      "num_tokens": 9061046.0,
      "step": 4140
    },
    {
      "entropy": 5.8500120639801025,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.6697,
      "mean_token_accuracy": 0.13418863266706466,
      "num_tokens": 9071899.0,
      "step": 4145
    },
    {
      "entropy": 5.858259487152099,
      "epoch": 0.41186978959904724,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.6601,
      "mean_token_accuracy": 0.14277462810277938,
      "num_tokens": 9081781.0,
      "step": 4150
    },
    {
      "entropy": 5.799009466171265,
      "epoch": 0.4123660182612148,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.6478,
      "mean_token_accuracy": 0.13711302280426024,
      "num_tokens": 9092284.0,
      "step": 4155
    },
    {
      "entropy": 5.805934286117553,
      "epoch": 0.4128622469233823,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.6595,
      "mean_token_accuracy": 0.13749522119760513,
      "num_tokens": 9102031.0,
      "step": 4160
    },
    {
      "entropy": 5.835699653625488,
      "epoch": 0.41335847558554983,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.6291,
      "mean_token_accuracy": 0.13785969614982604,
      "num_tokens": 9113312.0,
      "step": 4165
    },
    {
      "entropy": 5.765735816955567,
      "epoch": 0.41385470424771736,
      "grad_norm": 0.84375,
      "learning_rate": 0.000498880277528397,
      "loss": 5.6663,
      "mean_token_accuracy": 0.14153355211019517,
      "num_tokens": 9124111.0,
      "step": 4170
    },
    {
      "entropy": 5.843842172622681,
      "epoch": 0.4143509329098849,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.6188,
      "mean_token_accuracy": 0.14437493830919265,
      "num_tokens": 9134463.0,
      "step": 4175
    },
    {
      "entropy": 5.913123416900635,
      "epoch": 0.4148471615720524,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.8069,
      "mean_token_accuracy": 0.13912550434470178,
      "num_tokens": 9146064.0,
      "step": 4180
    },
    {
      "entropy": 5.78485426902771,
      "epoch": 0.41534339023421996,
      "grad_norm": 0.828125,
      "learning_rate": 0.000498869661260144,
      "loss": 5.5761,
      "mean_token_accuracy": 0.14577325507998468,
      "num_tokens": 9156272.0,
      "step": 4185
    },
    {
      "entropy": 5.768794250488281,
      "epoch": 0.41583961889638743,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498866111400972,
      "loss": 5.6309,
      "mean_token_accuracy": 0.13804599717259408,
      "num_tokens": 9166477.0,
      "step": 4190
    },
    {
      "entropy": 5.872252798080444,
      "epoch": 0.41633584755855496,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.6541,
      "mean_token_accuracy": 0.1345357172191143,
      "num_tokens": 9178577.0,
      "step": 4195
    },
    {
      "entropy": 5.759488153457641,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.5901,
      "mean_token_accuracy": 0.14053094610571862,
      "num_tokens": 9188020.0,
      "step": 4200
    },
    {
      "entropy": 5.81153302192688,
      "epoch": 0.41732830488289,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.6989,
      "mean_token_accuracy": 0.13900597989559174,
      "num_tokens": 9198571.0,
      "step": 4205
    },
    {
      "entropy": 5.817718124389648,
      "epoch": 0.41782453354505755,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.6435,
      "mean_token_accuracy": 0.14240401610732079,
      "num_tokens": 9210515.0,
      "step": 4210
    },
    {
      "entropy": 5.751219367980957,
      "epoch": 0.4183207622072251,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.5458,
      "mean_token_accuracy": 0.15108796954154968,
      "num_tokens": 9220756.0,
      "step": 4215
    },
    {
      "entropy": 5.835984897613526,
      "epoch": 0.4188169908693926,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.6065,
      "mean_token_accuracy": 0.1427240900695324,
      "num_tokens": 9231713.0,
      "step": 4220
    },
    {
      "entropy": 5.815691089630127,
      "epoch": 0.41931321953156014,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.6373,
      "mean_token_accuracy": 0.13878222405910492,
      "num_tokens": 9242430.0,
      "step": 4225
    },
    {
      "entropy": 5.872125101089478,
      "epoch": 0.4198094481937277,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.7332,
      "mean_token_accuracy": 0.13745660558342934,
      "num_tokens": 9254202.0,
      "step": 4230
    },
    {
      "entropy": 5.760603618621826,
      "epoch": 0.4203056768558952,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.6494,
      "mean_token_accuracy": 0.1339892975986004,
      "num_tokens": 9266065.0,
      "step": 4235
    },
    {
      "entropy": 5.827087020874023,
      "epoch": 0.42080190551806274,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498830307497001,
      "loss": 5.6108,
      "mean_token_accuracy": 0.14749082624912263,
      "num_tokens": 9277022.0,
      "step": 4240
    },
    {
      "entropy": 5.818345737457276,
      "epoch": 0.42129813418023027,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498826696577332,
      "loss": 5.6476,
      "mean_token_accuracy": 0.14149464368820192,
      "num_tokens": 9287144.0,
      "step": 4245
    },
    {
      "entropy": 5.757974863052368,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.6034,
      "mean_token_accuracy": 0.14171627312898635,
      "num_tokens": 9298559.0,
      "step": 4250
    },
    {
      "entropy": 5.7662519931793215,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.6121,
      "mean_token_accuracy": 0.13865947276353835,
      "num_tokens": 9309275.0,
      "step": 4255
    },
    {
      "entropy": 5.735612678527832,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.5498,
      "mean_token_accuracy": 0.14119256883859635,
      "num_tokens": 9319442.0,
      "step": 4260
    },
    {
      "entropy": 5.712081670761108,
      "epoch": 0.42328304882890033,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000498812197395356,
      "loss": 5.5393,
      "mean_token_accuracy": 0.1511388584971428,
      "num_tokens": 9329762.0,
      "step": 4265
    },
    {
      "entropy": 5.928845739364624,
      "epoch": 0.42377927749106786,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.7577,
      "mean_token_accuracy": 0.1394852250814438,
      "num_tokens": 9340417.0,
      "step": 4270
    },
    {
      "entropy": 5.835004281997681,
      "epoch": 0.4242755061532354,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.5924,
      "mean_token_accuracy": 0.14120562374591827,
      "num_tokens": 9351538.0,
      "step": 4275
    },
    {
      "entropy": 5.6838963508605955,
      "epoch": 0.4247717348154029,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.539,
      "mean_token_accuracy": 0.1515072450041771,
      "num_tokens": 9360916.0,
      "step": 4280
    },
    {
      "entropy": 5.770059251785279,
      "epoch": 0.42526796347757045,
      "grad_norm": 0.828125,
      "learning_rate": 0.000498797609412419,
      "loss": 5.6463,
      "mean_token_accuracy": 0.13973931297659875,
      "num_tokens": 9372959.0,
      "step": 4285
    },
    {
      "entropy": 5.817070579528808,
      "epoch": 0.425764192139738,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.6419,
      "mean_token_accuracy": 0.14023181200027465,
      "num_tokens": 9383887.0,
      "step": 4290
    },
    {
      "entropy": 5.808628129959106,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.71875,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.6319,
      "mean_token_accuracy": 0.14302112758159638,
      "num_tokens": 9397735.0,
      "step": 4295
    },
    {
      "entropy": 5.730823183059693,
      "epoch": 0.42675664946407305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.488,
      "mean_token_accuracy": 0.14955226927995682,
      "num_tokens": 9407614.0,
      "step": 4300
    },
    {
      "entropy": 5.750316524505616,
      "epoch": 0.4272528781262406,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.6242,
      "mean_token_accuracy": 0.13437720984220505,
      "num_tokens": 9420077.0,
      "step": 4305
    },
    {
      "entropy": 5.7564538478851315,
      "epoch": 0.4277491067884081,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.5636,
      "mean_token_accuracy": 0.14153434187173844,
      "num_tokens": 9431719.0,
      "step": 4310
    },
    {
      "entropy": 5.7765909194946286,
      "epoch": 0.42824533545057564,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498775560948883,
      "loss": 5.5705,
      "mean_token_accuracy": 0.14171234220266343,
      "num_tokens": 9441462.0,
      "step": 4315
    },
    {
      "entropy": 5.755573511123657,
      "epoch": 0.42874156411274317,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.5835,
      "mean_token_accuracy": 0.14282924234867095,
      "num_tokens": 9451968.0,
      "step": 4320
    },
    {
      "entropy": 5.788213109970092,
      "epoch": 0.4292377927749107,
      "grad_norm": 0.7734375,
      "learning_rate": 0.00049876816706685,
      "loss": 5.6225,
      "mean_token_accuracy": 0.1437190592288971,
      "num_tokens": 9463643.0,
      "step": 4325
    },
    {
      "entropy": 5.862869071960449,
      "epoch": 0.42973402143707823,
      "grad_norm": 0.796875,
      "learning_rate": 0.000498764461802335,
      "loss": 5.658,
      "mean_token_accuracy": 0.13717727959156037,
      "num_tokens": 9476067.0,
      "step": 4330
    },
    {
      "entropy": 5.769567584991455,
      "epoch": 0.43023025009924576,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.5133,
      "mean_token_accuracy": 0.14948484748601915,
      "num_tokens": 9485697.0,
      "step": 4335
    },
    {
      "entropy": 5.860392141342163,
      "epoch": 0.43072647876141323,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.7128,
      "mean_token_accuracy": 0.13932884708046914,
      "num_tokens": 9496104.0,
      "step": 4340
    },
    {
      "entropy": 5.883248043060303,
      "epoch": 0.43122270742358076,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.6981,
      "mean_token_accuracy": 0.13705390840768814,
      "num_tokens": 9507391.0,
      "step": 4345
    },
    {
      "entropy": 5.716363573074341,
      "epoch": 0.4317189360857483,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.6374,
      "mean_token_accuracy": 0.14488721787929534,
      "num_tokens": 9518713.0,
      "step": 4350
    },
    {
      "entropy": 5.765891647338867,
      "epoch": 0.4322151647479158,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.5963,
      "mean_token_accuracy": 0.14442972615361213,
      "num_tokens": 9529930.0,
      "step": 4355
    },
    {
      "entropy": 5.753257894515992,
      "epoch": 0.43271139341008336,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.5214,
      "mean_token_accuracy": 0.1493165947496891,
      "num_tokens": 9539455.0,
      "step": 4360
    },
    {
      "entropy": 5.804065704345703,
      "epoch": 0.4332076220722509,
      "grad_norm": 0.734375,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.6349,
      "mean_token_accuracy": 0.14414166957139968,
      "num_tokens": 9551197.0,
      "step": 4365
    },
    {
      "entropy": 5.768270540237427,
      "epoch": 0.4337038507344184,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.5961,
      "mean_token_accuracy": 0.14060084372758866,
      "num_tokens": 9561240.0,
      "step": 4370
    },
    {
      "entropy": 5.790508127212524,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.6799,
      "mean_token_accuracy": 0.1315781854093075,
      "num_tokens": 9572876.0,
      "step": 4375
    },
    {
      "entropy": 5.8483240604400635,
      "epoch": 0.4346963080587535,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.6106,
      "mean_token_accuracy": 0.1432049058377743,
      "num_tokens": 9583400.0,
      "step": 4380
    },
    {
      "entropy": 5.762222099304199,
      "epoch": 0.435192536720921,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.5258,
      "mean_token_accuracy": 0.14699009358882903,
      "num_tokens": 9594637.0,
      "step": 4385
    },
    {
      "entropy": 5.743005323410034,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.6742,
      "mean_token_accuracy": 0.14157632738351822,
      "num_tokens": 9605564.0,
      "step": 4390
    },
    {
      "entropy": 5.783427429199219,
      "epoch": 0.43618499404525607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.6197,
      "mean_token_accuracy": 0.1403976060450077,
      "num_tokens": 9615601.0,
      "step": 4395
    },
    {
      "entropy": 5.743398284912109,
      "epoch": 0.4366812227074236,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.5265,
      "mean_token_accuracy": 0.1420717403292656,
      "num_tokens": 9626182.0,
      "step": 4400
    },
    {
      "entropy": 5.868048667907715,
      "epoch": 0.43717745136959113,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1418082222342491,
      "num_tokens": 9636290.0,
      "step": 4405
    },
    {
      "entropy": 5.7547532081604,
      "epoch": 0.43767368003175866,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.5669,
      "mean_token_accuracy": 0.14294910728931426,
      "num_tokens": 9648891.0,
      "step": 4410
    },
    {
      "entropy": 5.740622091293335,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.6691,
      "mean_token_accuracy": 0.13969032391905783,
      "num_tokens": 9662452.0,
      "step": 4415
    },
    {
      "entropy": 5.731381464004516,
      "epoch": 0.43866613735609367,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.5281,
      "mean_token_accuracy": 0.14229729771614075,
      "num_tokens": 9672685.0,
      "step": 4420
    },
    {
      "entropy": 5.834894180297852,
      "epoch": 0.4391623660182612,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.6568,
      "mean_token_accuracy": 0.1371921069920063,
      "num_tokens": 9683904.0,
      "step": 4425
    },
    {
      "entropy": 5.798441934585571,
      "epoch": 0.4396585946804287,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.5377,
      "mean_token_accuracy": 0.1421339616179466,
      "num_tokens": 9695463.0,
      "step": 4430
    },
    {
      "entropy": 5.793214607238769,
      "epoch": 0.44015482334259626,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.6289,
      "mean_token_accuracy": 0.14237457364797593,
      "num_tokens": 9706922.0,
      "step": 4435
    },
    {
      "entropy": 5.799474287033081,
      "epoch": 0.4406510520047638,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.6341,
      "mean_token_accuracy": 0.13740638121962548,
      "num_tokens": 9716591.0,
      "step": 4440
    },
    {
      "entropy": 5.796828365325927,
      "epoch": 0.4411472806669313,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.7006,
      "mean_token_accuracy": 0.1401236653327942,
      "num_tokens": 9727745.0,
      "step": 4445
    },
    {
      "entropy": 5.842713260650635,
      "epoch": 0.44164350932909885,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.6507,
      "mean_token_accuracy": 0.13934939727187157,
      "num_tokens": 9738663.0,
      "step": 4450
    },
    {
      "entropy": 5.718237113952637,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.5118,
      "mean_token_accuracy": 0.14976126551628113,
      "num_tokens": 9749815.0,
      "step": 4455
    },
    {
      "entropy": 5.848182296752929,
      "epoch": 0.4426359666534339,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.6391,
      "mean_token_accuracy": 0.1392246350646019,
      "num_tokens": 9760038.0,
      "step": 4460
    },
    {
      "entropy": 5.756622552871704,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.6253,
      "mean_token_accuracy": 0.14274290949106216,
      "num_tokens": 9770322.0,
      "step": 4465
    },
    {
      "entropy": 5.764551448822021,
      "epoch": 0.44362842397776897,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.6102,
      "mean_token_accuracy": 0.14604611694812775,
      "num_tokens": 9781649.0,
      "step": 4470
    },
    {
      "entropy": 5.751196050643921,
      "epoch": 0.4441246526399365,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.6174,
      "mean_token_accuracy": 0.13738834485411644,
      "num_tokens": 9793146.0,
      "step": 4475
    },
    {
      "entropy": 5.821500253677368,
      "epoch": 0.44462088130210403,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.5428,
      "mean_token_accuracy": 0.14439707696437837,
      "num_tokens": 9804645.0,
      "step": 4480
    },
    {
      "entropy": 5.754297876358033,
      "epoch": 0.44511710996427156,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.566,
      "mean_token_accuracy": 0.14867610186338426,
      "num_tokens": 9815589.0,
      "step": 4485
    },
    {
      "entropy": 5.698452997207641,
      "epoch": 0.44561333862643904,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.5089,
      "mean_token_accuracy": 0.14647819399833678,
      "num_tokens": 9826064.0,
      "step": 4490
    },
    {
      "entropy": 5.6576714515686035,
      "epoch": 0.44610956728860657,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.5813,
      "mean_token_accuracy": 0.14719886407256128,
      "num_tokens": 9835682.0,
      "step": 4495
    },
    {
      "entropy": 5.7849688053131105,
      "epoch": 0.4466057959507741,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.6224,
      "mean_token_accuracy": 0.1433667279779911,
      "num_tokens": 9846445.0,
      "step": 4500
    },
    {
      "entropy": 5.802335596084594,
      "epoch": 0.44710202461294163,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.6107,
      "mean_token_accuracy": 0.14189146533608438,
      "num_tokens": 9856537.0,
      "step": 4505
    },
    {
      "entropy": 5.826911115646363,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.6992,
      "mean_token_accuracy": 0.13836220204830169,
      "num_tokens": 9869464.0,
      "step": 4510
    },
    {
      "entropy": 5.741131401062011,
      "epoch": 0.4480944819372767,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.5264,
      "mean_token_accuracy": 0.14098931327462197,
      "num_tokens": 9879896.0,
      "step": 4515
    },
    {
      "entropy": 5.750592660903931,
      "epoch": 0.4485907105994442,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.6056,
      "mean_token_accuracy": 0.14312511458992958,
      "num_tokens": 9890261.0,
      "step": 4520
    },
    {
      "entropy": 5.79586353302002,
      "epoch": 0.44908693926161175,
      "grad_norm": 0.875,
      "learning_rate": 0.000498615629284329,
      "loss": 5.7019,
      "mean_token_accuracy": 0.1382159896194935,
      "num_tokens": 9903447.0,
      "step": 4525
    },
    {
      "entropy": 5.8367531299591064,
      "epoch": 0.4495831679237793,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.5607,
      "mean_token_accuracy": 0.1461690679192543,
      "num_tokens": 9913665.0,
      "step": 4530
    },
    {
      "entropy": 5.791276550292968,
      "epoch": 0.4500793965859468,
      "grad_norm": 0.78125,
      "learning_rate": 0.000498607769446362,
      "loss": 5.6197,
      "mean_token_accuracy": 0.14112822711467743,
      "num_tokens": 9926313.0,
      "step": 4535
    },
    {
      "entropy": 5.799275541305542,
      "epoch": 0.45057562524811434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.5776,
      "mean_token_accuracy": 0.14516764432191848,
      "num_tokens": 9936797.0,
      "step": 4540
    },
    {
      "entropy": 5.749766206741333,
      "epoch": 0.4510718539102819,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.5845,
      "mean_token_accuracy": 0.14456166476011276,
      "num_tokens": 9948623.0,
      "step": 4545
    },
    {
      "entropy": 5.755304145812988,
      "epoch": 0.4515680825724494,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.6135,
      "mean_token_accuracy": 0.1402042731642723,
      "num_tokens": 9959097.0,
      "step": 4550
    },
    {
      "entropy": 5.791376781463623,
      "epoch": 0.45206431123461693,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.638,
      "mean_token_accuracy": 0.14308259412646293,
      "num_tokens": 9969098.0,
      "step": 4555
    },
    {
      "entropy": 5.786107730865479,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.5532,
      "mean_token_accuracy": 0.14295581132173538,
      "num_tokens": 9979427.0,
      "step": 4560
    },
    {
      "entropy": 5.794799947738648,
      "epoch": 0.45305676855895194,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.7124,
      "mean_token_accuracy": 0.1363362714648247,
      "num_tokens": 9990543.0,
      "step": 4565
    },
    {
      "entropy": 5.79382004737854,
      "epoch": 0.45355299722111947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.6002,
      "mean_token_accuracy": 0.13433463424444197,
      "num_tokens": 10001119.0,
      "step": 4570
    },
    {
      "entropy": 5.810307884216309,
      "epoch": 0.454049225883287,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.6935,
      "mean_token_accuracy": 0.1423911526799202,
      "num_tokens": 10011721.0,
      "step": 4575
    },
    {
      "entropy": 5.84243597984314,
      "epoch": 0.45454545454545453,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.6066,
      "mean_token_accuracy": 0.1378902815282345,
      "num_tokens": 10025463.0,
      "step": 4580
    },
    {
      "entropy": 5.859374952316284,
      "epoch": 0.45504168320762206,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.7511,
      "mean_token_accuracy": 0.13759656250476837,
      "num_tokens": 10036250.0,
      "step": 4585
    },
    {
      "entropy": 5.793245410919189,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.6023,
      "mean_token_accuracy": 0.1445884197950363,
      "num_tokens": 10048344.0,
      "step": 4590
    },
    {
      "entropy": 5.848655414581299,
      "epoch": 0.4560341405319571,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.6626,
      "mean_token_accuracy": 0.1377781480550766,
      "num_tokens": 10059798.0,
      "step": 4595
    },
    {
      "entropy": 5.719696712493897,
      "epoch": 0.45653036919412465,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.5542,
      "mean_token_accuracy": 0.14258502498269082,
      "num_tokens": 10069905.0,
      "step": 4600
    },
    {
      "entropy": 5.751314640045166,
      "epoch": 0.4570265978562922,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.552,
      "mean_token_accuracy": 0.1417912483215332,
      "num_tokens": 10081226.0,
      "step": 4605
    },
    {
      "entropy": 5.841175079345703,
      "epoch": 0.4575228265184597,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.6644,
      "mean_token_accuracy": 0.14452986270189286,
      "num_tokens": 10092673.0,
      "step": 4610
    },
    {
      "entropy": 5.790229415893554,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.5375,
      "mean_token_accuracy": 0.147946497797966,
      "num_tokens": 10103243.0,
      "step": 4615
    },
    {
      "entropy": 5.607552480697632,
      "epoch": 0.4585152838427948,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.4944,
      "mean_token_accuracy": 0.16138856112957,
      "num_tokens": 10114936.0,
      "step": 4620
    },
    {
      "entropy": 5.716821146011353,
      "epoch": 0.4590115125049623,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.4845,
      "mean_token_accuracy": 0.14877623617649077,
      "num_tokens": 10125505.0,
      "step": 4625
    },
    {
      "entropy": 5.773360824584961,
      "epoch": 0.45950774116712984,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.6469,
      "mean_token_accuracy": 0.13701683208346366,
      "num_tokens": 10137032.0,
      "step": 4630
    },
    {
      "entropy": 5.778048896789551,
      "epoch": 0.46000396982929737,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.6209,
      "mean_token_accuracy": 0.14277000054717065,
      "num_tokens": 10147527.0,
      "step": 4635
    },
    {
      "entropy": 5.747213745117188,
      "epoch": 0.46050019849146484,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1428982563316822,
      "num_tokens": 10158848.0,
      "step": 4640
    },
    {
      "entropy": 5.769062137603759,
      "epoch": 0.46099642715363237,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000498519847524242,
      "loss": 5.6425,
      "mean_token_accuracy": 0.13826091587543488,
      "num_tokens": 10169594.0,
      "step": 4645
    },
    {
      "entropy": 5.758398199081421,
      "epoch": 0.4614926558157999,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.5947,
      "mean_token_accuracy": 0.14429235756397246,
      "num_tokens": 10180843.0,
      "step": 4650
    },
    {
      "entropy": 5.781599521636963,
      "epoch": 0.46198888447796743,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.6591,
      "mean_token_accuracy": 0.14339326396584512,
      "num_tokens": 10193226.0,
      "step": 4655
    },
    {
      "entropy": 5.770064926147461,
      "epoch": 0.46248511314013496,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.5023,
      "mean_token_accuracy": 0.14674944579601287,
      "num_tokens": 10203023.0,
      "step": 4660
    },
    {
      "entropy": 5.758601713180542,
      "epoch": 0.4629813418023025,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000498503573462923,
      "loss": 5.5471,
      "mean_token_accuracy": 0.14791421741247177,
      "num_tokens": 10214814.0,
      "step": 4665
    },
    {
      "entropy": 5.691638612747193,
      "epoch": 0.46347757046447,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.5554,
      "mean_token_accuracy": 0.14694947376847267,
      "num_tokens": 10224961.0,
      "step": 4670
    },
    {
      "entropy": 5.720160102844238,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.5125,
      "mean_token_accuracy": 0.15420230478048325,
      "num_tokens": 10235923.0,
      "step": 4675
    },
    {
      "entropy": 5.738377094268799,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.6292,
      "mean_token_accuracy": 0.13787151053547858,
      "num_tokens": 10246065.0,
      "step": 4680
    },
    {
      "entropy": 5.7926074981689455,
      "epoch": 0.4649662564509726,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.5412,
      "mean_token_accuracy": 0.14564343318343162,
      "num_tokens": 10257085.0,
      "step": 4685
    },
    {
      "entropy": 5.830641603469848,
      "epoch": 0.46546248511314015,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.6244,
      "mean_token_accuracy": 0.14183780327439308,
      "num_tokens": 10267192.0,
      "step": 4690
    },
    {
      "entropy": 5.721994161605835,
      "epoch": 0.4659587137753077,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.5808,
      "mean_token_accuracy": 0.14262180626392365,
      "num_tokens": 10278580.0,
      "step": 4695
    },
    {
      "entropy": 5.801725006103515,
      "epoch": 0.4664549424374752,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.5959,
      "mean_token_accuracy": 0.14173402413725852,
      "num_tokens": 10290034.0,
      "step": 4700
    },
    {
      "entropy": 5.76532883644104,
      "epoch": 0.46695117109964274,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.5547,
      "mean_token_accuracy": 0.13979748636484146,
      "num_tokens": 10301279.0,
      "step": 4705
    },
    {
      "entropy": 5.752497339248658,
      "epoch": 0.46744739976181027,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.5254,
      "mean_token_accuracy": 0.14921521246433259,
      "num_tokens": 10311981.0,
      "step": 4710
    },
    {
      "entropy": 5.725309419631958,
      "epoch": 0.46794362842397774,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.5436,
      "mean_token_accuracy": 0.14275271743535994,
      "num_tokens": 10323315.0,
      "step": 4715
    },
    {
      "entropy": 5.754794073104859,
      "epoch": 0.4684398570861453,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.6004,
      "mean_token_accuracy": 0.13739793747663498,
      "num_tokens": 10333736.0,
      "step": 4720
    },
    {
      "entropy": 5.806239128112793,
      "epoch": 0.4689360857483128,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498454219233913,
      "loss": 5.6768,
      "mean_token_accuracy": 0.14256877303123475,
      "num_tokens": 10344736.0,
      "step": 4725
    },
    {
      "entropy": 5.799949693679809,
      "epoch": 0.46943231441048033,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.6034,
      "mean_token_accuracy": 0.1360436648130417,
      "num_tokens": 10355914.0,
      "step": 4730
    },
    {
      "entropy": 5.717201519012451,
      "epoch": 0.46992854307264786,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.5069,
      "mean_token_accuracy": 0.1432778298854828,
      "num_tokens": 10366748.0,
      "step": 4735
    },
    {
      "entropy": 5.74832329750061,
      "epoch": 0.4704247717348154,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.6366,
      "mean_token_accuracy": 0.13813115283846855,
      "num_tokens": 10378993.0,
      "step": 4740
    },
    {
      "entropy": 5.785438346862793,
      "epoch": 0.4709210003969829,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.5254,
      "mean_token_accuracy": 0.1435009367763996,
      "num_tokens": 10388813.0,
      "step": 4745
    },
    {
      "entropy": 5.745747184753418,
      "epoch": 0.47141722905915046,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.531,
      "mean_token_accuracy": 0.14421454221010208,
      "num_tokens": 10399264.0,
      "step": 4750
    },
    {
      "entropy": 5.6879270553588865,
      "epoch": 0.471913457721318,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.523,
      "mean_token_accuracy": 0.1475764036178589,
      "num_tokens": 10408580.0,
      "step": 4755
    },
    {
      "entropy": 5.671563291549683,
      "epoch": 0.4724096863834855,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.4679,
      "mean_token_accuracy": 0.14338286966085434,
      "num_tokens": 10418246.0,
      "step": 4760
    },
    {
      "entropy": 5.740840244293213,
      "epoch": 0.47290591504565305,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.5027,
      "mean_token_accuracy": 0.15183788239955903,
      "num_tokens": 10428068.0,
      "step": 4765
    },
    {
      "entropy": 5.705519914627075,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.5603,
      "mean_token_accuracy": 0.1430072493851185,
      "num_tokens": 10438402.0,
      "step": 4770
    },
    {
      "entropy": 5.714328622817993,
      "epoch": 0.4738983723699881,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.5561,
      "mean_token_accuracy": 0.14558149576187135,
      "num_tokens": 10449280.0,
      "step": 4775
    },
    {
      "entropy": 5.800537776947022,
      "epoch": 0.47439460103215564,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.655,
      "mean_token_accuracy": 0.1406233288347721,
      "num_tokens": 10460688.0,
      "step": 4780
    },
    {
      "entropy": 5.732371997833252,
      "epoch": 0.47489082969432317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.5625,
      "mean_token_accuracy": 0.14215604737401008,
      "num_tokens": 10471652.0,
      "step": 4785
    },
    {
      "entropy": 5.65198073387146,
      "epoch": 0.47538705835649064,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.5392,
      "mean_token_accuracy": 0.14276112467050553,
      "num_tokens": 10481907.0,
      "step": 4790
    },
    {
      "entropy": 5.830527257919312,
      "epoch": 0.4758832870186582,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.5726,
      "mean_token_accuracy": 0.14328495860099794,
      "num_tokens": 10494490.0,
      "step": 4795
    },
    {
      "entropy": 5.680215311050415,
      "epoch": 0.4763795156808257,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.4991,
      "mean_token_accuracy": 0.14917379915714263,
      "num_tokens": 10505331.0,
      "step": 4800
    },
    {
      "entropy": 5.765805625915528,
      "epoch": 0.47687574434299324,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.6443,
      "mean_token_accuracy": 0.13716130703687668,
      "num_tokens": 10516282.0,
      "step": 4805
    },
    {
      "entropy": 5.749814176559449,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.4716,
      "mean_token_accuracy": 0.15196210145950317,
      "num_tokens": 10527454.0,
      "step": 4810
    },
    {
      "entropy": 5.69330244064331,
      "epoch": 0.4778682016673283,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498378691827247,
      "loss": 5.57,
      "mean_token_accuracy": 0.1436673015356064,
      "num_tokens": 10537609.0,
      "step": 4815
    },
    {
      "entropy": 5.723602724075318,
      "epoch": 0.4783644303294958,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.5325,
      "mean_token_accuracy": 0.14168880209326745,
      "num_tokens": 10548875.0,
      "step": 4820
    },
    {
      "entropy": 5.762793350219726,
      "epoch": 0.47886065899166336,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.503,
      "mean_token_accuracy": 0.1494440570473671,
      "num_tokens": 10559778.0,
      "step": 4825
    },
    {
      "entropy": 5.68077507019043,
      "epoch": 0.4793568876538309,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.5241,
      "mean_token_accuracy": 0.14728284776210784,
      "num_tokens": 10571517.0,
      "step": 4830
    },
    {
      "entropy": 5.722755718231201,
      "epoch": 0.4798531163159984,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.512,
      "mean_token_accuracy": 0.14505938664078713,
      "num_tokens": 10582209.0,
      "step": 4835
    },
    {
      "entropy": 5.706402349472046,
      "epoch": 0.48034934497816595,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.487,
      "mean_token_accuracy": 0.15195719823241233,
      "num_tokens": 10592164.0,
      "step": 4840
    },
    {
      "entropy": 5.671671056747437,
      "epoch": 0.4808455736403335,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.4838,
      "mean_token_accuracy": 0.14761274307966232,
      "num_tokens": 10602175.0,
      "step": 4845
    },
    {
      "entropy": 5.796689987182617,
      "epoch": 0.481341802302501,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.4715,
      "mean_token_accuracy": 0.15492843464016914,
      "num_tokens": 10612808.0,
      "step": 4850
    },
    {
      "entropy": 5.745422887802124,
      "epoch": 0.48183803096466854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.5714,
      "mean_token_accuracy": 0.14109451994299888,
      "num_tokens": 10622918.0,
      "step": 4855
    },
    {
      "entropy": 5.748815298080444,
      "epoch": 0.48233425962683607,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.6228,
      "mean_token_accuracy": 0.1434580571949482,
      "num_tokens": 10634607.0,
      "step": 4860
    },
    {
      "entropy": 5.698299264907837,
      "epoch": 0.48283048828900355,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.524,
      "mean_token_accuracy": 0.14604240506887436,
      "num_tokens": 10644951.0,
      "step": 4865
    },
    {
      "entropy": 5.751831769943237,
      "epoch": 0.4833267169511711,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.5602,
      "mean_token_accuracy": 0.14837105572223663,
      "num_tokens": 10655839.0,
      "step": 4870
    },
    {
      "entropy": 5.731594610214233,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.57,
      "mean_token_accuracy": 0.14353922083973886,
      "num_tokens": 10667642.0,
      "step": 4875
    },
    {
      "entropy": 5.70718560218811,
      "epoch": 0.48431917427550614,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.6143,
      "mean_token_accuracy": 0.1420133762061596,
      "num_tokens": 10677870.0,
      "step": 4880
    },
    {
      "entropy": 5.7234118461608885,
      "epoch": 0.48481540293767367,
      "grad_norm": 0.875,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.5291,
      "mean_token_accuracy": 0.14341830611228942,
      "num_tokens": 10688547.0,
      "step": 4885
    },
    {
      "entropy": 5.778681707382202,
      "epoch": 0.4853116315998412,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.5465,
      "mean_token_accuracy": 0.14223891943693162,
      "num_tokens": 10699565.0,
      "step": 4890
    },
    {
      "entropy": 5.765140056610107,
      "epoch": 0.48580786026200873,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.6569,
      "mean_token_accuracy": 0.14225051552057266,
      "num_tokens": 10711611.0,
      "step": 4895
    },
    {
      "entropy": 5.706524229049682,
      "epoch": 0.48630408892417626,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498305712433429,
      "loss": 5.5166,
      "mean_token_accuracy": 0.14108336716890335,
      "num_tokens": 10722719.0,
      "step": 4900
    },
    {
      "entropy": 5.810617971420288,
      "epoch": 0.4868003175863438,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000498301369685268,
      "loss": 5.5993,
      "mean_token_accuracy": 0.14669429659843444,
      "num_tokens": 10735308.0,
      "step": 4905
    },
    {
      "entropy": 5.677530193328858,
      "epoch": 0.4872965462485113,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.5094,
      "mean_token_accuracy": 0.15335984379053116,
      "num_tokens": 10745849.0,
      "step": 4910
    },
    {
      "entropy": 5.6549098014831545,
      "epoch": 0.48779277491067885,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.483,
      "mean_token_accuracy": 0.15263262242078782,
      "num_tokens": 10755702.0,
      "step": 4915
    },
    {
      "entropy": 5.750715446472168,
      "epoch": 0.4882890035728464,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.5674,
      "mean_token_accuracy": 0.14263114631175994,
      "num_tokens": 10766586.0,
      "step": 4920
    },
    {
      "entropy": 5.755090045928955,
      "epoch": 0.4887852322350139,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.6241,
      "mean_token_accuracy": 0.14211804419755936,
      "num_tokens": 10778634.0,
      "step": 4925
    },
    {
      "entropy": 5.757094669342041,
      "epoch": 0.48928146089718144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00049827957288573,
      "loss": 5.6214,
      "mean_token_accuracy": 0.14086363837122917,
      "num_tokens": 10789807.0,
      "step": 4930
    },
    {
      "entropy": 5.7073039531707765,
      "epoch": 0.48977768955934897,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.5214,
      "mean_token_accuracy": 0.14596231430768966,
      "num_tokens": 10801481.0,
      "step": 4935
    },
    {
      "entropy": 5.770408725738525,
      "epoch": 0.49027391822151645,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.6366,
      "mean_token_accuracy": 0.13673219382762908,
      "num_tokens": 10813787.0,
      "step": 4940
    },
    {
      "entropy": 5.721285676956176,
      "epoch": 0.490770146883684,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.5212,
      "mean_token_accuracy": 0.14413397759199142,
      "num_tokens": 10824698.0,
      "step": 4945
    },
    {
      "entropy": 5.726984357833862,
      "epoch": 0.4912663755458515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.5451,
      "mean_token_accuracy": 0.14104425087571143,
      "num_tokens": 10836308.0,
      "step": 4950
    },
    {
      "entropy": 5.7905865669250485,
      "epoch": 0.49176260420801904,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.5125,
      "mean_token_accuracy": 0.1454315721988678,
      "num_tokens": 10846844.0,
      "step": 4955
    },
    {
      "entropy": 5.655481433868408,
      "epoch": 0.49225883287018657,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.4114,
      "mean_token_accuracy": 0.15307837426662446,
      "num_tokens": 10857097.0,
      "step": 4960
    },
    {
      "entropy": 5.577139949798584,
      "epoch": 0.4927550615323541,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.4244,
      "mean_token_accuracy": 0.14815596491098404,
      "num_tokens": 10868005.0,
      "step": 4965
    },
    {
      "entropy": 5.741294574737549,
      "epoch": 0.49325129019452163,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.5194,
      "mean_token_accuracy": 0.14900355488061906,
      "num_tokens": 10878708.0,
      "step": 4970
    },
    {
      "entropy": 5.767526912689209,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.5253,
      "mean_token_accuracy": 0.1481163889169693,
      "num_tokens": 10889340.0,
      "step": 4975
    },
    {
      "entropy": 5.690439510345459,
      "epoch": 0.4942437475188567,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000498235564036352,
      "loss": 5.5572,
      "mean_token_accuracy": 0.14344785660505294,
      "num_tokens": 10901182.0,
      "step": 4980
    },
    {
      "entropy": 5.720120239257812,
      "epoch": 0.4947399761810242,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.536,
      "mean_token_accuracy": 0.13983044475317002,
      "num_tokens": 10911898.0,
      "step": 4985
    },
    {
      "entropy": 5.659397125244141,
      "epoch": 0.49523620484319175,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.5257,
      "mean_token_accuracy": 0.14723979830741882,
      "num_tokens": 10922848.0,
      "step": 4990
    },
    {
      "entropy": 5.674997425079345,
      "epoch": 0.4957324335053593,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.4523,
      "mean_token_accuracy": 0.15189297795295714,
      "num_tokens": 10933252.0,
      "step": 4995
    },
    {
      "entropy": 5.7184357166290285,
      "epoch": 0.4962286621675268,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1462848648428917,
      "num_tokens": 10944240.0,
      "step": 5000
    },
    {
      "entropy": 5.7290411472320555,
      "epoch": 0.49672489082969434,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.5457,
      "mean_token_accuracy": 0.1431691415607929,
      "num_tokens": 10954233.0,
      "step": 5005
    },
    {
      "entropy": 5.847318840026856,
      "epoch": 0.4972211194918619,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.5932,
      "mean_token_accuracy": 0.13924277797341347,
      "num_tokens": 10965252.0,
      "step": 5010
    },
    {
      "entropy": 5.801607608795166,
      "epoch": 0.4977173481540294,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.606,
      "mean_token_accuracy": 0.14378827288746834,
      "num_tokens": 10974362.0,
      "step": 5015
    },
    {
      "entropy": 5.749516296386719,
      "epoch": 0.4982135768161969,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.6852,
      "mean_token_accuracy": 0.14219343066215515,
      "num_tokens": 10987481.0,
      "step": 5020
    },
    {
      "entropy": 5.775265407562256,
      "epoch": 0.4987098054783644,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000498195482761308,
      "loss": 5.5463,
      "mean_token_accuracy": 0.14539667814970017,
      "num_tokens": 10997987.0,
      "step": 5025
    },
    {
      "entropy": 5.766027021408081,
      "epoch": 0.49920603414053194,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.6403,
      "mean_token_accuracy": 0.13737524002790452,
      "num_tokens": 11008349.0,
      "step": 5030
    },
    {
      "entropy": 5.753103017807007,
      "epoch": 0.49970226280269947,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.4851,
      "mean_token_accuracy": 0.14920002669095994,
      "num_tokens": 11017925.0,
      "step": 5035
    },
    {
      "entropy": 5.75523157119751,
      "epoch": 0.5001984914648671,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.5573,
      "mean_token_accuracy": 0.14050936996936797,
      "num_tokens": 11028710.0,
      "step": 5040
    },
    {
      "entropy": 5.7299353122711185,
      "epoch": 0.5006947201270345,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.5491,
      "mean_token_accuracy": 0.14111147299408913,
      "num_tokens": 11039827.0,
      "step": 5045
    },
    {
      "entropy": 5.686654472351075,
      "epoch": 0.5011909487892021,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.5898,
      "mean_token_accuracy": 0.14161250442266465,
      "num_tokens": 11050875.0,
      "step": 5050
    },
    {
      "entropy": 5.771305179595947,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.5932,
      "mean_token_accuracy": 0.14453488439321518,
      "num_tokens": 11063530.0,
      "step": 5055
    },
    {
      "entropy": 5.776243591308594,
      "epoch": 0.5021834061135371,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.594,
      "mean_token_accuracy": 0.1460576295852661,
      "num_tokens": 11074644.0,
      "step": 5060
    },
    {
      "entropy": 5.771616506576538,
      "epoch": 0.5026796347757047,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.5607,
      "mean_token_accuracy": 0.14273366034030915,
      "num_tokens": 11086204.0,
      "step": 5065
    },
    {
      "entropy": 5.715749454498291,
      "epoch": 0.5031758634378721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.5525,
      "mean_token_accuracy": 0.14529728889465332,
      "num_tokens": 11096547.0,
      "step": 5070
    },
    {
      "entropy": 5.75760703086853,
      "epoch": 0.5036720921000397,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498150422212006,
      "loss": 5.5224,
      "mean_token_accuracy": 0.1408478021621704,
      "num_tokens": 11106692.0,
      "step": 5075
    },
    {
      "entropy": 5.784019184112549,
      "epoch": 0.5041683207622072,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.4917,
      "mean_token_accuracy": 0.1443358115851879,
      "num_tokens": 11118065.0,
      "step": 5080
    },
    {
      "entropy": 5.660696649551392,
      "epoch": 0.5046645494243748,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.4779,
      "mean_token_accuracy": 0.15095531791448594,
      "num_tokens": 11127795.0,
      "step": 5085
    },
    {
      "entropy": 5.589940595626831,
      "epoch": 0.5051607780865423,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1452270820736885,
      "num_tokens": 11138936.0,
      "step": 5090
    },
    {
      "entropy": 5.896628999710083,
      "epoch": 0.5056570067487098,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.6485,
      "mean_token_accuracy": 0.14233459681272506,
      "num_tokens": 11150671.0,
      "step": 5095
    },
    {
      "entropy": 5.746099233627319,
      "epoch": 0.5061532354108773,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.5258,
      "mean_token_accuracy": 0.15028978437185286,
      "num_tokens": 11161413.0,
      "step": 5100
    },
    {
      "entropy": 5.683275032043457,
      "epoch": 0.5066494640730449,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.5928,
      "mean_token_accuracy": 0.14552089124917983,
      "num_tokens": 11171773.0,
      "step": 5105
    },
    {
      "entropy": 5.720244073867798,
      "epoch": 0.5071456927352124,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.611,
      "mean_token_accuracy": 0.1406140722334385,
      "num_tokens": 11183373.0,
      "step": 5110
    },
    {
      "entropy": 5.697904205322265,
      "epoch": 0.50764192139738,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.4462,
      "mean_token_accuracy": 0.1458640366792679,
      "num_tokens": 11195535.0,
      "step": 5115
    },
    {
      "entropy": 5.775056600570679,
      "epoch": 0.5081381500595474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.6412,
      "mean_token_accuracy": 0.14183022677898408,
      "num_tokens": 11206937.0,
      "step": 5120
    },
    {
      "entropy": 5.697936296463013,
      "epoch": 0.508634378721715,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.5702,
      "mean_token_accuracy": 0.1445106729865074,
      "num_tokens": 11218179.0,
      "step": 5125
    },
    {
      "entropy": 5.780670166015625,
      "epoch": 0.5091306073838825,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.5801,
      "mean_token_accuracy": 0.1428689740598202,
      "num_tokens": 11229444.0,
      "step": 5130
    },
    {
      "entropy": 5.751763439178466,
      "epoch": 0.50962683604605,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.5898,
      "mean_token_accuracy": 0.14459608271718025,
      "num_tokens": 11240869.0,
      "step": 5135
    },
    {
      "entropy": 5.712113332748413,
      "epoch": 0.5101230647082176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.5319,
      "mean_token_accuracy": 0.15055114701390265,
      "num_tokens": 11251110.0,
      "step": 5140
    },
    {
      "entropy": 5.682780170440674,
      "epoch": 0.510619293370385,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.5933,
      "mean_token_accuracy": 0.14577661454677582,
      "num_tokens": 11263522.0,
      "step": 5145
    },
    {
      "entropy": 5.6680196762084964,
      "epoch": 0.5111155220325526,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.3988,
      "mean_token_accuracy": 0.15136830508708954,
      "num_tokens": 11273703.0,
      "step": 5150
    },
    {
      "entropy": 5.781063032150269,
      "epoch": 0.5116117506947201,
      "grad_norm": 0.921875,
      "learning_rate": 0.000498077174381234,
      "loss": 5.6066,
      "mean_token_accuracy": 0.14043332785367965,
      "num_tokens": 11285834.0,
      "step": 5155
    },
    {
      "entropy": 5.753556966781616,
      "epoch": 0.5121079793568877,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.5369,
      "mean_token_accuracy": 0.14618612378835677,
      "num_tokens": 11297382.0,
      "step": 5160
    },
    {
      "entropy": 5.680868196487427,
      "epoch": 0.5126042080190552,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.5154,
      "mean_token_accuracy": 0.1420224815607071,
      "num_tokens": 11308869.0,
      "step": 5165
    },
    {
      "entropy": 5.693512916564941,
      "epoch": 0.5131004366812227,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.557,
      "mean_token_accuracy": 0.15055932104587555,
      "num_tokens": 11319273.0,
      "step": 5170
    },
    {
      "entropy": 5.761547708511353,
      "epoch": 0.5135966653433902,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.512,
      "mean_token_accuracy": 0.14337420240044593,
      "num_tokens": 11329996.0,
      "step": 5175
    },
    {
      "entropy": 5.773670721054077,
      "epoch": 0.5140928940055578,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.5039,
      "mean_token_accuracy": 0.14647472128272057,
      "num_tokens": 11341338.0,
      "step": 5180
    },
    {
      "entropy": 5.657513952255249,
      "epoch": 0.5145891226677253,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.4802,
      "mean_token_accuracy": 0.14851266294717788,
      "num_tokens": 11352546.0,
      "step": 5185
    },
    {
      "entropy": 5.711814117431641,
      "epoch": 0.5150853513298929,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.5685,
      "mean_token_accuracy": 0.1371039167046547,
      "num_tokens": 11363418.0,
      "step": 5190
    },
    {
      "entropy": 5.767662286758423,
      "epoch": 0.5155815799920603,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.5946,
      "mean_token_accuracy": 0.14328308179974555,
      "num_tokens": 11373715.0,
      "step": 5195
    },
    {
      "entropy": 5.756713438034057,
      "epoch": 0.5160778086542279,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.5379,
      "mean_token_accuracy": 0.1436440795660019,
      "num_tokens": 11385881.0,
      "step": 5200
    },
    {
      "entropy": 5.705041694641113,
      "epoch": 0.5165740373163954,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.5266,
      "mean_token_accuracy": 0.14371063709259033,
      "num_tokens": 11395917.0,
      "step": 5205
    },
    {
      "entropy": 5.72797532081604,
      "epoch": 0.5170702659785629,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.4824,
      "mean_token_accuracy": 0.14561500996351243,
      "num_tokens": 11407210.0,
      "step": 5210
    },
    {
      "entropy": 5.738058710098267,
      "epoch": 0.5175664946407305,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.5568,
      "mean_token_accuracy": 0.14780419021844865,
      "num_tokens": 11418715.0,
      "step": 5215
    },
    {
      "entropy": 5.7290088653564455,
      "epoch": 0.5180627233028979,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.5161,
      "mean_token_accuracy": 0.14652584344148636,
      "num_tokens": 11428964.0,
      "step": 5220
    },
    {
      "entropy": 5.678311729431153,
      "epoch": 0.5185589519650655,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.4704,
      "mean_token_accuracy": 0.14789997041225433,
      "num_tokens": 11439644.0,
      "step": 5225
    },
    {
      "entropy": 5.704117679595948,
      "epoch": 0.519055180627233,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.5684,
      "mean_token_accuracy": 0.14092440232634545,
      "num_tokens": 11450737.0,
      "step": 5230
    },
    {
      "entropy": 5.733443784713745,
      "epoch": 0.5195514092894006,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980025104019,
      "loss": 5.5499,
      "mean_token_accuracy": 0.1493789568543434,
      "num_tokens": 11461054.0,
      "step": 5235
    },
    {
      "entropy": 5.679115247726441,
      "epoch": 0.520047637951568,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.5115,
      "mean_token_accuracy": 0.14243048802018166,
      "num_tokens": 11470900.0,
      "step": 5240
    },
    {
      "entropy": 5.66800594329834,
      "epoch": 0.5205438666137356,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.476,
      "mean_token_accuracy": 0.15013523325324057,
      "num_tokens": 11480551.0,
      "step": 5245
    },
    {
      "entropy": 5.681051254272461,
      "epoch": 0.5210400952759031,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.4523,
      "mean_token_accuracy": 0.15104128569364547,
      "num_tokens": 11491317.0,
      "step": 5250
    },
    {
      "entropy": 5.682211923599243,
      "epoch": 0.5215363239380707,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.5658,
      "mean_token_accuracy": 0.14473358616232873,
      "num_tokens": 11502975.0,
      "step": 5255
    },
    {
      "entropy": 5.729284763336182,
      "epoch": 0.5220325526002382,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.535,
      "mean_token_accuracy": 0.1405423954129219,
      "num_tokens": 11514803.0,
      "step": 5260
    },
    {
      "entropy": 5.76097846031189,
      "epoch": 0.5225287812624058,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497974146405303,
      "loss": 5.5608,
      "mean_token_accuracy": 0.14489552080631257,
      "num_tokens": 11525782.0,
      "step": 5265
    },
    {
      "entropy": 5.802859878540039,
      "epoch": 0.5230250099245732,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.6009,
      "mean_token_accuracy": 0.14190295189619065,
      "num_tokens": 11535857.0,
      "step": 5270
    },
    {
      "entropy": 5.762614154815674,
      "epoch": 0.5235212385867408,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.6049,
      "mean_token_accuracy": 0.14444348067045212,
      "num_tokens": 11547385.0,
      "step": 5275
    },
    {
      "entropy": 5.735829925537109,
      "epoch": 0.5240174672489083,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.4827,
      "mean_token_accuracy": 0.15125673189759253,
      "num_tokens": 11559382.0,
      "step": 5280
    },
    {
      "entropy": 5.741348791122436,
      "epoch": 0.5245136959110758,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.5748,
      "mean_token_accuracy": 0.14045862033963202,
      "num_tokens": 11570096.0,
      "step": 5285
    },
    {
      "entropy": 5.687244176864624,
      "epoch": 0.5250099245732434,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.4595,
      "mean_token_accuracy": 0.15186235532164574,
      "num_tokens": 11580543.0,
      "step": 5290
    },
    {
      "entropy": 5.654762315750122,
      "epoch": 0.5255061532354108,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1539733052253723,
      "num_tokens": 11591132.0,
      "step": 5295
    },
    {
      "entropy": 5.703864574432373,
      "epoch": 0.5260023818975784,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1441838711500168,
      "num_tokens": 11602684.0,
      "step": 5300
    },
    {
      "entropy": 5.713519716262818,
      "epoch": 0.5264986105597459,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.5545,
      "mean_token_accuracy": 0.1504795104265213,
      "num_tokens": 11614009.0,
      "step": 5305
    },
    {
      "entropy": 5.7190515995025635,
      "epoch": 0.5269948392219135,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.6684,
      "mean_token_accuracy": 0.1400254748761654,
      "num_tokens": 11624577.0,
      "step": 5310
    },
    {
      "entropy": 5.790601539611816,
      "epoch": 0.527491067884081,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000497926430747989,
      "loss": 5.5737,
      "mean_token_accuracy": 0.1412855789065361,
      "num_tokens": 11636382.0,
      "step": 5315
    },
    {
      "entropy": 5.693267488479615,
      "epoch": 0.5279872965462485,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.5219,
      "mean_token_accuracy": 0.14752980172634125,
      "num_tokens": 11648606.0,
      "step": 5320
    },
    {
      "entropy": 5.758263397216797,
      "epoch": 0.528483525208416,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497916821275588,
      "loss": 5.5143,
      "mean_token_accuracy": 0.14610515534877777,
      "num_tokens": 11658191.0,
      "step": 5325
    },
    {
      "entropy": 5.800335216522217,
      "epoch": 0.5289797538705836,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.5913,
      "mean_token_accuracy": 0.14024471640586852,
      "num_tokens": 11668955.0,
      "step": 5330
    },
    {
      "entropy": 5.649293613433838,
      "epoch": 0.5294759825327511,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.5543,
      "mean_token_accuracy": 0.14026069939136504,
      "num_tokens": 11680634.0,
      "step": 5335
    },
    {
      "entropy": 5.814202690124512,
      "epoch": 0.5299722111949187,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.5892,
      "mean_token_accuracy": 0.14762159585952758,
      "num_tokens": 11692155.0,
      "step": 5340
    },
    {
      "entropy": 5.709968757629395,
      "epoch": 0.5304684398570861,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.5468,
      "mean_token_accuracy": 0.142489093542099,
      "num_tokens": 11702968.0,
      "step": 5345
    },
    {
      "entropy": 5.787270879745483,
      "epoch": 0.5309646685192537,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.5992,
      "mean_token_accuracy": 0.14081287980079651,
      "num_tokens": 11714329.0,
      "step": 5350
    },
    {
      "entropy": 5.648181056976318,
      "epoch": 0.5314608971814212,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.3987,
      "mean_token_accuracy": 0.1549681007862091,
      "num_tokens": 11724071.0,
      "step": 5355
    },
    {
      "entropy": 5.719262218475341,
      "epoch": 0.5319571258435887,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.5938,
      "mean_token_accuracy": 0.14241832122206688,
      "num_tokens": 11735939.0,
      "step": 5360
    },
    {
      "entropy": 5.7412763118743895,
      "epoch": 0.5324533545057563,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.6077,
      "mean_token_accuracy": 0.14213175773620607,
      "num_tokens": 11747965.0,
      "step": 5365
    },
    {
      "entropy": 5.771096277236938,
      "epoch": 0.5329495831679237,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.5144,
      "mean_token_accuracy": 0.1431013822555542,
      "num_tokens": 11758206.0,
      "step": 5370
    },
    {
      "entropy": 5.711185789108276,
      "epoch": 0.5334458118300913,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.5418,
      "mean_token_accuracy": 0.148122126609087,
      "num_tokens": 11770158.0,
      "step": 5375
    },
    {
      "entropy": 5.606020498275757,
      "epoch": 0.5339420404922588,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.3807,
      "mean_token_accuracy": 0.15766320377588272,
      "num_tokens": 11780332.0,
      "step": 5380
    },
    {
      "entropy": 5.6558021068572994,
      "epoch": 0.5344382691544264,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.5226,
      "mean_token_accuracy": 0.14899348244071006,
      "num_tokens": 11791178.0,
      "step": 5385
    },
    {
      "entropy": 5.658620691299438,
      "epoch": 0.5349344978165939,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.4923,
      "mean_token_accuracy": 0.1501137539744377,
      "num_tokens": 11800701.0,
      "step": 5390
    },
    {
      "entropy": 5.675882959365845,
      "epoch": 0.5354307264787614,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.5018,
      "mean_token_accuracy": 0.14781143814325332,
      "num_tokens": 11811727.0,
      "step": 5395
    },
    {
      "entropy": 5.751688480377197,
      "epoch": 0.5359269551409289,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.5785,
      "mean_token_accuracy": 0.14426472634077073,
      "num_tokens": 11822892.0,
      "step": 5400
    },
    {
      "entropy": 5.75275239944458,
      "epoch": 0.5364231838030965,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.5423,
      "mean_token_accuracy": 0.14597827419638634,
      "num_tokens": 11834547.0,
      "step": 5405
    },
    {
      "entropy": 5.709875154495239,
      "epoch": 0.536919412465264,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.5167,
      "mean_token_accuracy": 0.15175689309835433,
      "num_tokens": 11845667.0,
      "step": 5410
    },
    {
      "entropy": 5.701542901992798,
      "epoch": 0.5374156411274316,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.4614,
      "mean_token_accuracy": 0.1492401197552681,
      "num_tokens": 11856228.0,
      "step": 5415
    },
    {
      "entropy": 5.68884801864624,
      "epoch": 0.537911869789599,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.5095,
      "mean_token_accuracy": 0.14929999262094498,
      "num_tokens": 11867491.0,
      "step": 5420
    },
    {
      "entropy": 5.713821840286255,
      "epoch": 0.5384080984517666,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.604,
      "mean_token_accuracy": 0.14582066833972931,
      "num_tokens": 11878635.0,
      "step": 5425
    },
    {
      "entropy": 5.75053448677063,
      "epoch": 0.5389043271139341,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.518,
      "mean_token_accuracy": 0.14597933515906333,
      "num_tokens": 11890204.0,
      "step": 5430
    },
    {
      "entropy": 5.720219421386719,
      "epoch": 0.5394005557761016,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.5659,
      "mean_token_accuracy": 0.14796526730060577,
      "num_tokens": 11901049.0,
      "step": 5435
    },
    {
      "entropy": 5.674677467346191,
      "epoch": 0.5398967844382692,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.5177,
      "mean_token_accuracy": 0.1466207966208458,
      "num_tokens": 11912957.0,
      "step": 5440
    },
    {
      "entropy": 5.771763324737549,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.5386,
      "mean_token_accuracy": 0.14368943870067596,
      "num_tokens": 11924583.0,
      "step": 5445
    },
    {
      "entropy": 5.642049503326416,
      "epoch": 0.5408892417626042,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.4801,
      "mean_token_accuracy": 0.14956890642642975,
      "num_tokens": 11934898.0,
      "step": 5450
    },
    {
      "entropy": 5.658617067337036,
      "epoch": 0.5413854704247717,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.468,
      "mean_token_accuracy": 0.14964041709899903,
      "num_tokens": 11944577.0,
      "step": 5455
    },
    {
      "entropy": 5.6672618865966795,
      "epoch": 0.5418816990869393,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.5256,
      "mean_token_accuracy": 0.15061981230974197,
      "num_tokens": 11956184.0,
      "step": 5460
    },
    {
      "entropy": 5.7081310749053955,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.5468,
      "mean_token_accuracy": 0.14525931626558303,
      "num_tokens": 11966929.0,
      "step": 5465
    },
    {
      "entropy": 5.743371105194091,
      "epoch": 0.5428741564112743,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.4909,
      "mean_token_accuracy": 0.15600426346063614,
      "num_tokens": 11978166.0,
      "step": 5470
    },
    {
      "entropy": 5.728525018692016,
      "epoch": 0.5433703850734418,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.5724,
      "mean_token_accuracy": 0.14086554646492006,
      "num_tokens": 11988719.0,
      "step": 5475
    },
    {
      "entropy": 5.737892532348633,
      "epoch": 0.5438666137356094,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.5461,
      "mean_token_accuracy": 0.1428004577755928,
      "num_tokens": 11999877.0,
      "step": 5480
    },
    {
      "entropy": 5.69523024559021,
      "epoch": 0.5443628423977769,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.5103,
      "mean_token_accuracy": 0.1473350539803505,
      "num_tokens": 12011123.0,
      "step": 5485
    },
    {
      "entropy": 5.722416639328003,
      "epoch": 0.5448590710599445,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.4883,
      "mean_token_accuracy": 0.1468016877770424,
      "num_tokens": 12021494.0,
      "step": 5490
    },
    {
      "entropy": 5.753288125991821,
      "epoch": 0.5453552997221119,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.5503,
      "mean_token_accuracy": 0.14130862951278686,
      "num_tokens": 12032780.0,
      "step": 5495
    },
    {
      "entropy": 5.653288125991821,
      "epoch": 0.5458515283842795,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.5505,
      "mean_token_accuracy": 0.1433147445321083,
      "num_tokens": 12043910.0,
      "step": 5500
    },
    {
      "entropy": 5.670283603668213,
      "epoch": 0.546347757046447,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.4559,
      "mean_token_accuracy": 0.15128089636564254,
      "num_tokens": 12053722.0,
      "step": 5505
    },
    {
      "entropy": 5.686805295944214,
      "epoch": 0.5468439857086145,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.5672,
      "mean_token_accuracy": 0.14593221247196198,
      "num_tokens": 12065316.0,
      "step": 5510
    },
    {
      "entropy": 5.738336801528931,
      "epoch": 0.5473402143707821,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.505,
      "mean_token_accuracy": 0.1411941595375538,
      "num_tokens": 12076095.0,
      "step": 5515
    },
    {
      "entropy": 5.724178504943848,
      "epoch": 0.5478364430329495,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.4775,
      "mean_token_accuracy": 0.14833004772663116,
      "num_tokens": 12086960.0,
      "step": 5520
    },
    {
      "entropy": 5.671430826187134,
      "epoch": 0.5483326716951171,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497719989711627,
      "loss": 5.4652,
      "mean_token_accuracy": 0.1522202044725418,
      "num_tokens": 12097534.0,
      "step": 5525
    },
    {
      "entropy": 5.646196937561035,
      "epoch": 0.5488289003572846,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.4219,
      "mean_token_accuracy": 0.15497131198644637,
      "num_tokens": 12110348.0,
      "step": 5530
    },
    {
      "entropy": 5.652182626724243,
      "epoch": 0.5493251290194522,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.4425,
      "mean_token_accuracy": 0.14648700058460234,
      "num_tokens": 12121171.0,
      "step": 5535
    },
    {
      "entropy": 5.695895004272461,
      "epoch": 0.5498213576816197,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.5762,
      "mean_token_accuracy": 0.13877081498503685,
      "num_tokens": 12133054.0,
      "step": 5540
    },
    {
      "entropy": 5.719188833236695,
      "epoch": 0.5503175863437872,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.4826,
      "mean_token_accuracy": 0.14942405819892884,
      "num_tokens": 12144278.0,
      "step": 5545
    },
    {
      "entropy": 5.749782371520996,
      "epoch": 0.5508138150059547,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.6103,
      "mean_token_accuracy": 0.14077139124274254,
      "num_tokens": 12154446.0,
      "step": 5550
    },
    {
      "entropy": 5.73576135635376,
      "epoch": 0.5513100436681223,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.4866,
      "mean_token_accuracy": 0.14601529091596605,
      "num_tokens": 12165176.0,
      "step": 5555
    },
    {
      "entropy": 5.700368404388428,
      "epoch": 0.5518062723302898,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.5005,
      "mean_token_accuracy": 0.14673406779766082,
      "num_tokens": 12176329.0,
      "step": 5560
    },
    {
      "entropy": 5.766986751556397,
      "epoch": 0.5523025009924574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.5196,
      "mean_token_accuracy": 0.1453908458352089,
      "num_tokens": 12186927.0,
      "step": 5565
    },
    {
      "entropy": 5.7339111328125,
      "epoch": 0.5527987296546248,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.558,
      "mean_token_accuracy": 0.155231836438179,
      "num_tokens": 12196632.0,
      "step": 5570
    },
    {
      "entropy": 5.6834649562835695,
      "epoch": 0.5532949583167924,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.5048,
      "mean_token_accuracy": 0.14794395565986634,
      "num_tokens": 12207099.0,
      "step": 5575
    },
    {
      "entropy": 5.803468036651611,
      "epoch": 0.5537911869789599,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000497664311502617,
      "loss": 5.5711,
      "mean_token_accuracy": 0.13951099216938018,
      "num_tokens": 12218595.0,
      "step": 5580
    },
    {
      "entropy": 5.68735785484314,
      "epoch": 0.5542874156411274,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.5177,
      "mean_token_accuracy": 0.1455443173646927,
      "num_tokens": 12229280.0,
      "step": 5585
    },
    {
      "entropy": 5.665556478500366,
      "epoch": 0.554783644303295,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.4868,
      "mean_token_accuracy": 0.14853484332561492,
      "num_tokens": 12240079.0,
      "step": 5590
    },
    {
      "entropy": 5.649208879470825,
      "epoch": 0.5552798729654624,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.4664,
      "mean_token_accuracy": 0.14796095639467238,
      "num_tokens": 12250083.0,
      "step": 5595
    },
    {
      "entropy": 5.625807046890259,
      "epoch": 0.55577610162763,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1555221602320671,
      "num_tokens": 12260868.0,
      "step": 5600
    },
    {
      "entropy": 5.600586366653443,
      "epoch": 0.5562723302897975,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.4601,
      "mean_token_accuracy": 0.14932666048407556,
      "num_tokens": 12272268.0,
      "step": 5605
    },
    {
      "entropy": 5.714358758926392,
      "epoch": 0.5567685589519651,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.4988,
      "mean_token_accuracy": 0.14387020766735076,
      "num_tokens": 12283347.0,
      "step": 5610
    },
    {
      "entropy": 5.678732252120971,
      "epoch": 0.5572647876141326,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.5251,
      "mean_token_accuracy": 0.14423660859465598,
      "num_tokens": 12294639.0,
      "step": 5615
    },
    {
      "entropy": 5.7032191276550295,
      "epoch": 0.5577610162763001,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.5031,
      "mean_token_accuracy": 0.15360878258943558,
      "num_tokens": 12304573.0,
      "step": 5620
    },
    {
      "entropy": 5.78430814743042,
      "epoch": 0.5582572449384676,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.5368,
      "mean_token_accuracy": 0.14589423388242723,
      "num_tokens": 12314559.0,
      "step": 5625
    },
    {
      "entropy": 5.708568286895752,
      "epoch": 0.5587534736006352,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.4564,
      "mean_token_accuracy": 0.14999575465917586,
      "num_tokens": 12324433.0,
      "step": 5630
    },
    {
      "entropy": 5.655963993072509,
      "epoch": 0.5592497022628027,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.5973,
      "mean_token_accuracy": 0.13453896194696427,
      "num_tokens": 12335039.0,
      "step": 5635
    },
    {
      "entropy": 5.7656334400177,
      "epoch": 0.5597459309249703,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.475,
      "mean_token_accuracy": 0.15004511773586274,
      "num_tokens": 12345917.0,
      "step": 5640
    },
    {
      "entropy": 5.732184886932373,
      "epoch": 0.5602421595871377,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.539,
      "mean_token_accuracy": 0.14373888224363326,
      "num_tokens": 12357362.0,
      "step": 5645
    },
    {
      "entropy": 5.691069555282593,
      "epoch": 0.5607383882493053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.5193,
      "mean_token_accuracy": 0.15101567581295966,
      "num_tokens": 12368012.0,
      "step": 5650
    },
    {
      "entropy": 5.73072075843811,
      "epoch": 0.5612346169114728,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1452140413224697,
      "num_tokens": 12378745.0,
      "step": 5655
    },
    {
      "entropy": 5.679556798934937,
      "epoch": 0.5617308455736403,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.4516,
      "mean_token_accuracy": 0.14968680888414382,
      "num_tokens": 12388821.0,
      "step": 5660
    },
    {
      "entropy": 5.841395521163941,
      "epoch": 0.5622270742358079,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.6054,
      "mean_token_accuracy": 0.1351458415389061,
      "num_tokens": 12399798.0,
      "step": 5665
    },
    {
      "entropy": 5.776484537124634,
      "epoch": 0.5627233028979753,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.4968,
      "mean_token_accuracy": 0.1472727134823799,
      "num_tokens": 12410887.0,
      "step": 5670
    },
    {
      "entropy": 5.681442070007324,
      "epoch": 0.5632195315601429,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.5417,
      "mean_token_accuracy": 0.14771043360233307,
      "num_tokens": 12422430.0,
      "step": 5675
    },
    {
      "entropy": 5.682510662078857,
      "epoch": 0.5637157602223104,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.4886,
      "mean_token_accuracy": 0.14588471949100495,
      "num_tokens": 12432930.0,
      "step": 5680
    },
    {
      "entropy": 5.715798091888428,
      "epoch": 0.564211988884478,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497556161626688,
      "loss": 5.5359,
      "mean_token_accuracy": 0.14626526534557344,
      "num_tokens": 12444117.0,
      "step": 5685
    },
    {
      "entropy": 5.666935300827026,
      "epoch": 0.5647082175466455,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.4849,
      "mean_token_accuracy": 0.14816055819392204,
      "num_tokens": 12454619.0,
      "step": 5690
    },
    {
      "entropy": 5.71806058883667,
      "epoch": 0.565204446208813,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.5184,
      "mean_token_accuracy": 0.1513572871685028,
      "num_tokens": 12464666.0,
      "step": 5695
    },
    {
      "entropy": 5.757439851760864,
      "epoch": 0.5657006748709805,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.5773,
      "mean_token_accuracy": 0.14439682960510253,
      "num_tokens": 12474923.0,
      "step": 5700
    },
    {
      "entropy": 5.660067176818847,
      "epoch": 0.5661969035331481,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.4611,
      "mean_token_accuracy": 0.14791497886180877,
      "num_tokens": 12485505.0,
      "step": 5705
    },
    {
      "entropy": 5.636746072769165,
      "epoch": 0.5666931321953156,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.5397,
      "mean_token_accuracy": 0.14973339289426804,
      "num_tokens": 12497489.0,
      "step": 5710
    },
    {
      "entropy": 5.704427242279053,
      "epoch": 0.5671893608574832,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.4964,
      "mean_token_accuracy": 0.14984386861324311,
      "num_tokens": 12508117.0,
      "step": 5715
    },
    {
      "entropy": 5.671060371398926,
      "epoch": 0.5676855895196506,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.4942,
      "mean_token_accuracy": 0.14745090827345847,
      "num_tokens": 12519090.0,
      "step": 5720
    },
    {
      "entropy": 5.714696836471558,
      "epoch": 0.5681818181818182,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.6377,
      "mean_token_accuracy": 0.13868438005447387,
      "num_tokens": 12530427.0,
      "step": 5725
    },
    {
      "entropy": 5.682834196090698,
      "epoch": 0.5686780468439857,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.4093,
      "mean_token_accuracy": 0.1557854413986206,
      "num_tokens": 12542142.0,
      "step": 5730
    },
    {
      "entropy": 5.669261884689331,
      "epoch": 0.5691742755061532,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.4716,
      "mean_token_accuracy": 0.14626218229532242,
      "num_tokens": 12552716.0,
      "step": 5735
    },
    {
      "entropy": 5.664806842803955,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.4651,
      "mean_token_accuracy": 0.14478039592504502,
      "num_tokens": 12562659.0,
      "step": 5740
    },
    {
      "entropy": 5.700056409835815,
      "epoch": 0.5701667328304882,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.432,
      "mean_token_accuracy": 0.1550305038690567,
      "num_tokens": 12573982.0,
      "step": 5745
    },
    {
      "entropy": 5.675100469589234,
      "epoch": 0.5706629614926558,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.4179,
      "mean_token_accuracy": 0.1572493478655815,
      "num_tokens": 12584402.0,
      "step": 5750
    },
    {
      "entropy": 5.654999494552612,
      "epoch": 0.5711591901548233,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.5063,
      "mean_token_accuracy": 0.15086663961410524,
      "num_tokens": 12595342.0,
      "step": 5755
    },
    {
      "entropy": 5.6311304569244385,
      "epoch": 0.5716554188169909,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.4451,
      "mean_token_accuracy": 0.1464613065123558,
      "num_tokens": 12605634.0,
      "step": 5760
    },
    {
      "entropy": 5.712693929672241,
      "epoch": 0.5721516474791584,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.482,
      "mean_token_accuracy": 0.14805632829666138,
      "num_tokens": 12618000.0,
      "step": 5765
    },
    {
      "entropy": 5.629064559936523,
      "epoch": 0.572647876141326,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.3117,
      "mean_token_accuracy": 0.15821036249399184,
      "num_tokens": 12629744.0,
      "step": 5770
    },
    {
      "entropy": 5.592479276657104,
      "epoch": 0.5731441048034934,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.43,
      "mean_token_accuracy": 0.1496155872941017,
      "num_tokens": 12639752.0,
      "step": 5775
    },
    {
      "entropy": 5.625934076309204,
      "epoch": 0.573640333465661,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.4456,
      "mean_token_accuracy": 0.15022974014282225,
      "num_tokens": 12650219.0,
      "step": 5780
    },
    {
      "entropy": 5.731177139282226,
      "epoch": 0.5741365621278285,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.5599,
      "mean_token_accuracy": 0.14314404800534247,
      "num_tokens": 12661589.0,
      "step": 5785
    },
    {
      "entropy": 5.708392477035522,
      "epoch": 0.5746327907899961,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.5481,
      "mean_token_accuracy": 0.14462292194366455,
      "num_tokens": 12674794.0,
      "step": 5790
    },
    {
      "entropy": 5.726592636108398,
      "epoch": 0.5751290194521635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.4621,
      "mean_token_accuracy": 0.15062210634350776,
      "num_tokens": 12684573.0,
      "step": 5795
    },
    {
      "entropy": 5.6901463031768795,
      "epoch": 0.5756252481143311,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.4698,
      "mean_token_accuracy": 0.15178433656692505,
      "num_tokens": 12695319.0,
      "step": 5800
    },
    {
      "entropy": 5.66407151222229,
      "epoch": 0.5761214767764986,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.4646,
      "mean_token_accuracy": 0.15256776362657548,
      "num_tokens": 12706007.0,
      "step": 5805
    },
    {
      "entropy": 5.7142106056213375,
      "epoch": 0.5766177054386661,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.4826,
      "mean_token_accuracy": 0.1548708349466324,
      "num_tokens": 12716213.0,
      "step": 5810
    },
    {
      "entropy": 5.730684423446656,
      "epoch": 0.5771139341008337,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000497418889859794,
      "loss": 5.5054,
      "mean_token_accuracy": 0.1462215453386307,
      "num_tokens": 12727643.0,
      "step": 5815
    },
    {
      "entropy": 5.690935802459717,
      "epoch": 0.5776101627630011,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.5139,
      "mean_token_accuracy": 0.14824384078383446,
      "num_tokens": 12738535.0,
      "step": 5820
    },
    {
      "entropy": 5.684843587875366,
      "epoch": 0.5781063914251687,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.4944,
      "mean_token_accuracy": 0.15067617148160933,
      "num_tokens": 12750419.0,
      "step": 5825
    },
    {
      "entropy": 5.660930967330932,
      "epoch": 0.5786026200873362,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.467,
      "mean_token_accuracy": 0.14467859491705895,
      "num_tokens": 12759827.0,
      "step": 5830
    },
    {
      "entropy": 5.668339252471924,
      "epoch": 0.5790988487495038,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.4925,
      "mean_token_accuracy": 0.15156256407499313,
      "num_tokens": 12770328.0,
      "step": 5835
    },
    {
      "entropy": 5.733518838882446,
      "epoch": 0.5795950774116713,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.4657,
      "mean_token_accuracy": 0.1480121061205864,
      "num_tokens": 12781788.0,
      "step": 5840
    },
    {
      "entropy": 5.595726871490479,
      "epoch": 0.5800913060738389,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.3876,
      "mean_token_accuracy": 0.15593918263912201,
      "num_tokens": 12792134.0,
      "step": 5845
    },
    {
      "entropy": 5.611249542236328,
      "epoch": 0.5805875347360063,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.4282,
      "mean_token_accuracy": 0.14744565784931182,
      "num_tokens": 12801810.0,
      "step": 5850
    },
    {
      "entropy": 5.7450704097747805,
      "epoch": 0.5810837633981739,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.568,
      "mean_token_accuracy": 0.14706992655992507,
      "num_tokens": 12814412.0,
      "step": 5855
    },
    {
      "entropy": 5.669381999969483,
      "epoch": 0.5815799920603414,
      "grad_norm": 0.875,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.5154,
      "mean_token_accuracy": 0.14339750856161118,
      "num_tokens": 12826344.0,
      "step": 5860
    },
    {
      "entropy": 5.684680652618408,
      "epoch": 0.582076220722509,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.5293,
      "mean_token_accuracy": 0.14462660253047943,
      "num_tokens": 12837434.0,
      "step": 5865
    },
    {
      "entropy": 5.691894435882569,
      "epoch": 0.5825724493846764,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.5075,
      "mean_token_accuracy": 0.14309678077697754,
      "num_tokens": 12848681.0,
      "step": 5870
    },
    {
      "entropy": 5.704286384582519,
      "epoch": 0.583068678046844,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.6055,
      "mean_token_accuracy": 0.1407423920929432,
      "num_tokens": 12860300.0,
      "step": 5875
    },
    {
      "entropy": 5.734279632568359,
      "epoch": 0.5835649067090115,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497348855650453,
      "loss": 5.4837,
      "mean_token_accuracy": 0.14364191889762878,
      "num_tokens": 12871079.0,
      "step": 5880
    },
    {
      "entropy": 5.62366623878479,
      "epoch": 0.584061135371179,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.4802,
      "mean_token_accuracy": 0.15506502836942673,
      "num_tokens": 12881168.0,
      "step": 5885
    },
    {
      "entropy": 5.723691940307617,
      "epoch": 0.5845573640333466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.5361,
      "mean_token_accuracy": 0.14893506616353988,
      "num_tokens": 12891912.0,
      "step": 5890
    },
    {
      "entropy": 5.660767841339111,
      "epoch": 0.585053592695514,
      "grad_norm": 0.875,
      "learning_rate": 0.000497332561568575,
      "loss": 5.4902,
      "mean_token_accuracy": 0.1484683483839035,
      "num_tokens": 12902984.0,
      "step": 5895
    },
    {
      "entropy": 5.703382062911987,
      "epoch": 0.5855498213576816,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.53,
      "mean_token_accuracy": 0.14851990342140198,
      "num_tokens": 12914635.0,
      "step": 5900
    },
    {
      "entropy": 5.714743709564209,
      "epoch": 0.5860460500198491,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.4407,
      "mean_token_accuracy": 0.15131992548704148,
      "num_tokens": 12924575.0,
      "step": 5905
    },
    {
      "entropy": 5.642918872833252,
      "epoch": 0.5865422786820167,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497316217866382,
      "loss": 5.3906,
      "mean_token_accuracy": 0.15495836287736892,
      "num_tokens": 12935404.0,
      "step": 5910
    },
    {
      "entropy": 5.64871129989624,
      "epoch": 0.5870385073441842,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1580519050359726,
      "num_tokens": 12946772.0,
      "step": 5915
    },
    {
      "entropy": 5.7082726001739506,
      "epoch": 0.5875347360063518,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.5364,
      "mean_token_accuracy": 0.1475169390439987,
      "num_tokens": 12956651.0,
      "step": 5920
    },
    {
      "entropy": 5.684748649597168,
      "epoch": 0.5880309646685192,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.5727,
      "mean_token_accuracy": 0.1417674317955971,
      "num_tokens": 12968206.0,
      "step": 5925
    },
    {
      "entropy": 5.761432743072509,
      "epoch": 0.5885271933306868,
      "grad_norm": 0.76953125,
      "learning_rate": 0.000497294349082607,
      "loss": 5.5347,
      "mean_token_accuracy": 0.15004124194383622,
      "num_tokens": 12980050.0,
      "step": 5930
    },
    {
      "entropy": 5.679399108886718,
      "epoch": 0.5890234219928543,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.5269,
      "mean_token_accuracy": 0.14604859352111815,
      "num_tokens": 12990514.0,
      "step": 5935
    },
    {
      "entropy": 5.699740409851074,
      "epoch": 0.5895196506550219,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.5481,
      "mean_token_accuracy": 0.1457527607679367,
      "num_tokens": 13000803.0,
      "step": 5940
    },
    {
      "entropy": 5.664505529403686,
      "epoch": 0.5900158793171894,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.4369,
      "mean_token_accuracy": 0.15405270159244538,
      "num_tokens": 13010877.0,
      "step": 5945
    },
    {
      "entropy": 5.648125219345093,
      "epoch": 0.5905121079793569,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.4775,
      "mean_token_accuracy": 0.14735304266214372,
      "num_tokens": 13021906.0,
      "step": 5950
    },
    {
      "entropy": 5.721935272216797,
      "epoch": 0.5910083366415244,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.5298,
      "mean_token_accuracy": 0.14616427272558213,
      "num_tokens": 13033977.0,
      "step": 5955
    },
    {
      "entropy": 5.7889549255371096,
      "epoch": 0.5915045653036919,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.6038,
      "mean_token_accuracy": 0.14398072883486748,
      "num_tokens": 13044078.0,
      "step": 5960
    },
    {
      "entropy": 5.697792863845825,
      "epoch": 0.5920007939658595,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.5275,
      "mean_token_accuracy": 0.14884334802627563,
      "num_tokens": 13055779.0,
      "step": 5965
    },
    {
      "entropy": 5.657936763763428,
      "epoch": 0.592497022628027,
      "grad_norm": 0.78125,
      "learning_rate": 0.000497250346927576,
      "loss": 5.4393,
      "mean_token_accuracy": 0.14896177351474763,
      "num_tokens": 13066727.0,
      "step": 5970
    },
    {
      "entropy": 5.6672728061676025,
      "epoch": 0.5929932512901945,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.3889,
      "mean_token_accuracy": 0.15317624807357788,
      "num_tokens": 13077678.0,
      "step": 5975
    },
    {
      "entropy": 5.69059386253357,
      "epoch": 0.593489479952362,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.4937,
      "mean_token_accuracy": 0.13894937336444854,
      "num_tokens": 13088612.0,
      "step": 5980
    },
    {
      "entropy": 5.705918073654175,
      "epoch": 0.5939857086145296,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.4994,
      "mean_token_accuracy": 0.15010869204998017,
      "num_tokens": 13100960.0,
      "step": 5985
    },
    {
      "entropy": 5.619600057601929,
      "epoch": 0.5944819372766971,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.4586,
      "mean_token_accuracy": 0.14963005781173705,
      "num_tokens": 13110512.0,
      "step": 5990
    },
    {
      "entropy": 5.665920400619507,
      "epoch": 0.5949781659388647,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.429,
      "mean_token_accuracy": 0.149860779941082,
      "num_tokens": 13120569.0,
      "step": 5995
    },
    {
      "entropy": 5.710688781738281,
      "epoch": 0.5954743946010321,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.51,
      "mean_token_accuracy": 0.1443882778286934,
      "num_tokens": 13132006.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.514557613786084,
      "eval_loss": 5.547272682189941,
      "eval_mean_token_accuracy": 0.15050982425719975,
      "eval_num_tokens": 13132006.0,
      "eval_runtime": 20.4049,
      "eval_samples_per_second": 1650.048,
      "eval_steps_per_second": 206.274,
      "step": 6000
    },
    {
      "entropy": 5.636407136917114,
      "epoch": 0.5959706232631997,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.4212,
      "mean_token_accuracy": 0.15150929391384124,
      "num_tokens": 13143602.0,
      "step": 6005
    },
    {
      "entropy": 5.6786360263824465,
      "epoch": 0.5964668519253672,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.5065,
      "mean_token_accuracy": 0.1464373305439949,
      "num_tokens": 13154134.0,
      "step": 6010
    },
    {
      "entropy": 5.672341775894165,
      "epoch": 0.5969630805875348,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.4927,
      "mean_token_accuracy": 0.1511476755142212,
      "num_tokens": 13163460.0,
      "step": 6015
    },
    {
      "entropy": 5.7387994766235355,
      "epoch": 0.5974593092497023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.5617,
      "mean_token_accuracy": 0.1462268501520157,
      "num_tokens": 13173199.0,
      "step": 6020
    },
    {
      "entropy": 5.689201307296753,
      "epoch": 0.5979555379118698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.4924,
      "mean_token_accuracy": 0.14812639355659485,
      "num_tokens": 13183713.0,
      "step": 6025
    },
    {
      "entropy": 5.682174921035767,
      "epoch": 0.5984517665740373,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.4422,
      "mean_token_accuracy": 0.14377645552158355,
      "num_tokens": 13194483.0,
      "step": 6030
    },
    {
      "entropy": 5.744835281372071,
      "epoch": 0.5989479952362048,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.5226,
      "mean_token_accuracy": 0.1429636113345623,
      "num_tokens": 13206876.0,
      "step": 6035
    },
    {
      "entropy": 5.647784805297851,
      "epoch": 0.5994442238983724,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.4786,
      "mean_token_accuracy": 0.14632896780967714,
      "num_tokens": 13219196.0,
      "step": 6040
    },
    {
      "entropy": 5.634663867950439,
      "epoch": 0.5999404525605398,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1415524125099182,
      "num_tokens": 13230004.0,
      "step": 6045
    },
    {
      "entropy": 5.742028093338012,
      "epoch": 0.6004366812227074,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000497161284512744,
      "loss": 5.4891,
      "mean_token_accuracy": 0.14330911859869958,
      "num_tokens": 13240760.0,
      "step": 6050
    },
    {
      "entropy": 5.69619517326355,
      "epoch": 0.6009329098848749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.3768,
      "mean_token_accuracy": 0.16144133508205413,
      "num_tokens": 13250789.0,
      "step": 6055
    },
    {
      "entropy": 5.648048257827758,
      "epoch": 0.6014291385470425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.5231,
      "mean_token_accuracy": 0.15022530108690263,
      "num_tokens": 13261502.0,
      "step": 6060
    },
    {
      "entropy": 5.720826482772827,
      "epoch": 0.60192536720921,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.5108,
      "mean_token_accuracy": 0.15191833823919296,
      "num_tokens": 13272657.0,
      "step": 6065
    },
    {
      "entropy": 5.73655219078064,
      "epoch": 0.6024215958713776,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.511,
      "mean_token_accuracy": 0.15018166303634645,
      "num_tokens": 13285418.0,
      "step": 6070
    },
    {
      "entropy": 5.642021226882934,
      "epoch": 0.602917824533545,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.4639,
      "mean_token_accuracy": 0.14776795208454133,
      "num_tokens": 13296640.0,
      "step": 6075
    },
    {
      "entropy": 5.699698162078858,
      "epoch": 0.6034140531957126,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497127522474725,
      "loss": 5.5199,
      "mean_token_accuracy": 0.15716414302587509,
      "num_tokens": 13307418.0,
      "step": 6080
    },
    {
      "entropy": 5.688119840621948,
      "epoch": 0.6039102818578801,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.4121,
      "mean_token_accuracy": 0.15672177970409393,
      "num_tokens": 13317637.0,
      "step": 6085
    },
    {
      "entropy": 5.670992040634156,
      "epoch": 0.6044065105200477,
      "grad_norm": 0.875,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.4866,
      "mean_token_accuracy": 0.14688835591077803,
      "num_tokens": 13329472.0,
      "step": 6090
    },
    {
      "entropy": 5.739949369430542,
      "epoch": 0.6049027391822152,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.4797,
      "mean_token_accuracy": 0.1417884647846222,
      "num_tokens": 13340797.0,
      "step": 6095
    },
    {
      "entropy": 5.6924355030059814,
      "epoch": 0.6053989678443827,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497104904281841,
      "loss": 5.4157,
      "mean_token_accuracy": 0.14879040569067,
      "num_tokens": 13350389.0,
      "step": 6100
    },
    {
      "entropy": 5.588418054580688,
      "epoch": 0.6058951965065502,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.4388,
      "mean_token_accuracy": 0.15105883181095123,
      "num_tokens": 13361763.0,
      "step": 6105
    },
    {
      "entropy": 5.625319623947144,
      "epoch": 0.6063914251687177,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.4581,
      "mean_token_accuracy": 0.14965360164642333,
      "num_tokens": 13372229.0,
      "step": 6110
    },
    {
      "entropy": 5.813246631622315,
      "epoch": 0.6068876538308853,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.6008,
      "mean_token_accuracy": 0.1432671144604683,
      "num_tokens": 13384568.0,
      "step": 6115
    },
    {
      "entropy": 5.700366735458374,
      "epoch": 0.6073838824930528,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.4385,
      "mean_token_accuracy": 0.14977137744426727,
      "num_tokens": 13396093.0,
      "step": 6120
    },
    {
      "entropy": 5.7171636581420895,
      "epoch": 0.6078801111552203,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.6018,
      "mean_token_accuracy": 0.14215637519955635,
      "num_tokens": 13407061.0,
      "step": 6125
    },
    {
      "entropy": 5.577902603149414,
      "epoch": 0.6083763398173878,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.3291,
      "mean_token_accuracy": 0.15379580110311508,
      "num_tokens": 13417310.0,
      "step": 6130
    },
    {
      "entropy": 5.655209684371949,
      "epoch": 0.6088725684795554,
      "grad_norm": 0.875,
      "learning_rate": 0.000497065110402382,
      "loss": 5.4683,
      "mean_token_accuracy": 0.14650997817516326,
      "num_tokens": 13427187.0,
      "step": 6135
    },
    {
      "entropy": 5.670758199691773,
      "epoch": 0.6093687971417229,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.4259,
      "mean_token_accuracy": 0.15036226212978362,
      "num_tokens": 13439345.0,
      "step": 6140
    },
    {
      "entropy": 5.686715221405029,
      "epoch": 0.6098650258038905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.4667,
      "mean_token_accuracy": 0.1479718953371048,
      "num_tokens": 13448733.0,
      "step": 6145
    },
    {
      "entropy": 5.626626682281494,
      "epoch": 0.6103612544660579,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.417,
      "mean_token_accuracy": 0.1545220583677292,
      "num_tokens": 13460384.0,
      "step": 6150
    },
    {
      "entropy": 5.64982123374939,
      "epoch": 0.6108574831282255,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.5074,
      "mean_token_accuracy": 0.14217086285352706,
      "num_tokens": 13471573.0,
      "step": 6155
    },
    {
      "entropy": 5.645913934707641,
      "epoch": 0.611353711790393,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.4401,
      "mean_token_accuracy": 0.14568966925144194,
      "num_tokens": 13481550.0,
      "step": 6160
    },
    {
      "entropy": 5.675332736968994,
      "epoch": 0.6118499404525606,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.4654,
      "mean_token_accuracy": 0.14330592900514602,
      "num_tokens": 13491980.0,
      "step": 6165
    },
    {
      "entropy": 5.6928802013397215,
      "epoch": 0.6123461691147281,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.5053,
      "mean_token_accuracy": 0.14846292287111282,
      "num_tokens": 13503563.0,
      "step": 6170
    },
    {
      "entropy": 5.685325241088867,
      "epoch": 0.6128423977768956,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.4655,
      "mean_token_accuracy": 0.15476149171590806,
      "num_tokens": 13514123.0,
      "step": 6175
    },
    {
      "entropy": 5.646957683563232,
      "epoch": 0.6133386264390631,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.4572,
      "mean_token_accuracy": 0.15366399735212327,
      "num_tokens": 13524772.0,
      "step": 6180
    },
    {
      "entropy": 5.556572580337525,
      "epoch": 0.6138348551012306,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.3727,
      "mean_token_accuracy": 0.16014648824930192,
      "num_tokens": 13535423.0,
      "step": 6185
    },
    {
      "entropy": 5.692446756362915,
      "epoch": 0.6143310837633982,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.4817,
      "mean_token_accuracy": 0.14573622196912767,
      "num_tokens": 13546088.0,
      "step": 6190
    },
    {
      "entropy": 5.632522344589233,
      "epoch": 0.6148273124255657,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1472128950059414,
      "num_tokens": 13557035.0,
      "step": 6195
    },
    {
      "entropy": 5.714746141433716,
      "epoch": 0.6153235410877332,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049699049164464,
      "loss": 5.5837,
      "mean_token_accuracy": 0.1481762021780014,
      "num_tokens": 13568346.0,
      "step": 6200
    },
    {
      "entropy": 5.709953498840332,
      "epoch": 0.6158197697499007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.4189,
      "mean_token_accuracy": 0.15038838982582092,
      "num_tokens": 13579475.0,
      "step": 6205
    },
    {
      "entropy": 5.578711986541748,
      "epoch": 0.6163159984120683,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.4885,
      "mean_token_accuracy": 0.14924827218055725,
      "num_tokens": 13591944.0,
      "step": 6210
    },
    {
      "entropy": 5.627823066711426,
      "epoch": 0.6168122270742358,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.4503,
      "mean_token_accuracy": 0.15478098541498184,
      "num_tokens": 13603464.0,
      "step": 6215
    },
    {
      "entropy": 5.742721176147461,
      "epoch": 0.6173084557364034,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.4393,
      "mean_token_accuracy": 0.1509105958044529,
      "num_tokens": 13614670.0,
      "step": 6220
    },
    {
      "entropy": 5.65491132736206,
      "epoch": 0.6178046843985708,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000496961544385447,
      "loss": 5.5149,
      "mean_token_accuracy": 0.14313933849334717,
      "num_tokens": 13624963.0,
      "step": 6225
    },
    {
      "entropy": 5.6417089939117435,
      "epoch": 0.6183009130607384,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.5094,
      "mean_token_accuracy": 0.144009967148304,
      "num_tokens": 13636950.0,
      "step": 6230
    },
    {
      "entropy": 5.689790821075439,
      "epoch": 0.6187971417229059,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.4276,
      "mean_token_accuracy": 0.14905738905072213,
      "num_tokens": 13647462.0,
      "step": 6235
    },
    {
      "entropy": 5.643426418304443,
      "epoch": 0.6192933703850735,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.4053,
      "mean_token_accuracy": 0.15547471195459367,
      "num_tokens": 13657855.0,
      "step": 6240
    },
    {
      "entropy": 5.5904652118682865,
      "epoch": 0.619789599047241,
      "grad_norm": 0.890625,
      "learning_rate": 0.000496938287502209,
      "loss": 5.509,
      "mean_token_accuracy": 0.14851374328136444,
      "num_tokens": 13668942.0,
      "step": 6245
    },
    {
      "entropy": 5.65542368888855,
      "epoch": 0.6202858277094085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.454,
      "mean_token_accuracy": 0.15515210330486298,
      "num_tokens": 13678515.0,
      "step": 6250
    },
    {
      "entropy": 5.7327227115631105,
      "epoch": 0.620782056371576,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.49,
      "mean_token_accuracy": 0.15133336037397385,
      "num_tokens": 13690413.0,
      "step": 6255
    },
    {
      "entropy": 5.698162698745728,
      "epoch": 0.6212782850337435,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.4646,
      "mean_token_accuracy": 0.1533384531736374,
      "num_tokens": 13701302.0,
      "step": 6260
    },
    {
      "entropy": 5.753817462921143,
      "epoch": 0.6217745136959111,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496914942561513,
      "loss": 5.6009,
      "mean_token_accuracy": 0.14602674618363382,
      "num_tokens": 13711335.0,
      "step": 6265
    },
    {
      "entropy": 5.701289892196655,
      "epoch": 0.6222707423580786,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.4275,
      "mean_token_accuracy": 0.1502552628517151,
      "num_tokens": 13721470.0,
      "step": 6270
    },
    {
      "entropy": 5.628956270217896,
      "epoch": 0.6227669710202461,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.4526,
      "mean_token_accuracy": 0.14600661471486093,
      "num_tokens": 13731397.0,
      "step": 6275
    },
    {
      "entropy": 5.631623363494873,
      "epoch": 0.6232631996824136,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.418,
      "mean_token_accuracy": 0.15643495917320252,
      "num_tokens": 13741868.0,
      "step": 6280
    },
    {
      "entropy": 5.692174673080444,
      "epoch": 0.6237594283445812,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.4694,
      "mean_token_accuracy": 0.15341556668281556,
      "num_tokens": 13751538.0,
      "step": 6285
    },
    {
      "entropy": 5.633273267745972,
      "epoch": 0.6242556570067487,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.4703,
      "mean_token_accuracy": 0.14505940824747085,
      "num_tokens": 13762639.0,
      "step": 6290
    },
    {
      "entropy": 5.65932879447937,
      "epoch": 0.6247518856689163,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.3782,
      "mean_token_accuracy": 0.15584892630577088,
      "num_tokens": 13772721.0,
      "step": 6295
    },
    {
      "entropy": 5.665151691436767,
      "epoch": 0.6252481143310837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.5118,
      "mean_token_accuracy": 0.14421930462121962,
      "num_tokens": 13783520.0,
      "step": 6300
    },
    {
      "entropy": 5.778001689910889,
      "epoch": 0.6257443429932513,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.6009,
      "mean_token_accuracy": 0.14250447154045104,
      "num_tokens": 13793814.0,
      "step": 6305
    },
    {
      "entropy": 5.770637035369873,
      "epoch": 0.6262405716554188,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.6058,
      "mean_token_accuracy": 0.14679773151874542,
      "num_tokens": 13806298.0,
      "step": 6310
    },
    {
      "entropy": 5.6270795345306395,
      "epoch": 0.6267368003175864,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.4413,
      "mean_token_accuracy": 0.14869038611650467,
      "num_tokens": 13816585.0,
      "step": 6315
    },
    {
      "entropy": 5.674802970886231,
      "epoch": 0.6272330289797539,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.4475,
      "mean_token_accuracy": 0.14527407959103583,
      "num_tokens": 13827670.0,
      "step": 6320
    },
    {
      "entropy": 5.688569974899292,
      "epoch": 0.6277292576419214,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1535331502556801,
      "num_tokens": 13838601.0,
      "step": 6325
    },
    {
      "entropy": 5.762390613555908,
      "epoch": 0.6282254863040889,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.5516,
      "mean_token_accuracy": 0.13699284419417382,
      "num_tokens": 13849869.0,
      "step": 6330
    },
    {
      "entropy": 5.707345628738404,
      "epoch": 0.6287217149662565,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.5008,
      "mean_token_accuracy": 0.14350766241550444,
      "num_tokens": 13861930.0,
      "step": 6335
    },
    {
      "entropy": 5.5844354152679445,
      "epoch": 0.629217943628424,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.3724,
      "mean_token_accuracy": 0.15514791309833526,
      "num_tokens": 13873795.0,
      "step": 6340
    },
    {
      "entropy": 5.638135242462158,
      "epoch": 0.6297141722905915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.4219,
      "mean_token_accuracy": 0.14748230278491975,
      "num_tokens": 13883479.0,
      "step": 6345
    },
    {
      "entropy": 5.663779544830322,
      "epoch": 0.630210400952759,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.4619,
      "mean_token_accuracy": 0.14451525509357452,
      "num_tokens": 13894981.0,
      "step": 6350
    },
    {
      "entropy": 5.634831380844116,
      "epoch": 0.6307066296149265,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.4436,
      "mean_token_accuracy": 0.15460619181394578,
      "num_tokens": 13906414.0,
      "step": 6355
    },
    {
      "entropy": 5.668380975723267,
      "epoch": 0.6312028582770941,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.457,
      "mean_token_accuracy": 0.14542514458298683,
      "num_tokens": 13917500.0,
      "step": 6360
    },
    {
      "entropy": 5.689888334274292,
      "epoch": 0.6316990869392616,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.4542,
      "mean_token_accuracy": 0.1514497473835945,
      "num_tokens": 13927843.0,
      "step": 6365
    },
    {
      "entropy": 5.596258497238159,
      "epoch": 0.6321953156014292,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.3699,
      "mean_token_accuracy": 0.1531464248895645,
      "num_tokens": 13937939.0,
      "step": 6370
    },
    {
      "entropy": 5.64495062828064,
      "epoch": 0.6326915442635966,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.435,
      "mean_token_accuracy": 0.15303865522146226,
      "num_tokens": 13951374.0,
      "step": 6375
    },
    {
      "entropy": 5.574727869033813,
      "epoch": 0.6331877729257642,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.3854,
      "mean_token_accuracy": 0.1575678378343582,
      "num_tokens": 13963739.0,
      "step": 6380
    },
    {
      "entropy": 5.539268970489502,
      "epoch": 0.6336840015879317,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1523458942770958,
      "num_tokens": 13974375.0,
      "step": 6385
    },
    {
      "entropy": 5.592712020874023,
      "epoch": 0.6341802302500993,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496767042210332,
      "loss": 5.3759,
      "mean_token_accuracy": 0.1553490623831749,
      "num_tokens": 13984379.0,
      "step": 6390
    },
    {
      "entropy": 5.613452339172364,
      "epoch": 0.6346764589122668,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.3748,
      "mean_token_accuracy": 0.1488352194428444,
      "num_tokens": 13994638.0,
      "step": 6395
    },
    {
      "entropy": 5.685046100616455,
      "epoch": 0.6351726875744343,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000496755061669385,
      "loss": 5.5109,
      "mean_token_accuracy": 0.14800857827067376,
      "num_tokens": 14005835.0,
      "step": 6400
    },
    {
      "entropy": 5.617988443374633,
      "epoch": 0.6356689162366018,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.379,
      "mean_token_accuracy": 0.15325325429439546,
      "num_tokens": 14017549.0,
      "step": 6405
    },
    {
      "entropy": 5.671000576019287,
      "epoch": 0.6361651448987694,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.4458,
      "mean_token_accuracy": 0.1529954493045807,
      "num_tokens": 14028417.0,
      "step": 6410
    },
    {
      "entropy": 5.598294830322265,
      "epoch": 0.6366613735609369,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.3925,
      "mean_token_accuracy": 0.1539965257048607,
      "num_tokens": 14038535.0,
      "step": 6415
    },
    {
      "entropy": 5.684281349182129,
      "epoch": 0.6371576022231044,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.4545,
      "mean_token_accuracy": 0.15439267605543136,
      "num_tokens": 14050082.0,
      "step": 6420
    },
    {
      "entropy": 5.663456583023072,
      "epoch": 0.6376538308852719,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.4946,
      "mean_token_accuracy": 0.15071579664945603,
      "num_tokens": 14061419.0,
      "step": 6425
    },
    {
      "entropy": 5.675267314910888,
      "epoch": 0.6381500595474394,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.4497,
      "mean_token_accuracy": 0.1518915466964245,
      "num_tokens": 14073360.0,
      "step": 6430
    },
    {
      "entropy": 5.642609262466431,
      "epoch": 0.638646288209607,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.491,
      "mean_token_accuracy": 0.15084727853536606,
      "num_tokens": 14083665.0,
      "step": 6435
    },
    {
      "entropy": 5.633159160614014,
      "epoch": 0.6391425168717745,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.4276,
      "mean_token_accuracy": 0.14808937460184096,
      "num_tokens": 14094968.0,
      "step": 6440
    },
    {
      "entropy": 5.66090202331543,
      "epoch": 0.6396387455339421,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496700877049473,
      "loss": 5.4635,
      "mean_token_accuracy": 0.1452733501791954,
      "num_tokens": 14105325.0,
      "step": 6445
    },
    {
      "entropy": 5.666377878189087,
      "epoch": 0.6401349741961095,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.4662,
      "mean_token_accuracy": 0.14866262972354888,
      "num_tokens": 14115764.0,
      "step": 6450
    },
    {
      "entropy": 5.621350908279419,
      "epoch": 0.6406312028582771,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.4291,
      "mean_token_accuracy": 0.15154656767845154,
      "num_tokens": 14126389.0,
      "step": 6455
    },
    {
      "entropy": 5.670088386535644,
      "epoch": 0.6411274315204446,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.4458,
      "mean_token_accuracy": 0.14993944764137268,
      "num_tokens": 14136241.0,
      "step": 6460
    },
    {
      "entropy": 5.677780246734619,
      "epoch": 0.6416236601826122,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.4099,
      "mean_token_accuracy": 0.14660378396511078,
      "num_tokens": 14147641.0,
      "step": 6465
    },
    {
      "entropy": 5.583411312103271,
      "epoch": 0.6421198888447797,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.422,
      "mean_token_accuracy": 0.15972242653369903,
      "num_tokens": 14158780.0,
      "step": 6470
    },
    {
      "entropy": 5.565994215011597,
      "epoch": 0.6426161175069472,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.3854,
      "mean_token_accuracy": 0.15325071215629577,
      "num_tokens": 14170672.0,
      "step": 6475
    },
    {
      "entropy": 5.65286021232605,
      "epoch": 0.6431123461691147,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.3212,
      "mean_token_accuracy": 0.1561105728149414,
      "num_tokens": 14179835.0,
      "step": 6480
    },
    {
      "entropy": 5.642740678787232,
      "epoch": 0.6436085748312823,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.4489,
      "mean_token_accuracy": 0.15144938081502915,
      "num_tokens": 14190624.0,
      "step": 6485
    },
    {
      "entropy": 5.661019277572632,
      "epoch": 0.6441048034934498,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.4273,
      "mean_token_accuracy": 0.15639885812997817,
      "num_tokens": 14201350.0,
      "step": 6490
    },
    {
      "entropy": 5.630823183059692,
      "epoch": 0.6446010321556173,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1472921535372734,
      "num_tokens": 14212837.0,
      "step": 6495
    },
    {
      "entropy": 5.618098354339599,
      "epoch": 0.6450972608177848,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.424,
      "mean_token_accuracy": 0.15451672524213791,
      "num_tokens": 14223479.0,
      "step": 6500
    },
    {
      "entropy": 5.661688804626465,
      "epoch": 0.6455934894799523,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.4334,
      "mean_token_accuracy": 0.15535577684640883,
      "num_tokens": 14233735.0,
      "step": 6505
    },
    {
      "entropy": 5.760781860351562,
      "epoch": 0.6460897181421199,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.5333,
      "mean_token_accuracy": 0.14729659855365754,
      "num_tokens": 14246092.0,
      "step": 6510
    },
    {
      "entropy": 5.659719753265381,
      "epoch": 0.6465859468042874,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1503255233168602,
      "num_tokens": 14257652.0,
      "step": 6515
    },
    {
      "entropy": 5.69444727897644,
      "epoch": 0.647082175466455,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.4561,
      "mean_token_accuracy": 0.1535518139600754,
      "num_tokens": 14268033.0,
      "step": 6520
    },
    {
      "entropy": 5.722513008117676,
      "epoch": 0.6475784041286224,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.6264,
      "mean_token_accuracy": 0.1441034719347954,
      "num_tokens": 14279224.0,
      "step": 6525
    },
    {
      "entropy": 5.8233589172363285,
      "epoch": 0.64807463279079,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.545,
      "mean_token_accuracy": 0.14491087198257446,
      "num_tokens": 14288908.0,
      "step": 6530
    },
    {
      "entropy": 5.644676971435547,
      "epoch": 0.6485708614529575,
      "grad_norm": 0.859375,
      "learning_rate": 0.000496591171977522,
      "loss": 5.5145,
      "mean_token_accuracy": 0.14537544697523117,
      "num_tokens": 14299654.0,
      "step": 6535
    },
    {
      "entropy": 5.621025228500367,
      "epoch": 0.6490670901151251,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.4345,
      "mean_token_accuracy": 0.15244705975055695,
      "num_tokens": 14310689.0,
      "step": 6540
    },
    {
      "entropy": 5.63126220703125,
      "epoch": 0.6495633187772926,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.4516,
      "mean_token_accuracy": 0.1496985822916031,
      "num_tokens": 14321913.0,
      "step": 6545
    },
    {
      "entropy": 5.662426233291626,
      "epoch": 0.6500595474394602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.3937,
      "mean_token_accuracy": 0.15601938664913179,
      "num_tokens": 14332869.0,
      "step": 6550
    },
    {
      "entropy": 5.6015643119812015,
      "epoch": 0.6505557761016276,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.3535,
      "mean_token_accuracy": 0.1575906530022621,
      "num_tokens": 14344363.0,
      "step": 6555
    },
    {
      "entropy": 5.548624658584595,
      "epoch": 0.6510520047637952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.4108,
      "mean_token_accuracy": 0.15710614770650863,
      "num_tokens": 14355149.0,
      "step": 6560
    },
    {
      "entropy": 5.742516422271729,
      "epoch": 0.6515482334259627,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.5024,
      "mean_token_accuracy": 0.1474456399679184,
      "num_tokens": 14366833.0,
      "step": 6565
    },
    {
      "entropy": 5.65959324836731,
      "epoch": 0.6520444620881302,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.4434,
      "mean_token_accuracy": 0.15156612694263458,
      "num_tokens": 14376945.0,
      "step": 6570
    },
    {
      "entropy": 5.620908212661743,
      "epoch": 0.6525406907502977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.4692,
      "mean_token_accuracy": 0.146368108689785,
      "num_tokens": 14388376.0,
      "step": 6575
    },
    {
      "entropy": 5.6574407577514645,
      "epoch": 0.6530369194124652,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.4095,
      "mean_token_accuracy": 0.15505319610238075,
      "num_tokens": 14398826.0,
      "step": 6580
    },
    {
      "entropy": 5.675036859512329,
      "epoch": 0.6535331480746328,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.4983,
      "mean_token_accuracy": 0.15082188099622726,
      "num_tokens": 14410758.0,
      "step": 6585
    },
    {
      "entropy": 5.674936771392822,
      "epoch": 0.6540293767368003,
      "grad_norm": 0.921875,
      "learning_rate": 0.000496523253482968,
      "loss": 5.4282,
      "mean_token_accuracy": 0.152885240316391,
      "num_tokens": 14420097.0,
      "step": 6590
    },
    {
      "entropy": 5.681108236312866,
      "epoch": 0.6545256053989679,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.4645,
      "mean_token_accuracy": 0.15146567970514296,
      "num_tokens": 14430998.0,
      "step": 6595
    },
    {
      "entropy": 5.654100513458252,
      "epoch": 0.6550218340611353,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.3722,
      "mean_token_accuracy": 0.1596202000975609,
      "num_tokens": 14442234.0,
      "step": 6600
    },
    {
      "entropy": 5.720690727233887,
      "epoch": 0.6555180627233029,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.6626,
      "mean_token_accuracy": 0.13380816876888274,
      "num_tokens": 14453270.0,
      "step": 6605
    },
    {
      "entropy": 5.6776495456695555,
      "epoch": 0.6560142913854704,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.4159,
      "mean_token_accuracy": 0.15594621300697326,
      "num_tokens": 14463818.0,
      "step": 6610
    },
    {
      "entropy": 5.707467174530029,
      "epoch": 0.656510520047638,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.5095,
      "mean_token_accuracy": 0.1520865574479103,
      "num_tokens": 14476342.0,
      "step": 6615
    },
    {
      "entropy": 5.600308656692505,
      "epoch": 0.6570067487098055,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.3472,
      "mean_token_accuracy": 0.15081407725811005,
      "num_tokens": 14488683.0,
      "step": 6620
    },
    {
      "entropy": 5.6171181201934814,
      "epoch": 0.657502977371973,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.3601,
      "mean_token_accuracy": 0.15405872762203215,
      "num_tokens": 14499650.0,
      "step": 6625
    },
    {
      "entropy": 5.612332534790039,
      "epoch": 0.6579992060341405,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.4652,
      "mean_token_accuracy": 0.1452098786830902,
      "num_tokens": 14511390.0,
      "step": 6630
    },
    {
      "entropy": 5.652116918563843,
      "epoch": 0.6584954346963081,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.4753,
      "mean_token_accuracy": 0.14711032062768936,
      "num_tokens": 14522973.0,
      "step": 6635
    },
    {
      "entropy": 5.686647367477417,
      "epoch": 0.6589916633584756,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.4543,
      "mean_token_accuracy": 0.16025857031345367,
      "num_tokens": 14534202.0,
      "step": 6640
    },
    {
      "entropy": 5.666857433319092,
      "epoch": 0.6594878920206431,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.4535,
      "mean_token_accuracy": 0.1473967120051384,
      "num_tokens": 14545855.0,
      "step": 6645
    },
    {
      "entropy": 5.620961904525757,
      "epoch": 0.6599841206828106,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496448402511793,
      "loss": 5.4128,
      "mean_token_accuracy": 0.15478780195116998,
      "num_tokens": 14556716.0,
      "step": 6650
    },
    {
      "entropy": 5.622297859191894,
      "epoch": 0.6604803493449781,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.328,
      "mean_token_accuracy": 0.1579113334417343,
      "num_tokens": 14566680.0,
      "step": 6655
    },
    {
      "entropy": 5.645349073410034,
      "epoch": 0.6609765780071457,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.4459,
      "mean_token_accuracy": 0.1504395991563797,
      "num_tokens": 14578106.0,
      "step": 6660
    },
    {
      "entropy": 5.543342542648316,
      "epoch": 0.6614728066693132,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.3833,
      "mean_token_accuracy": 0.15685123354196548,
      "num_tokens": 14588601.0,
      "step": 6665
    },
    {
      "entropy": 5.669635200500489,
      "epoch": 0.6619690353314808,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.3898,
      "mean_token_accuracy": 0.14961127191781998,
      "num_tokens": 14599686.0,
      "step": 6670
    },
    {
      "entropy": 5.680667924880981,
      "epoch": 0.6624652639936482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.4968,
      "mean_token_accuracy": 0.1492050275206566,
      "num_tokens": 14610704.0,
      "step": 6675
    },
    {
      "entropy": 5.624146795272827,
      "epoch": 0.6629614926558158,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.4171,
      "mean_token_accuracy": 0.15624218583106994,
      "num_tokens": 14621004.0,
      "step": 6680
    },
    {
      "entropy": 5.624433946609497,
      "epoch": 0.6634577213179833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.4108,
      "mean_token_accuracy": 0.15480295121669768,
      "num_tokens": 14631584.0,
      "step": 6685
    },
    {
      "entropy": 5.631601333618164,
      "epoch": 0.6639539499801509,
      "grad_norm": 0.875,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.4506,
      "mean_token_accuracy": 0.14667699560523034,
      "num_tokens": 14642687.0,
      "step": 6690
    },
    {
      "entropy": 5.652053308486939,
      "epoch": 0.6644501786423184,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496391745327459,
      "loss": 5.4217,
      "mean_token_accuracy": 0.15319797098636628,
      "num_tokens": 14652569.0,
      "step": 6695
    },
    {
      "entropy": 5.715272426605225,
      "epoch": 0.664946407304486,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.4692,
      "mean_token_accuracy": 0.15704635232686998,
      "num_tokens": 14664018.0,
      "step": 6700
    },
    {
      "entropy": 5.5942573070526125,
      "epoch": 0.6654426359666534,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.3457,
      "mean_token_accuracy": 0.16082138419151307,
      "num_tokens": 14673416.0,
      "step": 6705
    },
    {
      "entropy": 5.5431019306182865,
      "epoch": 0.665938864628821,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.4051,
      "mean_token_accuracy": 0.15510378628969193,
      "num_tokens": 14685393.0,
      "step": 6710
    },
    {
      "entropy": 5.730377197265625,
      "epoch": 0.6664350932909885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.5035,
      "mean_token_accuracy": 0.15221727937459945,
      "num_tokens": 14697133.0,
      "step": 6715
    },
    {
      "entropy": 5.739580488204956,
      "epoch": 0.666931321953156,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.4929,
      "mean_token_accuracy": 0.15139237344264983,
      "num_tokens": 14708340.0,
      "step": 6720
    },
    {
      "entropy": 5.647880268096924,
      "epoch": 0.6674275506153236,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.4708,
      "mean_token_accuracy": 0.15366054326295853,
      "num_tokens": 14719195.0,
      "step": 6725
    },
    {
      "entropy": 5.619507646560669,
      "epoch": 0.667923779277491,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.4348,
      "mean_token_accuracy": 0.1514870524406433,
      "num_tokens": 14730999.0,
      "step": 6730
    },
    {
      "entropy": 5.666649389266968,
      "epoch": 0.6684200079396586,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.3992,
      "mean_token_accuracy": 0.16101981699466705,
      "num_tokens": 14741548.0,
      "step": 6735
    },
    {
      "entropy": 5.65791802406311,
      "epoch": 0.6689162366018261,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.476,
      "mean_token_accuracy": 0.15005208179354668,
      "num_tokens": 14753257.0,
      "step": 6740
    },
    {
      "entropy": 5.6409228324890135,
      "epoch": 0.6694124652639937,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.475,
      "mean_token_accuracy": 0.14824454337358475,
      "num_tokens": 14765340.0,
      "step": 6745
    },
    {
      "entropy": 5.695513153076172,
      "epoch": 0.6699086939261611,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.468,
      "mean_token_accuracy": 0.14830066859722138,
      "num_tokens": 14775650.0,
      "step": 6750
    },
    {
      "entropy": 5.645009231567383,
      "epoch": 0.6704049225883287,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.4136,
      "mean_token_accuracy": 0.15091665089130402,
      "num_tokens": 14786991.0,
      "step": 6755
    },
    {
      "entropy": 5.620834684371948,
      "epoch": 0.6709011512504962,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.4068,
      "mean_token_accuracy": 0.15230662152171134,
      "num_tokens": 14798992.0,
      "step": 6760
    },
    {
      "entropy": 5.685542058944702,
      "epoch": 0.6713973799126638,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000496302728075779,
      "loss": 5.4999,
      "mean_token_accuracy": 0.1488227553665638,
      "num_tokens": 14809691.0,
      "step": 6765
    },
    {
      "entropy": 5.607803773880005,
      "epoch": 0.6718936085748313,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.3259,
      "mean_token_accuracy": 0.15773825198411942,
      "num_tokens": 14820217.0,
      "step": 6770
    },
    {
      "entropy": 5.665990114212036,
      "epoch": 0.6723898372369989,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.5263,
      "mean_token_accuracy": 0.14524677395820618,
      "num_tokens": 14831030.0,
      "step": 6775
    },
    {
      "entropy": 5.619019079208374,
      "epoch": 0.6728860658991663,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.3753,
      "mean_token_accuracy": 0.16171567142009735,
      "num_tokens": 14841422.0,
      "step": 6780
    },
    {
      "entropy": 5.645073175430298,
      "epoch": 0.6733822945613339,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.3769,
      "mean_token_accuracy": 0.15103233605623245,
      "num_tokens": 14852492.0,
      "step": 6785
    },
    {
      "entropy": 5.640752696990967,
      "epoch": 0.6738785232235014,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496270675505693,
      "loss": 5.4152,
      "mean_token_accuracy": 0.15012281462550164,
      "num_tokens": 14862933.0,
      "step": 6790
    },
    {
      "entropy": 5.614436531066895,
      "epoch": 0.6743747518856689,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.3785,
      "mean_token_accuracy": 0.15544779896736144,
      "num_tokens": 14873718.0,
      "step": 6795
    },
    {
      "entropy": 5.639614725112915,
      "epoch": 0.6748709805478365,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.4354,
      "mean_token_accuracy": 0.15213156566023828,
      "num_tokens": 14886113.0,
      "step": 6800
    },
    {
      "entropy": 5.64730429649353,
      "epoch": 0.6753672092100039,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.4818,
      "mean_token_accuracy": 0.1510591670870781,
      "num_tokens": 14897162.0,
      "step": 6805
    },
    {
      "entropy": 5.701270580291748,
      "epoch": 0.6758634378721715,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.4536,
      "mean_token_accuracy": 0.15445428118109703,
      "num_tokens": 14908387.0,
      "step": 6810
    },
    {
      "entropy": 5.666077661514282,
      "epoch": 0.676359666534339,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.4811,
      "mean_token_accuracy": 0.14743232354521751,
      "num_tokens": 14919163.0,
      "step": 6815
    },
    {
      "entropy": 5.6422357082366945,
      "epoch": 0.6768558951965066,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.3388,
      "mean_token_accuracy": 0.15469984114170074,
      "num_tokens": 14929431.0,
      "step": 6820
    },
    {
      "entropy": 5.624116373062134,
      "epoch": 0.677352123858674,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.3818,
      "mean_token_accuracy": 0.1528071641921997,
      "num_tokens": 14940395.0,
      "step": 6825
    },
    {
      "entropy": 5.51373257637024,
      "epoch": 0.6778483525208416,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.3488,
      "mean_token_accuracy": 0.16014270931482316,
      "num_tokens": 14952268.0,
      "step": 6830
    },
    {
      "entropy": 5.712094497680664,
      "epoch": 0.6783445811830091,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.4934,
      "mean_token_accuracy": 0.14743401110172272,
      "num_tokens": 14964880.0,
      "step": 6835
    },
    {
      "entropy": 5.608614826202393,
      "epoch": 0.6788408098451767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.4312,
      "mean_token_accuracy": 0.15244411379098893,
      "num_tokens": 14975512.0,
      "step": 6840
    },
    {
      "entropy": 5.599345922470093,
      "epoch": 0.6793370385073442,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.4347,
      "mean_token_accuracy": 0.1578333243727684,
      "num_tokens": 14986373.0,
      "step": 6845
    },
    {
      "entropy": 5.647356414794922,
      "epoch": 0.6798332671695118,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.3793,
      "mean_token_accuracy": 0.15593601912260055,
      "num_tokens": 14997449.0,
      "step": 6850
    },
    {
      "entropy": 5.599250316619873,
      "epoch": 0.6803294958316792,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.3318,
      "mean_token_accuracy": 0.1556403636932373,
      "num_tokens": 15008424.0,
      "step": 6855
    },
    {
      "entropy": 5.658540916442871,
      "epoch": 0.6808257244938468,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.4869,
      "mean_token_accuracy": 0.1529270425438881,
      "num_tokens": 15018933.0,
      "step": 6860
    },
    {
      "entropy": 5.680090856552124,
      "epoch": 0.6813219531560143,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.5091,
      "mean_token_accuracy": 0.14452311992645264,
      "num_tokens": 15030973.0,
      "step": 6865
    },
    {
      "entropy": 5.604967975616455,
      "epoch": 0.6818181818181818,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.3518,
      "mean_token_accuracy": 0.1516082003712654,
      "num_tokens": 15042412.0,
      "step": 6870
    },
    {
      "entropy": 5.6257671356201175,
      "epoch": 0.6823144104803494,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.3683,
      "mean_token_accuracy": 0.15602472871541978,
      "num_tokens": 15053301.0,
      "step": 6875
    },
    {
      "entropy": 5.572913599014282,
      "epoch": 0.6828106391425168,
      "grad_norm": 0.875,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.3442,
      "mean_token_accuracy": 0.1557287871837616,
      "num_tokens": 15064753.0,
      "step": 6880
    },
    {
      "entropy": 5.651091289520264,
      "epoch": 0.6833068678046844,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.4448,
      "mean_token_accuracy": 0.14951328039169312,
      "num_tokens": 15074991.0,
      "step": 6885
    },
    {
      "entropy": 5.629896545410157,
      "epoch": 0.6838030964668519,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.4798,
      "mean_token_accuracy": 0.1454489827156067,
      "num_tokens": 15084567.0,
      "step": 6890
    },
    {
      "entropy": 5.6943597316741945,
      "epoch": 0.6842993251290195,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.481,
      "mean_token_accuracy": 0.15312527120113373,
      "num_tokens": 15096222.0,
      "step": 6895
    },
    {
      "entropy": 5.744639539718628,
      "epoch": 0.684795553791187,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.4506,
      "mean_token_accuracy": 0.15132685154676437,
      "num_tokens": 15107019.0,
      "step": 6900
    },
    {
      "entropy": 5.668001794815064,
      "epoch": 0.6852917824533545,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.3445,
      "mean_token_accuracy": 0.15503085255622864,
      "num_tokens": 15117695.0,
      "step": 6905
    },
    {
      "entropy": 5.5599480152130125,
      "epoch": 0.685788011115522,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.4095,
      "mean_token_accuracy": 0.15311924368143082,
      "num_tokens": 15130115.0,
      "step": 6910
    },
    {
      "entropy": 5.618098974227905,
      "epoch": 0.6862842397776896,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.3734,
      "mean_token_accuracy": 0.15242569297552108,
      "num_tokens": 15140626.0,
      "step": 6915
    },
    {
      "entropy": 5.688369369506836,
      "epoch": 0.6867804684398571,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1514328971505165,
      "num_tokens": 15151586.0,
      "step": 6920
    },
    {
      "entropy": 5.640116119384766,
      "epoch": 0.6872766971020247,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.4022,
      "mean_token_accuracy": 0.15450709015130998,
      "num_tokens": 15162662.0,
      "step": 6925
    },
    {
      "entropy": 5.611700391769409,
      "epoch": 0.6877729257641921,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.3903,
      "mean_token_accuracy": 0.15380929261446,
      "num_tokens": 15172691.0,
      "step": 6930
    },
    {
      "entropy": 5.649933624267578,
      "epoch": 0.6882691544263597,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.4601,
      "mean_token_accuracy": 0.14420425295829772,
      "num_tokens": 15184655.0,
      "step": 6935
    },
    {
      "entropy": 5.708760738372803,
      "epoch": 0.6887653830885272,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.4594,
      "mean_token_accuracy": 0.15817063748836518,
      "num_tokens": 15196031.0,
      "step": 6940
    },
    {
      "entropy": 5.710674715042114,
      "epoch": 0.6892616117506947,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.4747,
      "mean_token_accuracy": 0.14608751758933067,
      "num_tokens": 15206094.0,
      "step": 6945
    },
    {
      "entropy": 5.671129035949707,
      "epoch": 0.6897578404128623,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.5308,
      "mean_token_accuracy": 0.1472563214600086,
      "num_tokens": 15216758.0,
      "step": 6950
    },
    {
      "entropy": 5.633934164047242,
      "epoch": 0.6902540690750297,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.4073,
      "mean_token_accuracy": 0.15675952434539794,
      "num_tokens": 15226594.0,
      "step": 6955
    },
    {
      "entropy": 5.532809400558472,
      "epoch": 0.6907502977371973,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.2756,
      "mean_token_accuracy": 0.16181354969739914,
      "num_tokens": 15237225.0,
      "step": 6960
    },
    {
      "entropy": 5.6459267139434814,
      "epoch": 0.6912465263993648,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.4325,
      "mean_token_accuracy": 0.15588307455182077,
      "num_tokens": 15248215.0,
      "step": 6965
    },
    {
      "entropy": 5.638271236419678,
      "epoch": 0.6917427550615324,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.3935,
      "mean_token_accuracy": 0.15396419316530227,
      "num_tokens": 15258056.0,
      "step": 6970
    },
    {
      "entropy": 5.608838939666748,
      "epoch": 0.6922389837236999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.3112,
      "mean_token_accuracy": 0.16119545549154282,
      "num_tokens": 15268438.0,
      "step": 6975
    },
    {
      "entropy": 5.656685829162598,
      "epoch": 0.6927352123858674,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000496022594508192,
      "loss": 5.4831,
      "mean_token_accuracy": 0.15047807842493058,
      "num_tokens": 15279308.0,
      "step": 6980
    },
    {
      "entropy": 5.623466348648071,
      "epoch": 0.6932314410480349,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.4331,
      "mean_token_accuracy": 0.1543915554881096,
      "num_tokens": 15290804.0,
      "step": 6985
    },
    {
      "entropy": 5.666862154006958,
      "epoch": 0.6937276697102025,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.3945,
      "mean_token_accuracy": 0.14858111143112182,
      "num_tokens": 15302543.0,
      "step": 6990
    },
    {
      "entropy": 5.614682197570801,
      "epoch": 0.69422389837237,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.3972,
      "mean_token_accuracy": 0.15324713587760924,
      "num_tokens": 15312841.0,
      "step": 6995
    },
    {
      "entropy": 5.60940899848938,
      "epoch": 0.6947201270345376,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.3947,
      "mean_token_accuracy": 0.15326911807060242,
      "num_tokens": 15323940.0,
      "step": 7000
    },
    {
      "entropy": 5.731871891021728,
      "epoch": 0.695216355696705,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.5208,
      "mean_token_accuracy": 0.1442640818655491,
      "num_tokens": 15335094.0,
      "step": 7005
    },
    {
      "entropy": 5.681009006500244,
      "epoch": 0.6957125843588726,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.4714,
      "mean_token_accuracy": 0.14773209691047667,
      "num_tokens": 15346383.0,
      "step": 7010
    },
    {
      "entropy": 5.623705053329468,
      "epoch": 0.6962088130210401,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1609782949090004,
      "num_tokens": 15356243.0,
      "step": 7015
    },
    {
      "entropy": 5.639126014709473,
      "epoch": 0.6967050416832076,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.4617,
      "mean_token_accuracy": 0.15161948204040526,
      "num_tokens": 15367706.0,
      "step": 7020
    },
    {
      "entropy": 5.650139904022216,
      "epoch": 0.6972012703453752,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.3392,
      "mean_token_accuracy": 0.15602014511823653,
      "num_tokens": 15377208.0,
      "step": 7025
    },
    {
      "entropy": 5.573210191726685,
      "epoch": 0.6976974990075426,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.3545,
      "mean_token_accuracy": 0.15573587417602539,
      "num_tokens": 15387901.0,
      "step": 7030
    },
    {
      "entropy": 5.6079521656036375,
      "epoch": 0.6981937276697102,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.3973,
      "mean_token_accuracy": 0.15515853762626647,
      "num_tokens": 15398562.0,
      "step": 7035
    },
    {
      "entropy": 5.637833929061889,
      "epoch": 0.6986899563318777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.327,
      "mean_token_accuracy": 0.15844229310750962,
      "num_tokens": 15409498.0,
      "step": 7040
    },
    {
      "entropy": 5.603020763397216,
      "epoch": 0.6991861849940453,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.3522,
      "mean_token_accuracy": 0.1574825882911682,
      "num_tokens": 15419737.0,
      "step": 7045
    },
    {
      "entropy": 5.6288899898529055,
      "epoch": 0.6996824136562128,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.4827,
      "mean_token_accuracy": 0.14918294548988342,
      "num_tokens": 15431384.0,
      "step": 7050
    },
    {
      "entropy": 5.619852209091187,
      "epoch": 0.7001786423183803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000495922488650937,
      "loss": 5.3743,
      "mean_token_accuracy": 0.15252328366041185,
      "num_tokens": 15441253.0,
      "step": 7055
    },
    {
      "entropy": 5.526697778701783,
      "epoch": 0.7006748709805478,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.3395,
      "mean_token_accuracy": 0.15528876781463624,
      "num_tokens": 15452388.0,
      "step": 7060
    },
    {
      "entropy": 5.587579965591431,
      "epoch": 0.7011710996427154,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.4173,
      "mean_token_accuracy": 0.15193360447883605,
      "num_tokens": 15463192.0,
      "step": 7065
    },
    {
      "entropy": 5.620156526565552,
      "epoch": 0.7016673283048829,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.3811,
      "mean_token_accuracy": 0.15177375450730324,
      "num_tokens": 15473978.0,
      "step": 7070
    },
    {
      "entropy": 5.646933650970459,
      "epoch": 0.7021635569670505,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000495895585553282,
      "loss": 5.4439,
      "mean_token_accuracy": 0.1537823498249054,
      "num_tokens": 15484466.0,
      "step": 7075
    },
    {
      "entropy": 5.6090333461761475,
      "epoch": 0.7026597856292179,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.3568,
      "mean_token_accuracy": 0.1543872684240341,
      "num_tokens": 15495407.0,
      "step": 7080
    },
    {
      "entropy": 5.601065635681152,
      "epoch": 0.7031560142913855,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.3385,
      "mean_token_accuracy": 0.15164850652217865,
      "num_tokens": 15506012.0,
      "step": 7085
    },
    {
      "entropy": 5.635050678253174,
      "epoch": 0.703652242953553,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.4797,
      "mean_token_accuracy": 0.14751836359500886,
      "num_tokens": 15516829.0,
      "step": 7090
    },
    {
      "entropy": 5.695350217819214,
      "epoch": 0.7041484716157205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.4597,
      "mean_token_accuracy": 0.14946115165948867,
      "num_tokens": 15528611.0,
      "step": 7095
    },
    {
      "entropy": 5.629154300689697,
      "epoch": 0.7046447002778881,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.4142,
      "mean_token_accuracy": 0.14929258972406387,
      "num_tokens": 15540010.0,
      "step": 7100
    },
    {
      "entropy": 5.700001001358032,
      "epoch": 0.7051409289400555,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.4549,
      "mean_token_accuracy": 0.14938557147979736,
      "num_tokens": 15550670.0,
      "step": 7105
    },
    {
      "entropy": 5.647877836227417,
      "epoch": 0.7056371576022231,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.4182,
      "mean_token_accuracy": 0.14997131526470184,
      "num_tokens": 15562222.0,
      "step": 7110
    },
    {
      "entropy": 5.603520059585572,
      "epoch": 0.7061333862643906,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.3958,
      "mean_token_accuracy": 0.15193863436579705,
      "num_tokens": 15573217.0,
      "step": 7115
    },
    {
      "entropy": 5.560222864151001,
      "epoch": 0.7066296149265582,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.2629,
      "mean_token_accuracy": 0.16524704545736313,
      "num_tokens": 15584043.0,
      "step": 7120
    },
    {
      "entropy": 5.591237354278564,
      "epoch": 0.7071258435887257,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.3812,
      "mean_token_accuracy": 0.1548437297344208,
      "num_tokens": 15595844.0,
      "step": 7125
    },
    {
      "entropy": 5.617861652374268,
      "epoch": 0.7076220722508932,
      "grad_norm": 0.890625,
      "learning_rate": 0.000495821150153876,
      "loss": 5.4821,
      "mean_token_accuracy": 0.14688558578491212,
      "num_tokens": 15606748.0,
      "step": 7130
    },
    {
      "entropy": 5.583647108078003,
      "epoch": 0.7081183009130607,
      "grad_norm": 0.828125,
      "learning_rate": 0.000495814350440689,
      "loss": 5.3752,
      "mean_token_accuracy": 0.15514082461595535,
      "num_tokens": 15617415.0,
      "step": 7135
    },
    {
      "entropy": 5.6666765213012695,
      "epoch": 0.7086145295752283,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.4537,
      "mean_token_accuracy": 0.14832925349473952,
      "num_tokens": 15628953.0,
      "step": 7140
    },
    {
      "entropy": 5.594859647750854,
      "epoch": 0.7091107582373958,
      "grad_norm": 0.875,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.3269,
      "mean_token_accuracy": 0.15243512839078904,
      "num_tokens": 15638937.0,
      "step": 7145
    },
    {
      "entropy": 5.614637994766236,
      "epoch": 0.7096069868995634,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.5105,
      "mean_token_accuracy": 0.1519908756017685,
      "num_tokens": 15650522.0,
      "step": 7150
    },
    {
      "entropy": 5.639059257507324,
      "epoch": 0.7101032155617308,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.4576,
      "mean_token_accuracy": 0.14506929814815522,
      "num_tokens": 15661899.0,
      "step": 7155
    },
    {
      "entropy": 5.625447607040405,
      "epoch": 0.7105994442238984,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.3261,
      "mean_token_accuracy": 0.158203125,
      "num_tokens": 15673234.0,
      "step": 7160
    },
    {
      "entropy": 5.598785018920898,
      "epoch": 0.7110956728860659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.3628,
      "mean_token_accuracy": 0.14904156178236008,
      "num_tokens": 15684567.0,
      "step": 7165
    },
    {
      "entropy": 5.595896816253662,
      "epoch": 0.7115919015482334,
      "grad_norm": 0.875,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.4932,
      "mean_token_accuracy": 0.14844507426023484,
      "num_tokens": 15695726.0,
      "step": 7170
    },
    {
      "entropy": 5.653880167007446,
      "epoch": 0.712088130210401,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.5092,
      "mean_token_accuracy": 0.15123874694108963,
      "num_tokens": 15706598.0,
      "step": 7175
    },
    {
      "entropy": 5.666147899627686,
      "epoch": 0.7125843588725684,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.3692,
      "mean_token_accuracy": 0.15533101558685303,
      "num_tokens": 15717915.0,
      "step": 7180
    },
    {
      "entropy": 5.659155750274659,
      "epoch": 0.713080587534736,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.4344,
      "mean_token_accuracy": 0.15139723122119902,
      "num_tokens": 15730466.0,
      "step": 7185
    },
    {
      "entropy": 5.617693519592285,
      "epoch": 0.7135768161969035,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000495739192235298,
      "loss": 5.4133,
      "mean_token_accuracy": 0.1624623268842697,
      "num_tokens": 15741944.0,
      "step": 7190
    },
    {
      "entropy": 5.618962287902832,
      "epoch": 0.7140730448590711,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.4611,
      "mean_token_accuracy": 0.15146422758698463,
      "num_tokens": 15753839.0,
      "step": 7195
    },
    {
      "entropy": 5.65963225364685,
      "epoch": 0.7145692735212386,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.3899,
      "mean_token_accuracy": 0.14947567582130433,
      "num_tokens": 15765443.0,
      "step": 7200
    },
    {
      "entropy": 5.638018465042114,
      "epoch": 0.7150655021834061,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.3232,
      "mean_token_accuracy": 0.15821360796689987,
      "num_tokens": 15775134.0,
      "step": 7205
    },
    {
      "entropy": 5.549161911010742,
      "epoch": 0.7155617308455736,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.3162,
      "mean_token_accuracy": 0.15003215447068213,
      "num_tokens": 15784900.0,
      "step": 7210
    },
    {
      "entropy": 5.617185020446778,
      "epoch": 0.7160579595077412,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.4136,
      "mean_token_accuracy": 0.15256475210189818,
      "num_tokens": 15796383.0,
      "step": 7215
    },
    {
      "entropy": 5.627089786529541,
      "epoch": 0.7165541881699087,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.425,
      "mean_token_accuracy": 0.15112082660198212,
      "num_tokens": 15806879.0,
      "step": 7220
    },
    {
      "entropy": 5.645400190353394,
      "epoch": 0.7170504168320763,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.5124,
      "mean_token_accuracy": 0.15344634652137756,
      "num_tokens": 15818298.0,
      "step": 7225
    },
    {
      "entropy": 5.607115745544434,
      "epoch": 0.7175466454942437,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.3749,
      "mean_token_accuracy": 0.1519390568137169,
      "num_tokens": 15828288.0,
      "step": 7230
    },
    {
      "entropy": 5.62923789024353,
      "epoch": 0.7180428741564113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.3579,
      "mean_token_accuracy": 0.15898007154464722,
      "num_tokens": 15838275.0,
      "step": 7235
    },
    {
      "entropy": 5.696276521682739,
      "epoch": 0.7185391028185788,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.5331,
      "mean_token_accuracy": 0.1506061628460884,
      "num_tokens": 15850477.0,
      "step": 7240
    },
    {
      "entropy": 5.668019485473633,
      "epoch": 0.7190353314807463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.469,
      "mean_token_accuracy": 0.14844062328338622,
      "num_tokens": 15861620.0,
      "step": 7245
    },
    {
      "entropy": 5.6130983352661135,
      "epoch": 0.7195315601429139,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.351,
      "mean_token_accuracy": 0.15826002955436708,
      "num_tokens": 15872069.0,
      "step": 7250
    },
    {
      "entropy": 5.600798845291138,
      "epoch": 0.7200277888050813,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.3897,
      "mean_token_accuracy": 0.15497324019670486,
      "num_tokens": 15883164.0,
      "step": 7255
    },
    {
      "entropy": 5.602950096130371,
      "epoch": 0.7205240174672489,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.5056,
      "mean_token_accuracy": 0.14905053973197938,
      "num_tokens": 15893574.0,
      "step": 7260
    },
    {
      "entropy": 5.6311534404754635,
      "epoch": 0.7210202461294164,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.384,
      "mean_token_accuracy": 0.15679542273283004,
      "num_tokens": 15904999.0,
      "step": 7265
    },
    {
      "entropy": 5.752856254577637,
      "epoch": 0.721516474791584,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.4463,
      "mean_token_accuracy": 0.15702928751707076,
      "num_tokens": 15915839.0,
      "step": 7270
    },
    {
      "entropy": 5.611137056350708,
      "epoch": 0.7220127034537515,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1542869359254837,
      "num_tokens": 15927335.0,
      "step": 7275
    },
    {
      "entropy": 5.569781541824341,
      "epoch": 0.722508932115919,
      "grad_norm": 0.875,
      "learning_rate": 0.000495614777508889,
      "loss": 5.4046,
      "mean_token_accuracy": 0.15806076377630235,
      "num_tokens": 15937640.0,
      "step": 7280
    },
    {
      "entropy": 5.674206256866455,
      "epoch": 0.7230051607780865,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.3931,
      "mean_token_accuracy": 0.15428969115018845,
      "num_tokens": 15949044.0,
      "step": 7285
    },
    {
      "entropy": 5.655490350723267,
      "epoch": 0.7235013894402541,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.4608,
      "mean_token_accuracy": 0.14632925987243653,
      "num_tokens": 15961415.0,
      "step": 7290
    },
    {
      "entropy": 5.526196002960205,
      "epoch": 0.7239976181024216,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.3705,
      "mean_token_accuracy": 0.15875635296106339,
      "num_tokens": 15973689.0,
      "step": 7295
    },
    {
      "entropy": 5.615568351745606,
      "epoch": 0.7244938467645892,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.4385,
      "mean_token_accuracy": 0.1495039150118828,
      "num_tokens": 15986754.0,
      "step": 7300
    },
    {
      "entropy": 5.679232788085938,
      "epoch": 0.7249900754267566,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.3663,
      "mean_token_accuracy": 0.15502672493457795,
      "num_tokens": 15997268.0,
      "step": 7305
    },
    {
      "entropy": 5.590135145187378,
      "epoch": 0.7254863040889242,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.3575,
      "mean_token_accuracy": 0.15363836735486985,
      "num_tokens": 16008392.0,
      "step": 7310
    },
    {
      "entropy": 5.595528268814087,
      "epoch": 0.7259825327510917,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.4587,
      "mean_token_accuracy": 0.1487347014248371,
      "num_tokens": 16019674.0,
      "step": 7315
    },
    {
      "entropy": 5.628183507919312,
      "epoch": 0.7264787614132592,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.4377,
      "mean_token_accuracy": 0.1480069026350975,
      "num_tokens": 16030095.0,
      "step": 7320
    },
    {
      "entropy": 5.5880969047546385,
      "epoch": 0.7269749900754268,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.2883,
      "mean_token_accuracy": 0.15967124104499816,
      "num_tokens": 16041039.0,
      "step": 7325
    },
    {
      "entropy": 5.624117994308472,
      "epoch": 0.7274712187375942,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.3799,
      "mean_token_accuracy": 0.1528725281357765,
      "num_tokens": 16051251.0,
      "step": 7330
    },
    {
      "entropy": 5.542310190200806,
      "epoch": 0.7279674473997618,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.2788,
      "mean_token_accuracy": 0.15863536298274994,
      "num_tokens": 16061578.0,
      "step": 7335
    },
    {
      "entropy": 5.587097787857056,
      "epoch": 0.7284636760619293,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.3627,
      "mean_token_accuracy": 0.15709090381860732,
      "num_tokens": 16073532.0,
      "step": 7340
    },
    {
      "entropy": 5.5857377529144285,
      "epoch": 0.7289599047240969,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.3783,
      "mean_token_accuracy": 0.15069205611944197,
      "num_tokens": 16084689.0,
      "step": 7345
    },
    {
      "entropy": 5.620330429077148,
      "epoch": 0.7294561333862644,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.3636,
      "mean_token_accuracy": 0.15795191377401352,
      "num_tokens": 16094598.0,
      "step": 7350
    },
    {
      "entropy": 5.57075605392456,
      "epoch": 0.729952362048432,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.2985,
      "mean_token_accuracy": 0.16250543743371965,
      "num_tokens": 16105901.0,
      "step": 7355
    },
    {
      "entropy": 5.6399242877960205,
      "epoch": 0.7304485907105994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.5079,
      "mean_token_accuracy": 0.15133239775896073,
      "num_tokens": 16118060.0,
      "step": 7360
    },
    {
      "entropy": 5.61985330581665,
      "epoch": 0.730944819372767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.3053,
      "mean_token_accuracy": 0.15920507311820983,
      "num_tokens": 16129506.0,
      "step": 7365
    },
    {
      "entropy": 5.5957520484924315,
      "epoch": 0.7314410480349345,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1624893307685852,
      "num_tokens": 16140269.0,
      "step": 7370
    },
    {
      "entropy": 5.601445245742798,
      "epoch": 0.7319372766971021,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.4172,
      "mean_token_accuracy": 0.15268363505601884,
      "num_tokens": 16151139.0,
      "step": 7375
    },
    {
      "entropy": 5.612289571762085,
      "epoch": 0.7324335053592695,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.3409,
      "mean_token_accuracy": 0.16069180965423585,
      "num_tokens": 16162784.0,
      "step": 7380
    },
    {
      "entropy": 5.622614192962646,
      "epoch": 0.7329297340214371,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.3886,
      "mean_token_accuracy": 0.15400360226631166,
      "num_tokens": 16174496.0,
      "step": 7385
    },
    {
      "entropy": 5.628653907775879,
      "epoch": 0.7334259626836046,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.4382,
      "mean_token_accuracy": 0.14704638570547104,
      "num_tokens": 16186962.0,
      "step": 7390
    },
    {
      "entropy": 5.664592266082764,
      "epoch": 0.7339221913457721,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.496,
      "mean_token_accuracy": 0.15399871915578842,
      "num_tokens": 16196836.0,
      "step": 7395
    },
    {
      "entropy": 5.635286283493042,
      "epoch": 0.7344184200079397,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.4374,
      "mean_token_accuracy": 0.1515558183193207,
      "num_tokens": 16207326.0,
      "step": 7400
    },
    {
      "entropy": 5.627628040313721,
      "epoch": 0.7349146486701071,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.3761,
      "mean_token_accuracy": 0.15277141183614731,
      "num_tokens": 16218241.0,
      "step": 7405
    },
    {
      "entropy": 5.620658493041992,
      "epoch": 0.7354108773322747,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.4,
      "mean_token_accuracy": 0.15405254140496255,
      "num_tokens": 16228101.0,
      "step": 7410
    },
    {
      "entropy": 5.533949851989746,
      "epoch": 0.7359071059944422,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.2563,
      "mean_token_accuracy": 0.16437825858592986,
      "num_tokens": 16239080.0,
      "step": 7415
    },
    {
      "entropy": 5.568988418579101,
      "epoch": 0.7364033346566098,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.3084,
      "mean_token_accuracy": 0.16094768047332764,
      "num_tokens": 16250446.0,
      "step": 7420
    },
    {
      "entropy": 5.60447678565979,
      "epoch": 0.7368995633187773,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.3743,
      "mean_token_accuracy": 0.15510420650243759,
      "num_tokens": 16260414.0,
      "step": 7425
    },
    {
      "entropy": 5.612509727478027,
      "epoch": 0.7373957919809448,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.3885,
      "mean_token_accuracy": 0.15415119156241416,
      "num_tokens": 16270925.0,
      "step": 7430
    },
    {
      "entropy": 5.661929607391357,
      "epoch": 0.7378920206431123,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.4573,
      "mean_token_accuracy": 0.14561386108398439,
      "num_tokens": 16282761.0,
      "step": 7435
    },
    {
      "entropy": 5.584696054458618,
      "epoch": 0.7383882493052799,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.3167,
      "mean_token_accuracy": 0.15656638890504837,
      "num_tokens": 16294901.0,
      "step": 7440
    },
    {
      "entropy": 5.525906658172607,
      "epoch": 0.7388844779674474,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.2572,
      "mean_token_accuracy": 0.1594362586736679,
      "num_tokens": 16306310.0,
      "step": 7445
    },
    {
      "entropy": 5.601056337356567,
      "epoch": 0.739380706629615,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.4429,
      "mean_token_accuracy": 0.1526361346244812,
      "num_tokens": 16317325.0,
      "step": 7450
    },
    {
      "entropy": 5.6280937671661375,
      "epoch": 0.7398769352917824,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.3944,
      "mean_token_accuracy": 0.15684347599744797,
      "num_tokens": 16328115.0,
      "step": 7455
    },
    {
      "entropy": 5.609901237487793,
      "epoch": 0.74037316395395,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.3968,
      "mean_token_accuracy": 0.1560841456055641,
      "num_tokens": 16340692.0,
      "step": 7460
    },
    {
      "entropy": 5.640773582458496,
      "epoch": 0.7408693926161175,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000495353470741438,
      "loss": 5.3892,
      "mean_token_accuracy": 0.15599947273731232,
      "num_tokens": 16351704.0,
      "step": 7465
    },
    {
      "entropy": 5.591593456268311,
      "epoch": 0.741365621278285,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000495346304534622,
      "loss": 5.3835,
      "mean_token_accuracy": 0.15311512351036072,
      "num_tokens": 16363443.0,
      "step": 7470
    },
    {
      "entropy": 5.502955293655395,
      "epoch": 0.7418618499404526,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.298,
      "mean_token_accuracy": 0.1584179311990738,
      "num_tokens": 16373913.0,
      "step": 7475
    },
    {
      "entropy": 5.586470603942871,
      "epoch": 0.74235807860262,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.3959,
      "mean_token_accuracy": 0.15616491436958313,
      "num_tokens": 16384607.0,
      "step": 7480
    },
    {
      "entropy": 5.668504333496093,
      "epoch": 0.7428543072647876,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.4035,
      "mean_token_accuracy": 0.15533463209867476,
      "num_tokens": 16394769.0,
      "step": 7485
    },
    {
      "entropy": 5.540896797180176,
      "epoch": 0.7433505359269551,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.3224,
      "mean_token_accuracy": 0.15265467762947083,
      "num_tokens": 16404906.0,
      "step": 7490
    },
    {
      "entropy": 5.587875699996948,
      "epoch": 0.7438467645891227,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.4195,
      "mean_token_accuracy": 0.14555711299180984,
      "num_tokens": 16415779.0,
      "step": 7495
    },
    {
      "entropy": 5.684941673278809,
      "epoch": 0.7443429932512902,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.468,
      "mean_token_accuracy": 0.15213167369365693,
      "num_tokens": 16426612.0,
      "step": 7500
    },
    {
      "entropy": 5.682243394851684,
      "epoch": 0.7448392219134577,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.4462,
      "mean_token_accuracy": 0.14743935763835908,
      "num_tokens": 16437168.0,
      "step": 7505
    },
    {
      "entropy": 5.652225971221924,
      "epoch": 0.7453354505756252,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.3284,
      "mean_token_accuracy": 0.1516037866473198,
      "num_tokens": 16448073.0,
      "step": 7510
    },
    {
      "entropy": 5.546012926101684,
      "epoch": 0.7458316792377928,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.2682,
      "mean_token_accuracy": 0.16116264462471008,
      "num_tokens": 16458271.0,
      "step": 7515
    },
    {
      "entropy": 5.634888744354248,
      "epoch": 0.7463279078999603,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.5026,
      "mean_token_accuracy": 0.14879633635282516,
      "num_tokens": 16469667.0,
      "step": 7520
    },
    {
      "entropy": 5.677099084854126,
      "epoch": 0.7468241365621279,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.406,
      "mean_token_accuracy": 0.15829214900732042,
      "num_tokens": 16479449.0,
      "step": 7525
    },
    {
      "entropy": 5.666077995300293,
      "epoch": 0.7473203652242953,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.4791,
      "mean_token_accuracy": 0.15114178210496904,
      "num_tokens": 16490754.0,
      "step": 7530
    },
    {
      "entropy": 5.622817087173462,
      "epoch": 0.7478165938864629,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.3121,
      "mean_token_accuracy": 0.15742165148258208,
      "num_tokens": 16500017.0,
      "step": 7535
    },
    {
      "entropy": 5.665043354034424,
      "epoch": 0.7483128225486304,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.4489,
      "mean_token_accuracy": 0.1529522120952606,
      "num_tokens": 16512155.0,
      "step": 7540
    },
    {
      "entropy": 5.640869760513306,
      "epoch": 0.7488090512107979,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.4386,
      "mean_token_accuracy": 0.15592170655727386,
      "num_tokens": 16522735.0,
      "step": 7545
    },
    {
      "entropy": 5.588991308212281,
      "epoch": 0.7493052798729655,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.3197,
      "mean_token_accuracy": 0.16115929782390595,
      "num_tokens": 16533031.0,
      "step": 7550
    },
    {
      "entropy": 5.616295671463012,
      "epoch": 0.7498015085351329,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.4141,
      "mean_token_accuracy": 0.15109892338514327,
      "num_tokens": 16544106.0,
      "step": 7555
    },
    {
      "entropy": 5.5860758304595945,
      "epoch": 0.7502977371973005,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.3553,
      "mean_token_accuracy": 0.15483469069004058,
      "num_tokens": 16554585.0,
      "step": 7560
    },
    {
      "entropy": 5.577961206436157,
      "epoch": 0.750793965859468,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.3283,
      "mean_token_accuracy": 0.15794429928064346,
      "num_tokens": 16565660.0,
      "step": 7565
    },
    {
      "entropy": 5.643661785125732,
      "epoch": 0.7512901945216356,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.4965,
      "mean_token_accuracy": 0.1492744967341423,
      "num_tokens": 16576381.0,
      "step": 7570
    },
    {
      "entropy": 5.657181692123413,
      "epoch": 0.7517864231838031,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.443,
      "mean_token_accuracy": 0.14963263869285584,
      "num_tokens": 16587100.0,
      "step": 7575
    },
    {
      "entropy": 5.692304563522339,
      "epoch": 0.7522826518459707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.4036,
      "mean_token_accuracy": 0.15225370228290558,
      "num_tokens": 16598195.0,
      "step": 7580
    },
    {
      "entropy": 5.623695373535156,
      "epoch": 0.7527788805081381,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1622891202569008,
      "num_tokens": 16609225.0,
      "step": 7585
    },
    {
      "entropy": 5.594271087646485,
      "epoch": 0.7532751091703057,
      "grad_norm": 0.8125,
      "learning_rate": 0.000495172676747727,
      "loss": 5.404,
      "mean_token_accuracy": 0.15877650380134584,
      "num_tokens": 16620874.0,
      "step": 7590
    },
    {
      "entropy": 5.664577627182007,
      "epoch": 0.7537713378324732,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.4091,
      "mean_token_accuracy": 0.15990065783262253,
      "num_tokens": 16631758.0,
      "step": 7595
    },
    {
      "entropy": 5.666130876541137,
      "epoch": 0.7542675664946408,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.4119,
      "mean_token_accuracy": 0.15196043625473976,
      "num_tokens": 16642269.0,
      "step": 7600
    },
    {
      "entropy": 5.581426525115967,
      "epoch": 0.7547637951568082,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.2848,
      "mean_token_accuracy": 0.15674170553684236,
      "num_tokens": 16653842.0,
      "step": 7605
    },
    {
      "entropy": 5.584940624237061,
      "epoch": 0.7552600238189758,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1549353688955307,
      "num_tokens": 16664625.0,
      "step": 7610
    },
    {
      "entropy": 5.598399066925049,
      "epoch": 0.7557562524811433,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.4046,
      "mean_token_accuracy": 0.1578456997871399,
      "num_tokens": 16677076.0,
      "step": 7615
    },
    {
      "entropy": 5.630158805847168,
      "epoch": 0.7562524811433108,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.4247,
      "mean_token_accuracy": 0.14916593059897423,
      "num_tokens": 16688310.0,
      "step": 7620
    },
    {
      "entropy": 5.681772089004516,
      "epoch": 0.7567487098054784,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.5129,
      "mean_token_accuracy": 0.15009233355522156,
      "num_tokens": 16699992.0,
      "step": 7625
    },
    {
      "entropy": 5.790677738189697,
      "epoch": 0.7572449384676458,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.4238,
      "mean_token_accuracy": 0.15772709250450134,
      "num_tokens": 16710815.0,
      "step": 7630
    },
    {
      "entropy": 5.5672379493713375,
      "epoch": 0.7577411671298134,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.3439,
      "mean_token_accuracy": 0.16158160716295242,
      "num_tokens": 16721656.0,
      "step": 7635
    },
    {
      "entropy": 5.484818649291992,
      "epoch": 0.7582373957919809,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049509940352639,
      "loss": 5.2978,
      "mean_token_accuracy": 0.15768879652023315,
      "num_tokens": 16730977.0,
      "step": 7640
    },
    {
      "entropy": 5.584550285339356,
      "epoch": 0.7587336244541485,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.386,
      "mean_token_accuracy": 0.15792076587677,
      "num_tokens": 16741569.0,
      "step": 7645
    },
    {
      "entropy": 5.610457992553711,
      "epoch": 0.759229853116316,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.2926,
      "mean_token_accuracy": 0.15829338282346725,
      "num_tokens": 16752300.0,
      "step": 7650
    },
    {
      "entropy": 5.584188556671142,
      "epoch": 0.7597260817784836,
      "grad_norm": 0.875,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.3961,
      "mean_token_accuracy": 0.15859198719263076,
      "num_tokens": 16763202.0,
      "step": 7655
    },
    {
      "entropy": 5.587795352935791,
      "epoch": 0.760222310440651,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.3342,
      "mean_token_accuracy": 0.15597645193338394,
      "num_tokens": 16773792.0,
      "step": 7660
    },
    {
      "entropy": 5.6280210494995115,
      "epoch": 0.7607185391028186,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000495062562229694,
      "loss": 5.3998,
      "mean_token_accuracy": 0.15482239723205565,
      "num_tokens": 16785669.0,
      "step": 7665
    },
    {
      "entropy": 5.592633485794067,
      "epoch": 0.7612147677649861,
      "grad_norm": 0.7421875,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.3371,
      "mean_token_accuracy": 0.1603778690099716,
      "num_tokens": 16797141.0,
      "step": 7670
    },
    {
      "entropy": 5.565252637863159,
      "epoch": 0.7617109964271537,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.3759,
      "mean_token_accuracy": 0.155707186460495,
      "num_tokens": 16807904.0,
      "step": 7675
    },
    {
      "entropy": 5.593888282775879,
      "epoch": 0.7622072250893211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.4097,
      "mean_token_accuracy": 0.15516852140426635,
      "num_tokens": 16817270.0,
      "step": 7680
    },
    {
      "entropy": 5.655299139022827,
      "epoch": 0.7627034537514887,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.4095,
      "mean_token_accuracy": 0.15245045721530914,
      "num_tokens": 16828770.0,
      "step": 7685
    },
    {
      "entropy": 5.652985906600952,
      "epoch": 0.7631996824136562,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495025584506071,
      "loss": 5.4037,
      "mean_token_accuracy": 0.1481118842959404,
      "num_tokens": 16839811.0,
      "step": 7690
    },
    {
      "entropy": 5.610853862762451,
      "epoch": 0.7636959110758237,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.3908,
      "mean_token_accuracy": 0.15595915019512177,
      "num_tokens": 16850180.0,
      "step": 7695
    },
    {
      "entropy": 5.608128547668457,
      "epoch": 0.7641921397379913,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.3291,
      "mean_token_accuracy": 0.1587834984064102,
      "num_tokens": 16862401.0,
      "step": 7700
    },
    {
      "entropy": 5.646802854537964,
      "epoch": 0.7646883684001587,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.4246,
      "mean_token_accuracy": 0.14683643430471421,
      "num_tokens": 16872733.0,
      "step": 7705
    },
    {
      "entropy": 5.6853796482086185,
      "epoch": 0.7651845970623263,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.4752,
      "mean_token_accuracy": 0.14573515430092812,
      "num_tokens": 16883291.0,
      "step": 7710
    },
    {
      "entropy": 5.6609711170196535,
      "epoch": 0.7656808257244938,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.4196,
      "mean_token_accuracy": 0.15883346498012543,
      "num_tokens": 16894591.0,
      "step": 7715
    },
    {
      "entropy": 5.625853395462036,
      "epoch": 0.7661770543866614,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000494981031186463,
      "loss": 5.3935,
      "mean_token_accuracy": 0.15369686037302016,
      "num_tokens": 16905239.0,
      "step": 7720
    },
    {
      "entropy": 5.580170583724976,
      "epoch": 0.7666732830488289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.3167,
      "mean_token_accuracy": 0.15988347232341765,
      "num_tokens": 16914711.0,
      "step": 7725
    },
    {
      "entropy": 5.681525135040284,
      "epoch": 0.7671695117109965,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.4336,
      "mean_token_accuracy": 0.15401426404714585,
      "num_tokens": 16926184.0,
      "step": 7730
    },
    {
      "entropy": 5.638946390151977,
      "epoch": 0.7676657403731639,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1429605558514595,
      "num_tokens": 16937138.0,
      "step": 7735
    },
    {
      "entropy": 5.65322322845459,
      "epoch": 0.7681619690353315,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.3729,
      "mean_token_accuracy": 0.15339355170726776,
      "num_tokens": 16948797.0,
      "step": 7740
    },
    {
      "entropy": 5.715353393554688,
      "epoch": 0.768658197697499,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000494943753404413,
      "loss": 5.4642,
      "mean_token_accuracy": 0.14759013280272484,
      "num_tokens": 16959407.0,
      "step": 7745
    },
    {
      "entropy": 5.545732402801514,
      "epoch": 0.7691544263596666,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.4071,
      "mean_token_accuracy": 0.15611019730567932,
      "num_tokens": 16970600.0,
      "step": 7750
    },
    {
      "entropy": 5.626447916030884,
      "epoch": 0.769650655021834,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.406,
      "mean_token_accuracy": 0.15243962407112122,
      "num_tokens": 16980167.0,
      "step": 7755
    },
    {
      "entropy": 5.645298004150391,
      "epoch": 0.7701468836840016,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.3128,
      "mean_token_accuracy": 0.16106123626232147,
      "num_tokens": 16992209.0,
      "step": 7760
    },
    {
      "entropy": 5.582419490814209,
      "epoch": 0.7706431123461691,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000494913833003305,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1536666013300419,
      "num_tokens": 17003508.0,
      "step": 7765
    },
    {
      "entropy": 5.585568475723266,
      "epoch": 0.7711393410083366,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494906339269091,
      "loss": 5.3977,
      "mean_token_accuracy": 0.15811370611190795,
      "num_tokens": 17013160.0,
      "step": 7770
    },
    {
      "entropy": 5.648335266113281,
      "epoch": 0.7716355696705042,
      "grad_norm": 0.875,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.3239,
      "mean_token_accuracy": 0.1611305519938469,
      "num_tokens": 17023601.0,
      "step": 7775
    },
    {
      "entropy": 5.62475266456604,
      "epoch": 0.7721317983326716,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.3835,
      "mean_token_accuracy": 0.15101407766342162,
      "num_tokens": 17035530.0,
      "step": 7780
    },
    {
      "entropy": 5.58845591545105,
      "epoch": 0.7726280269948392,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.3093,
      "mean_token_accuracy": 0.16217434853315355,
      "num_tokens": 17046059.0,
      "step": 7785
    },
    {
      "entropy": 5.6219024658203125,
      "epoch": 0.7731242556570067,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.3811,
      "mean_token_accuracy": 0.15646445006132126,
      "num_tokens": 17056231.0,
      "step": 7790
    },
    {
      "entropy": 5.606713676452637,
      "epoch": 0.7736204843191743,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.3501,
      "mean_token_accuracy": 0.15891869515180587,
      "num_tokens": 17067144.0,
      "step": 7795
    },
    {
      "entropy": 5.577904081344604,
      "epoch": 0.7741167129813418,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.4022,
      "mean_token_accuracy": 0.15044188648462295,
      "num_tokens": 17077276.0,
      "step": 7800
    },
    {
      "entropy": 5.6347136974334715,
      "epoch": 0.7746129416435094,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.4548,
      "mean_token_accuracy": 0.1524588242173195,
      "num_tokens": 17089022.0,
      "step": 7805
    },
    {
      "entropy": 5.629032325744629,
      "epoch": 0.7751091703056768,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.4035,
      "mean_token_accuracy": 0.1579777330160141,
      "num_tokens": 17100657.0,
      "step": 7810
    },
    {
      "entropy": 5.6048362255096436,
      "epoch": 0.7756053989678444,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.3967,
      "mean_token_accuracy": 0.15557526648044587,
      "num_tokens": 17111203.0,
      "step": 7815
    },
    {
      "entropy": 5.65199909210205,
      "epoch": 0.7761016276300119,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.455,
      "mean_token_accuracy": 0.14781221970915795,
      "num_tokens": 17121999.0,
      "step": 7820
    },
    {
      "entropy": 5.6485401630401615,
      "epoch": 0.7765978562921795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.3759,
      "mean_token_accuracy": 0.154957877099514,
      "num_tokens": 17131872.0,
      "step": 7825
    },
    {
      "entropy": 5.586627244949341,
      "epoch": 0.777094084954347,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.3925,
      "mean_token_accuracy": 0.155118627846241,
      "num_tokens": 17143108.0,
      "step": 7830
    },
    {
      "entropy": 5.631199693679809,
      "epoch": 0.7775903136165145,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.3196,
      "mean_token_accuracy": 0.15906967222690582,
      "num_tokens": 17154034.0,
      "step": 7835
    },
    {
      "entropy": 5.6216942310333256,
      "epoch": 0.778086542278682,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049480085448854,
      "loss": 5.4446,
      "mean_token_accuracy": 0.15492319762706758,
      "num_tokens": 17164526.0,
      "step": 7840
    },
    {
      "entropy": 5.613617897033691,
      "epoch": 0.7785827709408495,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.3503,
      "mean_token_accuracy": 0.15849405378103257,
      "num_tokens": 17175653.0,
      "step": 7845
    },
    {
      "entropy": 5.564108180999756,
      "epoch": 0.7790789996030171,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494785698012787,
      "loss": 5.2883,
      "mean_token_accuracy": 0.16016782969236373,
      "num_tokens": 17186781.0,
      "step": 7850
    },
    {
      "entropy": 5.594603157043457,
      "epoch": 0.7795752282651845,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.3332,
      "mean_token_accuracy": 0.15670111179351806,
      "num_tokens": 17197375.0,
      "step": 7855
    },
    {
      "entropy": 5.639225149154663,
      "epoch": 0.7800714569273521,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.3815,
      "mean_token_accuracy": 0.15060267001390457,
      "num_tokens": 17208153.0,
      "step": 7860
    },
    {
      "entropy": 5.550888204574585,
      "epoch": 0.7805676855895196,
      "grad_norm": 0.875,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.3712,
      "mean_token_accuracy": 0.15728954672813417,
      "num_tokens": 17218914.0,
      "step": 7865
    },
    {
      "entropy": 5.634226894378662,
      "epoch": 0.7810639142516872,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.3626,
      "mean_token_accuracy": 0.15417295098304748,
      "num_tokens": 17229241.0,
      "step": 7870
    },
    {
      "entropy": 5.63611068725586,
      "epoch": 0.7815601429138547,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.428,
      "mean_token_accuracy": 0.15389227420091628,
      "num_tokens": 17239869.0,
      "step": 7875
    },
    {
      "entropy": 5.612786388397216,
      "epoch": 0.7820563715760223,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.347,
      "mean_token_accuracy": 0.16071650236845017,
      "num_tokens": 17251761.0,
      "step": 7880
    },
    {
      "entropy": 5.684227085113525,
      "epoch": 0.7825526002381897,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.4798,
      "mean_token_accuracy": 0.1493592232465744,
      "num_tokens": 17262812.0,
      "step": 7885
    },
    {
      "entropy": 5.630559062957763,
      "epoch": 0.7830488289003573,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.3433,
      "mean_token_accuracy": 0.15091573894023896,
      "num_tokens": 17273349.0,
      "step": 7890
    },
    {
      "entropy": 5.552906465530396,
      "epoch": 0.7835450575625248,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1502127081155777,
      "num_tokens": 17284187.0,
      "step": 7895
    },
    {
      "entropy": 5.591171932220459,
      "epoch": 0.7840412862246924,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.3434,
      "mean_token_accuracy": 0.15899887830018997,
      "num_tokens": 17295087.0,
      "step": 7900
    },
    {
      "entropy": 5.721424961090088,
      "epoch": 0.7845375148868599,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.3739,
      "mean_token_accuracy": 0.14994637817144393,
      "num_tokens": 17305375.0,
      "step": 7905
    },
    {
      "entropy": 5.638518762588501,
      "epoch": 0.7850337435490274,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.4362,
      "mean_token_accuracy": 0.15270688086748124,
      "num_tokens": 17315618.0,
      "step": 7910
    },
    {
      "entropy": 5.575199031829834,
      "epoch": 0.7855299722111949,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1530621036887169,
      "num_tokens": 17327063.0,
      "step": 7915
    },
    {
      "entropy": 5.589395236968994,
      "epoch": 0.7860262008733624,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15493138134479523,
      "num_tokens": 17338229.0,
      "step": 7920
    },
    {
      "entropy": 5.562542486190796,
      "epoch": 0.78652242953553,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.317,
      "mean_token_accuracy": 0.15572996884584428,
      "num_tokens": 17348917.0,
      "step": 7925
    },
    {
      "entropy": 5.583196687698364,
      "epoch": 0.7870186581976975,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.3296,
      "mean_token_accuracy": 0.15491889417171478,
      "num_tokens": 17359534.0,
      "step": 7930
    },
    {
      "entropy": 5.6295756816864015,
      "epoch": 0.787514886859865,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.4015,
      "mean_token_accuracy": 0.15565899312496184,
      "num_tokens": 17369758.0,
      "step": 7935
    },
    {
      "entropy": 5.6727861881256105,
      "epoch": 0.7880111155220325,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.4072,
      "mean_token_accuracy": 0.15707187727093697,
      "num_tokens": 17380541.0,
      "step": 7940
    },
    {
      "entropy": 5.646803140640259,
      "epoch": 0.7885073441842001,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.5134,
      "mean_token_accuracy": 0.15290991514921187,
      "num_tokens": 17391586.0,
      "step": 7945
    },
    {
      "entropy": 5.562468004226685,
      "epoch": 0.7890035728463676,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.3824,
      "mean_token_accuracy": 0.15632936656475066,
      "num_tokens": 17401324.0,
      "step": 7950
    },
    {
      "entropy": 5.607072496414185,
      "epoch": 0.7894998015085352,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.3879,
      "mean_token_accuracy": 0.15087226778268814,
      "num_tokens": 17413785.0,
      "step": 7955
    },
    {
      "entropy": 5.660722970962524,
      "epoch": 0.7899960301707026,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.4237,
      "mean_token_accuracy": 0.15096158236265184,
      "num_tokens": 17424571.0,
      "step": 7960
    },
    {
      "entropy": 5.586567020416259,
      "epoch": 0.7904922588328702,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000494609831985801,
      "loss": 5.324,
      "mean_token_accuracy": 0.15769710391759872,
      "num_tokens": 17435636.0,
      "step": 7965
    },
    {
      "entropy": 5.591307544708252,
      "epoch": 0.7909884874950377,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00049460212026984,
      "loss": 5.4149,
      "mean_token_accuracy": 0.15452817380428313,
      "num_tokens": 17446827.0,
      "step": 7970
    },
    {
      "entropy": 5.627238798141479,
      "epoch": 0.7914847161572053,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.492,
      "mean_token_accuracy": 0.14939460456371306,
      "num_tokens": 17458381.0,
      "step": 7975
    },
    {
      "entropy": 5.6771315574646,
      "epoch": 0.7919809448193728,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000494586680501131,
      "loss": 5.4246,
      "mean_token_accuracy": 0.15228963941335677,
      "num_tokens": 17469686.0,
      "step": 7980
    },
    {
      "entropy": 5.671817588806152,
      "epoch": 0.7924771734815403,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000494578952448766,
      "loss": 5.3797,
      "mean_token_accuracy": 0.15585313588380814,
      "num_tokens": 17481546.0,
      "step": 7985
    },
    {
      "entropy": 5.576861715316772,
      "epoch": 0.7929734021437078,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.3426,
      "mean_token_accuracy": 0.16077023446559907,
      "num_tokens": 17492245.0,
      "step": 7990
    },
    {
      "entropy": 5.568276643753052,
      "epoch": 0.7934696308058753,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.3692,
      "mean_token_accuracy": 0.16291145086288453,
      "num_tokens": 17502159.0,
      "step": 7995
    },
    {
      "entropy": 5.605306768417359,
      "epoch": 0.7939658594680429,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.4293,
      "mean_token_accuracy": 0.15206791535019876,
      "num_tokens": 17514576.0,
      "step": 8000
    },
    {
      "entropy": 5.640516138076782,
      "epoch": 0.7944620881302104,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.3336,
      "mean_token_accuracy": 0.15900408029556273,
      "num_tokens": 17525663.0,
      "step": 8005
    },
    {
      "entropy": 5.587938928604126,
      "epoch": 0.7949583167923779,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.4313,
      "mean_token_accuracy": 0.15825108885765077,
      "num_tokens": 17537955.0,
      "step": 8010
    },
    {
      "entropy": 5.556979179382324,
      "epoch": 0.7954545454545454,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.4029,
      "mean_token_accuracy": 0.1552339270710945,
      "num_tokens": 17548904.0,
      "step": 8015
    },
    {
      "entropy": 5.660978698730469,
      "epoch": 0.795950774116713,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.3853,
      "mean_token_accuracy": 0.15321238487958908,
      "num_tokens": 17560274.0,
      "step": 8020
    },
    {
      "entropy": 5.606451082229614,
      "epoch": 0.7964470027788805,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.4523,
      "mean_token_accuracy": 0.14679772555828094,
      "num_tokens": 17571308.0,
      "step": 8025
    },
    {
      "entropy": 5.613513851165772,
      "epoch": 0.7969432314410481,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.3426,
      "mean_token_accuracy": 0.14905406534671783,
      "num_tokens": 17582446.0,
      "step": 8030
    },
    {
      "entropy": 5.6312542915344235,
      "epoch": 0.7974394601032155,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.3708,
      "mean_token_accuracy": 0.1545553535223007,
      "num_tokens": 17594397.0,
      "step": 8035
    },
    {
      "entropy": 5.651876974105835,
      "epoch": 0.7979356887653831,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.4411,
      "mean_token_accuracy": 0.15946142226457596,
      "num_tokens": 17605131.0,
      "step": 8040
    },
    {
      "entropy": 5.634470987319946,
      "epoch": 0.7984319174275506,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.3192,
      "mean_token_accuracy": 0.15912223756313323,
      "num_tokens": 17615958.0,
      "step": 8045
    },
    {
      "entropy": 5.5749188423156735,
      "epoch": 0.7989281460897182,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.243,
      "mean_token_accuracy": 0.15980072617530822,
      "num_tokens": 17626571.0,
      "step": 8050
    },
    {
      "entropy": 5.663297986984253,
      "epoch": 0.7994243747518857,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.3733,
      "mean_token_accuracy": 0.15583392679691316,
      "num_tokens": 17638158.0,
      "step": 8055
    },
    {
      "entropy": 5.583486270904541,
      "epoch": 0.7999206034140532,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494462378360746,
      "loss": 5.4476,
      "mean_token_accuracy": 0.1550405353307724,
      "num_tokens": 17649384.0,
      "step": 8060
    },
    {
      "entropy": 5.613505077362061,
      "epoch": 0.8004168320762207,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.3699,
      "mean_token_accuracy": 0.15804558992385864,
      "num_tokens": 17661311.0,
      "step": 8065
    },
    {
      "entropy": 5.684753465652466,
      "epoch": 0.8009130607383882,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.3948,
      "mean_token_accuracy": 0.1540650889277458,
      "num_tokens": 17672710.0,
      "step": 8070
    },
    {
      "entropy": 5.65916805267334,
      "epoch": 0.8014092894005558,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494438916583036,
      "loss": 5.5062,
      "mean_token_accuracy": 0.1464451640844345,
      "num_tokens": 17683521.0,
      "step": 8075
    },
    {
      "entropy": 5.631662464141845,
      "epoch": 0.8019055180627233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.3362,
      "mean_token_accuracy": 0.15830335170030593,
      "num_tokens": 17693183.0,
      "step": 8080
    },
    {
      "entropy": 5.611684513092041,
      "epoch": 0.8024017467248908,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.3272,
      "mean_token_accuracy": 0.16114828288555144,
      "num_tokens": 17704503.0,
      "step": 8085
    },
    {
      "entropy": 5.574112319946289,
      "epoch": 0.8028979753870583,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494415405830816,
      "loss": 5.4399,
      "mean_token_accuracy": 0.15436345487833023,
      "num_tokens": 17715902.0,
      "step": 8090
    },
    {
      "entropy": 5.628669166564942,
      "epoch": 0.8033942040492259,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.382,
      "mean_token_accuracy": 0.15561255514621736,
      "num_tokens": 17727305.0,
      "step": 8095
    },
    {
      "entropy": 5.607144403457641,
      "epoch": 0.8038904327113934,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.3314,
      "mean_token_accuracy": 0.16143729090690612,
      "num_tokens": 17737906.0,
      "step": 8100
    },
    {
      "entropy": 5.635290479660034,
      "epoch": 0.804386661373561,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.3682,
      "mean_token_accuracy": 0.14989276826381684,
      "num_tokens": 17749503.0,
      "step": 8105
    },
    {
      "entropy": 5.568725299835205,
      "epoch": 0.8048828900357284,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.3508,
      "mean_token_accuracy": 0.15871076583862304,
      "num_tokens": 17759774.0,
      "step": 8110
    },
    {
      "entropy": 5.581037425994873,
      "epoch": 0.805379118697896,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.4202,
      "mean_token_accuracy": 0.1593910351395607,
      "num_tokens": 17772101.0,
      "step": 8115
    },
    {
      "entropy": 5.6613719940185545,
      "epoch": 0.8058753473600635,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.3735,
      "mean_token_accuracy": 0.15504423528909683,
      "num_tokens": 17781923.0,
      "step": 8120
    },
    {
      "entropy": 5.566691446304321,
      "epoch": 0.8063715760222311,
      "grad_norm": 0.796875,
      "learning_rate": 0.000494360356982038,
      "loss": 5.4037,
      "mean_token_accuracy": 0.15359729528427124,
      "num_tokens": 17794787.0,
      "step": 8125
    },
    {
      "entropy": 5.6643133640289305,
      "epoch": 0.8068678046843986,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.434,
      "mean_token_accuracy": 0.15398866832256317,
      "num_tokens": 17805370.0,
      "step": 8130
    },
    {
      "entropy": 5.648401927947998,
      "epoch": 0.8073640333465661,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000494344579779624,
      "loss": 5.4578,
      "mean_token_accuracy": 0.14570683762431144,
      "num_tokens": 17816910.0,
      "step": 8135
    },
    {
      "entropy": 5.63269624710083,
      "epoch": 0.8078602620087336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.4692,
      "mean_token_accuracy": 0.1559140294790268,
      "num_tokens": 17826792.0,
      "step": 8140
    },
    {
      "entropy": 5.598532342910767,
      "epoch": 0.8083564906709012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.3685,
      "mean_token_accuracy": 0.15593162924051285,
      "num_tokens": 17836986.0,
      "step": 8145
    },
    {
      "entropy": 5.612455987930298,
      "epoch": 0.8088527193330687,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.4497,
      "mean_token_accuracy": 0.15102608054876326,
      "num_tokens": 17849765.0,
      "step": 8150
    },
    {
      "entropy": 5.604673576354981,
      "epoch": 0.8093489479952362,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.2988,
      "mean_token_accuracy": 0.1542624846100807,
      "num_tokens": 17859666.0,
      "step": 8155
    },
    {
      "entropy": 5.627989912033081,
      "epoch": 0.8098451766574037,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494305041589742,
      "loss": 5.3585,
      "mean_token_accuracy": 0.15767708718776702,
      "num_tokens": 17869543.0,
      "step": 8160
    },
    {
      "entropy": 5.565008020401001,
      "epoch": 0.8103414053195712,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.3819,
      "mean_token_accuracy": 0.15892046838998794,
      "num_tokens": 17880347.0,
      "step": 8165
    },
    {
      "entropy": 5.587181615829468,
      "epoch": 0.8108376339817388,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.2574,
      "mean_token_accuracy": 0.16245560795068742,
      "num_tokens": 17889743.0,
      "step": 8170
    },
    {
      "entropy": 5.60626745223999,
      "epoch": 0.8113338626439063,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494281253414734,
      "loss": 5.2532,
      "mean_token_accuracy": 0.16442858278751374,
      "num_tokens": 17899853.0,
      "step": 8175
    },
    {
      "entropy": 5.584149694442749,
      "epoch": 0.8118300913060739,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1662215009331703,
      "num_tokens": 17909866.0,
      "step": 8180
    },
    {
      "entropy": 5.566713523864746,
      "epoch": 0.8123263199682413,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.4041,
      "mean_token_accuracy": 0.15417549759149551,
      "num_tokens": 17920086.0,
      "step": 8185
    },
    {
      "entropy": 5.643024539947509,
      "epoch": 0.8128225486304089,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.3454,
      "mean_token_accuracy": 0.15210518836975098,
      "num_tokens": 17930529.0,
      "step": 8190
    },
    {
      "entropy": 5.668810081481934,
      "epoch": 0.8133187772925764,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.4598,
      "mean_token_accuracy": 0.15129312872886658,
      "num_tokens": 17940818.0,
      "step": 8195
    },
    {
      "entropy": 5.61444058418274,
      "epoch": 0.813815005954744,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.3468,
      "mean_token_accuracy": 0.15352799147367477,
      "num_tokens": 17951241.0,
      "step": 8200
    },
    {
      "entropy": 5.671311759948731,
      "epoch": 0.8143112346169115,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494233530252071,
      "loss": 5.461,
      "mean_token_accuracy": 0.15286863148212432,
      "num_tokens": 17962353.0,
      "step": 8205
    },
    {
      "entropy": 5.62744107246399,
      "epoch": 0.814807463279079,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.3718,
      "mean_token_accuracy": 0.15298311412334442,
      "num_tokens": 17973672.0,
      "step": 8210
    },
    {
      "entropy": 5.577892780303955,
      "epoch": 0.8153036919412465,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.4122,
      "mean_token_accuracy": 0.15991712510585784,
      "num_tokens": 17985668.0,
      "step": 8215
    },
    {
      "entropy": 5.544806575775146,
      "epoch": 0.8157999206034141,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.3089,
      "mean_token_accuracy": 0.15600478947162627,
      "num_tokens": 17996326.0,
      "step": 8220
    },
    {
      "entropy": 5.634491348266602,
      "epoch": 0.8162961492655816,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.3967,
      "mean_token_accuracy": 0.1563344269990921,
      "num_tokens": 18007991.0,
      "step": 8225
    },
    {
      "entropy": 5.621115016937256,
      "epoch": 0.8167923779277491,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.3677,
      "mean_token_accuracy": 0.16104328036308288,
      "num_tokens": 18019356.0,
      "step": 8230
    },
    {
      "entropy": 5.5507493019104,
      "epoch": 0.8172886065899166,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.3372,
      "mean_token_accuracy": 0.15312866419553756,
      "num_tokens": 18030131.0,
      "step": 8235
    },
    {
      "entropy": 5.563556957244873,
      "epoch": 0.8177848352520841,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.2891,
      "mean_token_accuracy": 0.15343161374330522,
      "num_tokens": 18040848.0,
      "step": 8240
    },
    {
      "entropy": 5.6437132358551025,
      "epoch": 0.8182810639142517,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.5845,
      "mean_token_accuracy": 0.14782388806343078,
      "num_tokens": 18051888.0,
      "step": 8245
    },
    {
      "entropy": 5.663236951828003,
      "epoch": 0.8187772925764192,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.4058,
      "mean_token_accuracy": 0.15474457591772078,
      "num_tokens": 18064720.0,
      "step": 8250
    },
    {
      "entropy": 5.636355257034301,
      "epoch": 0.8192735212385868,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.4114,
      "mean_token_accuracy": 0.16236425191164017,
      "num_tokens": 18075208.0,
      "step": 8255
    },
    {
      "entropy": 5.568824243545532,
      "epoch": 0.8197697499007542,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494145529503543,
      "loss": 5.3415,
      "mean_token_accuracy": 0.15492145866155624,
      "num_tokens": 18086717.0,
      "step": 8260
    },
    {
      "entropy": 5.52168116569519,
      "epoch": 0.8202659785629218,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.2814,
      "mean_token_accuracy": 0.15663959905505181,
      "num_tokens": 18097593.0,
      "step": 8265
    },
    {
      "entropy": 5.588707017898559,
      "epoch": 0.8207622072250893,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.3356,
      "mean_token_accuracy": 0.15573811903595924,
      "num_tokens": 18108470.0,
      "step": 8270
    },
    {
      "entropy": 5.605103349685669,
      "epoch": 0.8212584358872569,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.3061,
      "mean_token_accuracy": 0.1623164527118206,
      "num_tokens": 18119854.0,
      "step": 8275
    },
    {
      "entropy": 5.579291915893554,
      "epoch": 0.8217546645494244,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.3199,
      "mean_token_accuracy": 0.15792210698127745,
      "num_tokens": 18130684.0,
      "step": 8280
    },
    {
      "entropy": 5.616350746154785,
      "epoch": 0.822250893211592,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.3674,
      "mean_token_accuracy": 0.15801014602184296,
      "num_tokens": 18140555.0,
      "step": 8285
    },
    {
      "entropy": 5.546979475021362,
      "epoch": 0.8227471218737594,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.2843,
      "mean_token_accuracy": 0.16096103340387344,
      "num_tokens": 18150478.0,
      "step": 8290
    },
    {
      "entropy": 5.650714969635009,
      "epoch": 0.823243350535927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.4737,
      "mean_token_accuracy": 0.14824917763471604,
      "num_tokens": 18160889.0,
      "step": 8295
    },
    {
      "entropy": 5.614264249801636,
      "epoch": 0.8237395791980945,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.3971,
      "mean_token_accuracy": 0.15588648915290831,
      "num_tokens": 18171086.0,
      "step": 8300
    },
    {
      "entropy": 5.682441234588623,
      "epoch": 0.824235807860262,
      "grad_norm": 0.875,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.3421,
      "mean_token_accuracy": 0.15635645389556885,
      "num_tokens": 18182711.0,
      "step": 8305
    },
    {
      "entropy": 5.612247943878174,
      "epoch": 0.8247320365224295,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.4132,
      "mean_token_accuracy": 0.154473115503788,
      "num_tokens": 18193655.0,
      "step": 8310
    },
    {
      "entropy": 5.629425096511841,
      "epoch": 0.825228265184597,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.3684,
      "mean_token_accuracy": 0.15860945135354995,
      "num_tokens": 18203921.0,
      "step": 8315
    },
    {
      "entropy": 5.599585676193238,
      "epoch": 0.8257244938467646,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.4648,
      "mean_token_accuracy": 0.155076764523983,
      "num_tokens": 18215581.0,
      "step": 8320
    },
    {
      "entropy": 5.674492502212525,
      "epoch": 0.8262207225089321,
      "grad_norm": 0.8125,
      "learning_rate": 0.000494040680883325,
      "loss": 5.3854,
      "mean_token_accuracy": 0.15098653882741928,
      "num_tokens": 18226273.0,
      "step": 8325
    },
    {
      "entropy": 5.604614114761352,
      "epoch": 0.8267169511710997,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.3385,
      "mean_token_accuracy": 0.1536204993724823,
      "num_tokens": 18236653.0,
      "step": 8330
    },
    {
      "entropy": 5.626929759979248,
      "epoch": 0.8272131798332671,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.4123,
      "mean_token_accuracy": 0.1565910443663597,
      "num_tokens": 18246421.0,
      "step": 8335
    },
    {
      "entropy": 5.62087116241455,
      "epoch": 0.8277094084954347,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.396,
      "mean_token_accuracy": 0.15819667428731918,
      "num_tokens": 18257698.0,
      "step": 8340
    },
    {
      "entropy": 5.690956068038941,
      "epoch": 0.8282056371576022,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.4658,
      "mean_token_accuracy": 0.14990609437227248,
      "num_tokens": 18267499.0,
      "step": 8345
    },
    {
      "entropy": 5.585352754592895,
      "epoch": 0.8287018658197698,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1565021187067032,
      "num_tokens": 18279070.0,
      "step": 8350
    },
    {
      "entropy": 5.632038831710815,
      "epoch": 0.8291980944819373,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.3456,
      "mean_token_accuracy": 0.15390770733356476,
      "num_tokens": 18289843.0,
      "step": 8355
    },
    {
      "entropy": 5.640030384063721,
      "epoch": 0.8296943231441049,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.3779,
      "mean_token_accuracy": 0.15570164024829863,
      "num_tokens": 18302089.0,
      "step": 8360
    },
    {
      "entropy": 5.627889394760132,
      "epoch": 0.8301905518062723,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.4384,
      "mean_token_accuracy": 0.15519084706902503,
      "num_tokens": 18313173.0,
      "step": 8365
    },
    {
      "entropy": 5.570294713973999,
      "epoch": 0.8306867804684399,
      "grad_norm": 0.875,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.3354,
      "mean_token_accuracy": 0.162067049741745,
      "num_tokens": 18324108.0,
      "step": 8370
    },
    {
      "entropy": 5.624648809432983,
      "epoch": 0.8311830091306074,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.2986,
      "mean_token_accuracy": 0.1590510353446007,
      "num_tokens": 18334146.0,
      "step": 8375
    },
    {
      "entropy": 5.562149524688721,
      "epoch": 0.8316792377927749,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.3103,
      "mean_token_accuracy": 0.15622681379318237,
      "num_tokens": 18345543.0,
      "step": 8380
    },
    {
      "entropy": 5.613871049880982,
      "epoch": 0.8321754664549424,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.3931,
      "mean_token_accuracy": 0.15306339040398598,
      "num_tokens": 18357923.0,
      "step": 8385
    },
    {
      "entropy": 5.612277221679688,
      "epoch": 0.8326716951171099,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.3974,
      "mean_token_accuracy": 0.15441624969244003,
      "num_tokens": 18369211.0,
      "step": 8390
    },
    {
      "entropy": 5.653699016571045,
      "epoch": 0.8331679237792775,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.2935,
      "mean_token_accuracy": 0.1570999026298523,
      "num_tokens": 18380307.0,
      "step": 8395
    },
    {
      "entropy": 5.563513565063476,
      "epoch": 0.833664152441445,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000493918561069992,
      "loss": 5.3011,
      "mean_token_accuracy": 0.16070234179496765,
      "num_tokens": 18391099.0,
      "step": 8400
    },
    {
      "entropy": 5.596113777160644,
      "epoch": 0.8341603811036126,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.3486,
      "mean_token_accuracy": 0.15528119951486588,
      "num_tokens": 18402882.0,
      "step": 8405
    },
    {
      "entropy": 5.690075922012329,
      "epoch": 0.83465660976578,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.3971,
      "mean_token_accuracy": 0.1567097343504429,
      "num_tokens": 18414815.0,
      "step": 8410
    },
    {
      "entropy": 5.606011056900025,
      "epoch": 0.8351528384279476,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.3722,
      "mean_token_accuracy": 0.15944744795560836,
      "num_tokens": 18426394.0,
      "step": 8415
    },
    {
      "entropy": 5.609368562698364,
      "epoch": 0.8356490670901151,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.3404,
      "mean_token_accuracy": 0.15961466431617738,
      "num_tokens": 18436147.0,
      "step": 8420
    },
    {
      "entropy": 5.551829862594604,
      "epoch": 0.8361452957522827,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1614237979054451,
      "num_tokens": 18447536.0,
      "step": 8425
    },
    {
      "entropy": 5.608613443374634,
      "epoch": 0.8366415244144502,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.3522,
      "mean_token_accuracy": 0.15601782351732255,
      "num_tokens": 18458956.0,
      "step": 8430
    },
    {
      "entropy": 5.575527048110962,
      "epoch": 0.8371377530766178,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.3563,
      "mean_token_accuracy": 0.15309559404850007,
      "num_tokens": 18470490.0,
      "step": 8435
    },
    {
      "entropy": 5.572996616363525,
      "epoch": 0.8376339817387852,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.3748,
      "mean_token_accuracy": 0.16257106214761735,
      "num_tokens": 18480512.0,
      "step": 8440
    },
    {
      "entropy": 5.600152826309204,
      "epoch": 0.8381302104009528,
      "grad_norm": 0.953125,
      "learning_rate": 0.000493844702837845,
      "loss": 5.3459,
      "mean_token_accuracy": 0.15522438585758208,
      "num_tokens": 18492657.0,
      "step": 8445
    },
    {
      "entropy": 5.63456621170044,
      "epoch": 0.8386264390631203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.2987,
      "mean_token_accuracy": 0.15847290754318238,
      "num_tokens": 18504108.0,
      "step": 8450
    },
    {
      "entropy": 5.643517541885376,
      "epoch": 0.8391226677252878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.4517,
      "mean_token_accuracy": 0.15475277751684188,
      "num_tokens": 18514760.0,
      "step": 8455
    },
    {
      "entropy": 5.631017065048217,
      "epoch": 0.8396188963874553,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.4265,
      "mean_token_accuracy": 0.16167930513620377,
      "num_tokens": 18526039.0,
      "step": 8460
    },
    {
      "entropy": 5.633409881591797,
      "epoch": 0.8401151250496228,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049381173584926,
      "loss": 5.4742,
      "mean_token_accuracy": 0.15149416625499726,
      "num_tokens": 18537818.0,
      "step": 8465
    },
    {
      "entropy": 5.652733993530274,
      "epoch": 0.8406113537117904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.4247,
      "mean_token_accuracy": 0.15181732326745986,
      "num_tokens": 18549028.0,
      "step": 8470
    },
    {
      "entropy": 5.593077802658081,
      "epoch": 0.8411075823739579,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.2955,
      "mean_token_accuracy": 0.16089700758457184,
      "num_tokens": 18559231.0,
      "step": 8475
    },
    {
      "entropy": 5.650155448913575,
      "epoch": 0.8416038110361255,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493786953633481,
      "loss": 5.3941,
      "mean_token_accuracy": 0.15679316222667694,
      "num_tokens": 18570232.0,
      "step": 8480
    },
    {
      "entropy": 5.648695278167724,
      "epoch": 0.842100039698293,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.3869,
      "mean_token_accuracy": 0.15618253797292708,
      "num_tokens": 18581856.0,
      "step": 8485
    },
    {
      "entropy": 5.638518524169922,
      "epoch": 0.8425962683604605,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.3917,
      "mean_token_accuracy": 0.15878208577632905,
      "num_tokens": 18592412.0,
      "step": 8490
    },
    {
      "entropy": 5.638802528381348,
      "epoch": 0.843092497022628,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1633645474910736,
      "num_tokens": 18602182.0,
      "step": 8495
    },
    {
      "entropy": 5.637797260284424,
      "epoch": 0.8435887256847956,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.4185,
      "mean_token_accuracy": 0.1555818259716034,
      "num_tokens": 18612701.0,
      "step": 8500
    },
    {
      "entropy": 5.575397396087647,
      "epoch": 0.8440849543469631,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000493745541434354,
      "loss": 5.3584,
      "mean_token_accuracy": 0.1490127384662628,
      "num_tokens": 18624231.0,
      "step": 8505
    },
    {
      "entropy": 5.658658599853515,
      "epoch": 0.8445811830091307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.3357,
      "mean_token_accuracy": 0.16234812438488005,
      "num_tokens": 18634424.0,
      "step": 8510
    },
    {
      "entropy": 5.575712966918945,
      "epoch": 0.8450774116712981,
      "grad_norm": 0.875,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.3587,
      "mean_token_accuracy": 0.1548302337527275,
      "num_tokens": 18646357.0,
      "step": 8515
    },
    {
      "entropy": 5.54103193283081,
      "epoch": 0.8455736403334657,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.27,
      "mean_token_accuracy": 0.15578805655241013,
      "num_tokens": 18657836.0,
      "step": 8520
    },
    {
      "entropy": 5.579985761642456,
      "epoch": 0.8460698689956332,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.3567,
      "mean_token_accuracy": 0.15607503801584244,
      "num_tokens": 18668426.0,
      "step": 8525
    },
    {
      "entropy": 5.612581777572632,
      "epoch": 0.8465660976578007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493703993624782,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15637078434228896,
      "num_tokens": 18677975.0,
      "step": 8530
    },
    {
      "entropy": 5.657680606842041,
      "epoch": 0.8470623263199683,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.4184,
      "mean_token_accuracy": 0.1554457426071167,
      "num_tokens": 18689944.0,
      "step": 8535
    },
    {
      "entropy": 5.5581377983093265,
      "epoch": 0.8475585549821357,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.3292,
      "mean_token_accuracy": 0.15647507160902024,
      "num_tokens": 18700524.0,
      "step": 8540
    },
    {
      "entropy": 5.589782762527466,
      "epoch": 0.8480547836443033,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.3387,
      "mean_token_accuracy": 0.15716854035854338,
      "num_tokens": 18711184.0,
      "step": 8545
    },
    {
      "entropy": 5.595709180831909,
      "epoch": 0.8485510123064708,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1532974734902382,
      "num_tokens": 18721393.0,
      "step": 8550
    },
    {
      "entropy": 5.631995677947998,
      "epoch": 0.8490472409686384,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.469,
      "mean_token_accuracy": 0.15430249124765397,
      "num_tokens": 18732690.0,
      "step": 8555
    },
    {
      "entropy": 5.594232273101807,
      "epoch": 0.8495434696308058,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.2703,
      "mean_token_accuracy": 0.16304569840431213,
      "num_tokens": 18744637.0,
      "step": 8560
    },
    {
      "entropy": 5.548993492126465,
      "epoch": 0.8500396982929734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.2382,
      "mean_token_accuracy": 0.16292684227228166,
      "num_tokens": 18754751.0,
      "step": 8565
    },
    {
      "entropy": 5.679952096939087,
      "epoch": 0.8505359269551409,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.4009,
      "mean_token_accuracy": 0.15540016666054726,
      "num_tokens": 18767412.0,
      "step": 8570
    },
    {
      "entropy": 5.564809846878052,
      "epoch": 0.8510321556173085,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000493628865911502,
      "loss": 5.3362,
      "mean_token_accuracy": 0.15500217378139497,
      "num_tokens": 18779287.0,
      "step": 8575
    },
    {
      "entropy": 5.594174003601074,
      "epoch": 0.851528384279476,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.3259,
      "mean_token_accuracy": 0.15980712622404097,
      "num_tokens": 18790461.0,
      "step": 8580
    },
    {
      "entropy": 5.6191596508026125,
      "epoch": 0.8520246129416436,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.3318,
      "mean_token_accuracy": 0.15613968148827553,
      "num_tokens": 18801310.0,
      "step": 8585
    },
    {
      "entropy": 5.5882275104522705,
      "epoch": 0.852520841603811,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.3678,
      "mean_token_accuracy": 0.16296720057725905,
      "num_tokens": 18811002.0,
      "step": 8590
    },
    {
      "entropy": 5.625974321365357,
      "epoch": 0.8530170702659786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493595334848088,
      "loss": 5.3617,
      "mean_token_accuracy": 0.16331041753292083,
      "num_tokens": 18821333.0,
      "step": 8595
    },
    {
      "entropy": 5.637843942642212,
      "epoch": 0.8535132989281461,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.4097,
      "mean_token_accuracy": 0.15526322796940803,
      "num_tokens": 18832173.0,
      "step": 8600
    },
    {
      "entropy": 5.619658994674682,
      "epoch": 0.8540095275903136,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.3584,
      "mean_token_accuracy": 0.159215746819973,
      "num_tokens": 18843298.0,
      "step": 8605
    },
    {
      "entropy": 5.656082487106323,
      "epoch": 0.8545057562524812,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.4037,
      "mean_token_accuracy": 0.14651786610484124,
      "num_tokens": 18854720.0,
      "step": 8610
    },
    {
      "entropy": 5.57837872505188,
      "epoch": 0.8550019849146486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.3597,
      "mean_token_accuracy": 0.1592981219291687,
      "num_tokens": 18866060.0,
      "step": 8615
    },
    {
      "entropy": 5.561797142028809,
      "epoch": 0.8554982135768162,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.364,
      "mean_token_accuracy": 0.15969037413597106,
      "num_tokens": 18877646.0,
      "step": 8620
    },
    {
      "entropy": 5.574878740310669,
      "epoch": 0.8559944422389837,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.3275,
      "mean_token_accuracy": 0.15330654829740525,
      "num_tokens": 18888426.0,
      "step": 8625
    },
    {
      "entropy": 5.621098184585572,
      "epoch": 0.8564906709011513,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.3515,
      "mean_token_accuracy": 0.15238586366176604,
      "num_tokens": 18897985.0,
      "step": 8630
    },
    {
      "entropy": 5.637457942962646,
      "epoch": 0.8569868995633187,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.403,
      "mean_token_accuracy": 0.15697371661663057,
      "num_tokens": 18908804.0,
      "step": 8635
    },
    {
      "entropy": 5.60301513671875,
      "epoch": 0.8574831282254863,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.2886,
      "mean_token_accuracy": 0.16138957291841508,
      "num_tokens": 18918347.0,
      "step": 8640
    },
    {
      "entropy": 5.555605125427246,
      "epoch": 0.8579793568876538,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000493511127770223,
      "loss": 5.2396,
      "mean_token_accuracy": 0.16767623722553254,
      "num_tokens": 18927996.0,
      "step": 8645
    },
    {
      "entropy": 5.52012448310852,
      "epoch": 0.8584755855498214,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.278,
      "mean_token_accuracy": 0.16145460307598114,
      "num_tokens": 18939603.0,
      "step": 8650
    },
    {
      "entropy": 5.639473247528076,
      "epoch": 0.8589718142119889,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.4137,
      "mean_token_accuracy": 0.1570653960108757,
      "num_tokens": 18950721.0,
      "step": 8655
    },
    {
      "entropy": 5.538495922088623,
      "epoch": 0.8594680428741565,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493485759974549,
      "loss": 5.277,
      "mean_token_accuracy": 0.16104861795902253,
      "num_tokens": 18961666.0,
      "step": 8660
    },
    {
      "entropy": 5.5623595237731935,
      "epoch": 0.8599642715363239,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1633254423737526,
      "num_tokens": 18972223.0,
      "step": 8665
    },
    {
      "entropy": 5.58285961151123,
      "epoch": 0.8604605001984915,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.3291,
      "mean_token_accuracy": 0.16010155230760575,
      "num_tokens": 18982983.0,
      "step": 8670
    },
    {
      "entropy": 5.594792604446411,
      "epoch": 0.860956728860659,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.3538,
      "mean_token_accuracy": 0.15369837135076522,
      "num_tokens": 18993521.0,
      "step": 8675
    },
    {
      "entropy": 5.645217847824097,
      "epoch": 0.8614529575228265,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.4396,
      "mean_token_accuracy": 0.1565738722681999,
      "num_tokens": 19004901.0,
      "step": 8680
    },
    {
      "entropy": 5.5139528751373295,
      "epoch": 0.861949186184994,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.2608,
      "mean_token_accuracy": 0.16335452049970628,
      "num_tokens": 19016123.0,
      "step": 8685
    },
    {
      "entropy": 5.655598974227905,
      "epoch": 0.8624454148471615,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.4042,
      "mean_token_accuracy": 0.15062565207481385,
      "num_tokens": 19026760.0,
      "step": 8690
    },
    {
      "entropy": 5.588661623001099,
      "epoch": 0.8629416435093291,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493426378831957,
      "loss": 5.2998,
      "mean_token_accuracy": 0.16091118305921553,
      "num_tokens": 19036665.0,
      "step": 8695
    },
    {
      "entropy": 5.576855039596557,
      "epoch": 0.8634378721714966,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1584911748766899,
      "num_tokens": 19047779.0,
      "step": 8700
    },
    {
      "entropy": 5.57593789100647,
      "epoch": 0.8639341008336642,
      "grad_norm": 0.90625,
      "learning_rate": 0.000493409364039526,
      "loss": 5.3917,
      "mean_token_accuracy": 0.15438186824321748,
      "num_tokens": 19058292.0,
      "step": 8705
    },
    {
      "entropy": 5.630038738250732,
      "epoch": 0.8644303294958317,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.3597,
      "mean_token_accuracy": 0.16072400957345961,
      "num_tokens": 19069938.0,
      "step": 8710
    },
    {
      "entropy": 5.663787031173706,
      "epoch": 0.8649265581579992,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.3861,
      "mean_token_accuracy": 0.151009963452816,
      "num_tokens": 19080492.0,
      "step": 8715
    },
    {
      "entropy": 5.60584602355957,
      "epoch": 0.8654227868201667,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.453,
      "mean_token_accuracy": 0.15526572465896607,
      "num_tokens": 19091723.0,
      "step": 8720
    },
    {
      "entropy": 5.620947933197021,
      "epoch": 0.8659190154823343,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493375269463375,
      "loss": 5.353,
      "mean_token_accuracy": 0.15451942384243011,
      "num_tokens": 19102652.0,
      "step": 8725
    },
    {
      "entropy": 5.582439756393432,
      "epoch": 0.8664152441445018,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.2549,
      "mean_token_accuracy": 0.16099116951227188,
      "num_tokens": 19114139.0,
      "step": 8730
    },
    {
      "entropy": 5.545179748535157,
      "epoch": 0.8669114728066694,
      "grad_norm": 0.84375,
      "learning_rate": 0.000493358189683037,
      "loss": 5.3944,
      "mean_token_accuracy": 0.15553621500730513,
      "num_tokens": 19126293.0,
      "step": 8735
    },
    {
      "entropy": 5.539413690567017,
      "epoch": 0.8674077014688368,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.2578,
      "mean_token_accuracy": 0.16513072103261947,
      "num_tokens": 19136947.0,
      "step": 8740
    },
    {
      "entropy": 5.590924406051636,
      "epoch": 0.8679039301310044,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.3116,
      "mean_token_accuracy": 0.1593662090599537,
      "num_tokens": 19146987.0,
      "step": 8745
    },
    {
      "entropy": 5.576338243484497,
      "epoch": 0.8684001587931719,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.3825,
      "mean_token_accuracy": 0.15438729673624038,
      "num_tokens": 19157972.0,
      "step": 8750
    },
    {
      "entropy": 5.591269445419312,
      "epoch": 0.8688963874553394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.3963,
      "mean_token_accuracy": 0.1547196127474308,
      "num_tokens": 19170313.0,
      "step": 8755
    },
    {
      "entropy": 5.648441314697266,
      "epoch": 0.869392616117507,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.3929,
      "mean_token_accuracy": 0.1565806858241558,
      "num_tokens": 19183134.0,
      "step": 8760
    },
    {
      "entropy": 5.652920722961426,
      "epoch": 0.8698888447796744,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.3202,
      "mean_token_accuracy": 0.1614924803376198,
      "num_tokens": 19194900.0,
      "step": 8765
    },
    {
      "entropy": 5.58241901397705,
      "epoch": 0.870385073441842,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000493298239896396,
      "loss": 5.3882,
      "mean_token_accuracy": 0.1591202899813652,
      "num_tokens": 19204980.0,
      "step": 8770
    },
    {
      "entropy": 5.574018907546997,
      "epoch": 0.8708813021040095,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1572340741753578,
      "num_tokens": 19217581.0,
      "step": 8775
    },
    {
      "entropy": 5.578329229354859,
      "epoch": 0.8713775307661771,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.2644,
      "mean_token_accuracy": 0.16128665804862977,
      "num_tokens": 19227694.0,
      "step": 8780
    },
    {
      "entropy": 5.587977695465088,
      "epoch": 0.8718737594283446,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.4033,
      "mean_token_accuracy": 0.15560183078050613,
      "num_tokens": 19238579.0,
      "step": 8785
    },
    {
      "entropy": 5.576746988296509,
      "epoch": 0.8723699880905121,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.3388,
      "mean_token_accuracy": 0.16043743789196013,
      "num_tokens": 19249658.0,
      "step": 8790
    },
    {
      "entropy": 5.551037979125977,
      "epoch": 0.8728662167526796,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000493255256216201,
      "loss": 5.3199,
      "mean_token_accuracy": 0.16095189452171327,
      "num_tokens": 19260273.0,
      "step": 8795
    },
    {
      "entropy": 5.6395526885986325,
      "epoch": 0.8733624454148472,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.3776,
      "mean_token_accuracy": 0.15290962308645248,
      "num_tokens": 19270344.0,
      "step": 8800
    },
    {
      "entropy": 5.5717244148254395,
      "epoch": 0.8738586740770147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.261,
      "mean_token_accuracy": 0.16190992146730424,
      "num_tokens": 19280934.0,
      "step": 8805
    },
    {
      "entropy": 5.602689743041992,
      "epoch": 0.8743549027391823,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1527844101190567,
      "num_tokens": 19291028.0,
      "step": 8810
    },
    {
      "entropy": 5.5577061653137205,
      "epoch": 0.8748511314013497,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.2867,
      "mean_token_accuracy": 0.1621830329298973,
      "num_tokens": 19302436.0,
      "step": 8815
    },
    {
      "entropy": 5.633925151824951,
      "epoch": 0.8753473600635173,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.3425,
      "mean_token_accuracy": 0.15692461431026458,
      "num_tokens": 19312650.0,
      "step": 8820
    },
    {
      "entropy": 5.507609462738037,
      "epoch": 0.8758435887256848,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.2717,
      "mean_token_accuracy": 0.16204639077186583,
      "num_tokens": 19323727.0,
      "step": 8825
    },
    {
      "entropy": 5.553243160247803,
      "epoch": 0.8763398173878523,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.3034,
      "mean_token_accuracy": 0.15697551220655442,
      "num_tokens": 19333286.0,
      "step": 8830
    },
    {
      "entropy": 5.615463161468506,
      "epoch": 0.8768360460500199,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.3334,
      "mean_token_accuracy": 0.15909288227558135,
      "num_tokens": 19343766.0,
      "step": 8835
    },
    {
      "entropy": 5.623085069656372,
      "epoch": 0.8773322747121873,
      "grad_norm": 0.890625,
      "learning_rate": 0.000493177544637444,
      "loss": 5.3052,
      "mean_token_accuracy": 0.15700812488794327,
      "num_tokens": 19354944.0,
      "step": 8840
    },
    {
      "entropy": 5.617136383056641,
      "epoch": 0.8778285033743549,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.3375,
      "mean_token_accuracy": 0.1600698336958885,
      "num_tokens": 19366832.0,
      "step": 8845
    },
    {
      "entropy": 5.637832498550415,
      "epoch": 0.8783247320365224,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.4179,
      "mean_token_accuracy": 0.15691168308258058,
      "num_tokens": 19379067.0,
      "step": 8850
    },
    {
      "entropy": 5.5958826541900635,
      "epoch": 0.87882096069869,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.333,
      "mean_token_accuracy": 0.155780066549778,
      "num_tokens": 19390419.0,
      "step": 8855
    },
    {
      "entropy": 5.617925071716309,
      "epoch": 0.8793171893608575,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.3751,
      "mean_token_accuracy": 0.15275435149669647,
      "num_tokens": 19401482.0,
      "step": 8860
    },
    {
      "entropy": 5.587957668304443,
      "epoch": 0.879813418023025,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.298,
      "mean_token_accuracy": 0.15809966623783112,
      "num_tokens": 19411620.0,
      "step": 8865
    },
    {
      "entropy": 5.553340101242066,
      "epoch": 0.8803096466851925,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.3695,
      "mean_token_accuracy": 0.15897068828344346,
      "num_tokens": 19422247.0,
      "step": 8870
    },
    {
      "entropy": 5.5806817531585695,
      "epoch": 0.8808058753473601,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.3553,
      "mean_token_accuracy": 0.15655985176563264,
      "num_tokens": 19433687.0,
      "step": 8875
    },
    {
      "entropy": 5.59553165435791,
      "epoch": 0.8813021040095276,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.2679,
      "mean_token_accuracy": 0.16422978639602662,
      "num_tokens": 19443884.0,
      "step": 8880
    },
    {
      "entropy": 5.6546924114227295,
      "epoch": 0.8817983326716952,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.4162,
      "mean_token_accuracy": 0.150820492208004,
      "num_tokens": 19456425.0,
      "step": 8885
    },
    {
      "entropy": 5.672200536727905,
      "epoch": 0.8822945613338626,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.3971,
      "mean_token_accuracy": 0.1519812598824501,
      "num_tokens": 19468838.0,
      "step": 8890
    },
    {
      "entropy": 5.578950500488281,
      "epoch": 0.8827907899960302,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.2988,
      "mean_token_accuracy": 0.1615019902586937,
      "num_tokens": 19479603.0,
      "step": 8895
    },
    {
      "entropy": 5.528980255126953,
      "epoch": 0.8832870186581977,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000493073247524365,
      "loss": 5.2999,
      "mean_token_accuracy": 0.1615740418434143,
      "num_tokens": 19491793.0,
      "step": 8900
    },
    {
      "entropy": 5.582971858978271,
      "epoch": 0.8837832473203652,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.3266,
      "mean_token_accuracy": 0.1594070553779602,
      "num_tokens": 19502869.0,
      "step": 8905
    },
    {
      "entropy": 5.569485759735107,
      "epoch": 0.8842794759825328,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000493055788953656,
      "loss": 5.2931,
      "mean_token_accuracy": 0.15838595032691954,
      "num_tokens": 19513766.0,
      "step": 8910
    },
    {
      "entropy": 5.556046676635742,
      "epoch": 0.8847757046447002,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.2945,
      "mean_token_accuracy": 0.16853482872247696,
      "num_tokens": 19524561.0,
      "step": 8915
    },
    {
      "entropy": 5.600092554092408,
      "epoch": 0.8852719333068678,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.2459,
      "mean_token_accuracy": 0.1625062644481659,
      "num_tokens": 19535253.0,
      "step": 8920
    },
    {
      "entropy": 5.535506629943848,
      "epoch": 0.8857681619690353,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.306,
      "mean_token_accuracy": 0.16020966172218323,
      "num_tokens": 19546650.0,
      "step": 8925
    },
    {
      "entropy": 5.595127725601197,
      "epoch": 0.8862643906312029,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.338,
      "mean_token_accuracy": 0.1646319642663002,
      "num_tokens": 19558012.0,
      "step": 8930
    },
    {
      "entropy": 5.573107576370239,
      "epoch": 0.8867606192933704,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.3363,
      "mean_token_accuracy": 0.15970392301678657,
      "num_tokens": 19569491.0,
      "step": 8935
    },
    {
      "entropy": 5.514723777770996,
      "epoch": 0.8872568479555379,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.3253,
      "mean_token_accuracy": 0.16029038578271865,
      "num_tokens": 19581116.0,
      "step": 8940
    },
    {
      "entropy": 5.6267228603363035,
      "epoch": 0.8877530766177054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492994513637178,
      "loss": 5.3945,
      "mean_token_accuracy": 0.15274353474378585,
      "num_tokens": 19591351.0,
      "step": 8945
    },
    {
      "entropy": 5.621697854995728,
      "epoch": 0.888249305279873,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.3508,
      "mean_token_accuracy": 0.15978964567184448,
      "num_tokens": 19602071.0,
      "step": 8950
    },
    {
      "entropy": 5.600676822662353,
      "epoch": 0.8887455339420405,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.3557,
      "mean_token_accuracy": 0.1630178526043892,
      "num_tokens": 19612799.0,
      "step": 8955
    },
    {
      "entropy": 5.592666816711426,
      "epoch": 0.8892417626042081,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000492968171696001,
      "loss": 5.3243,
      "mean_token_accuracy": 0.1548623263835907,
      "num_tokens": 19623901.0,
      "step": 8960
    },
    {
      "entropy": 5.6259071826934814,
      "epoch": 0.8897379912663755,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.2593,
      "mean_token_accuracy": 0.15817082226276397,
      "num_tokens": 19633963.0,
      "step": 8965
    },
    {
      "entropy": 5.483339309692383,
      "epoch": 0.8902342199285431,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1647270515561104,
      "num_tokens": 19644009.0,
      "step": 8970
    },
    {
      "entropy": 5.512039422988892,
      "epoch": 0.8907304485907106,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.215,
      "mean_token_accuracy": 0.16818392127752305,
      "num_tokens": 19654051.0,
      "step": 8975
    },
    {
      "entropy": 5.585809230804443,
      "epoch": 0.8912266772528781,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.3731,
      "mean_token_accuracy": 0.16150182783603667,
      "num_tokens": 19664118.0,
      "step": 8980
    },
    {
      "entropy": 5.578886842727661,
      "epoch": 0.8917229059150457,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.347,
      "mean_token_accuracy": 0.15661486685276033,
      "num_tokens": 19674598.0,
      "step": 8985
    },
    {
      "entropy": 5.511644315719605,
      "epoch": 0.8922191345772131,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492915341880733,
      "loss": 5.3418,
      "mean_token_accuracy": 0.15590587109327317,
      "num_tokens": 19686311.0,
      "step": 8990
    },
    {
      "entropy": 5.60111927986145,
      "epoch": 0.8927153632393807,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.3646,
      "mean_token_accuracy": 0.15273165926337243,
      "num_tokens": 19697518.0,
      "step": 8995
    },
    {
      "entropy": 5.6285388469696045,
      "epoch": 0.8932115919015482,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.3665,
      "mean_token_accuracy": 0.16206228882074356,
      "num_tokens": 19709768.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.453627224063669,
      "eval_loss": 5.396178722381592,
      "eval_mean_token_accuracy": 0.16025147596885286,
      "eval_num_tokens": 19709768.0,
      "eval_runtime": 20.5245,
      "eval_samples_per_second": 1640.433,
      "eval_steps_per_second": 205.072,
      "step": 9000
    },
    {
      "entropy": 5.563939046859741,
      "epoch": 0.8937078205637158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.261,
      "mean_token_accuracy": 0.15785569846630096,
      "num_tokens": 19721154.0,
      "step": 9005
    },
    {
      "entropy": 5.549664735794067,
      "epoch": 0.8942040492258833,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.2432,
      "mean_token_accuracy": 0.15936800241470336,
      "num_tokens": 19731965.0,
      "step": 9010
    },
    {
      "entropy": 5.574559926986694,
      "epoch": 0.8947002778880508,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.3779,
      "mean_token_accuracy": 0.14554269537329673,
      "num_tokens": 19742385.0,
      "step": 9015
    },
    {
      "entropy": 5.641107368469238,
      "epoch": 0.8951965065502183,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.3688,
      "mean_token_accuracy": 0.1551334723830223,
      "num_tokens": 19753027.0,
      "step": 9020
    },
    {
      "entropy": 5.627093029022217,
      "epoch": 0.8956927352123859,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.3634,
      "mean_token_accuracy": 0.15058883726596833,
      "num_tokens": 19763625.0,
      "step": 9025
    },
    {
      "entropy": 5.613175868988037,
      "epoch": 0.8961889638745534,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.3249,
      "mean_token_accuracy": 0.16086267828941345,
      "num_tokens": 19775425.0,
      "step": 9030
    },
    {
      "entropy": 5.579009342193603,
      "epoch": 0.896685192536721,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.3613,
      "mean_token_accuracy": 0.15699477791786193,
      "num_tokens": 19786343.0,
      "step": 9035
    },
    {
      "entropy": 5.543609619140625,
      "epoch": 0.8971814211988884,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.2566,
      "mean_token_accuracy": 0.16013272255659103,
      "num_tokens": 19796903.0,
      "step": 9040
    },
    {
      "entropy": 5.626705741882324,
      "epoch": 0.897677649861056,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.3737,
      "mean_token_accuracy": 0.15626833736896514,
      "num_tokens": 19808026.0,
      "step": 9045
    },
    {
      "entropy": 5.593157339096069,
      "epoch": 0.8981738785232235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.3136,
      "mean_token_accuracy": 0.15683972239494323,
      "num_tokens": 19819343.0,
      "step": 9050
    },
    {
      "entropy": 5.669935321807861,
      "epoch": 0.898670107185391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492800209971077,
      "loss": 5.4058,
      "mean_token_accuracy": 0.15618302449584007,
      "num_tokens": 19831264.0,
      "step": 9055
    },
    {
      "entropy": 5.665454912185669,
      "epoch": 0.8991663358475586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.4255,
      "mean_token_accuracy": 0.15055545270442963,
      "num_tokens": 19842101.0,
      "step": 9060
    },
    {
      "entropy": 5.56555027961731,
      "epoch": 0.899662564509726,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.33,
      "mean_token_accuracy": 0.15823139548301696,
      "num_tokens": 19853789.0,
      "step": 9065
    },
    {
      "entropy": 5.575005292892456,
      "epoch": 0.9001587931718936,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1617904022336006,
      "num_tokens": 19865855.0,
      "step": 9070
    },
    {
      "entropy": 5.640824270248413,
      "epoch": 0.9006550218340611,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.4448,
      "mean_token_accuracy": 0.15410556495189667,
      "num_tokens": 19877443.0,
      "step": 9075
    },
    {
      "entropy": 5.5719592571258545,
      "epoch": 0.9011512504962287,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.2923,
      "mean_token_accuracy": 0.16394501328468322,
      "num_tokens": 19889503.0,
      "step": 9080
    },
    {
      "entropy": 5.635532093048096,
      "epoch": 0.9016474791583962,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.3954,
      "mean_token_accuracy": 0.15535518079996108,
      "num_tokens": 19899969.0,
      "step": 9085
    },
    {
      "entropy": 5.597247695922851,
      "epoch": 0.9021437078205637,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.3596,
      "mean_token_accuracy": 0.1539223909378052,
      "num_tokens": 19911326.0,
      "step": 9090
    },
    {
      "entropy": 5.6681952476501465,
      "epoch": 0.9026399364827312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492728905852412,
      "loss": 5.3744,
      "mean_token_accuracy": 0.15609387010335923,
      "num_tokens": 19922297.0,
      "step": 9095
    },
    {
      "entropy": 5.664239835739136,
      "epoch": 0.9031361651448988,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.3458,
      "mean_token_accuracy": 0.1604857474565506,
      "num_tokens": 19934727.0,
      "step": 9100
    },
    {
      "entropy": 5.5298487663269045,
      "epoch": 0.9036323938070663,
      "grad_norm": 0.84375,
      "learning_rate": 0.000492711025826243,
      "loss": 5.2293,
      "mean_token_accuracy": 0.16582711338996886,
      "num_tokens": 19945485.0,
      "step": 9105
    },
    {
      "entropy": 5.563791227340698,
      "epoch": 0.9041286224692339,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.3171,
      "mean_token_accuracy": 0.1616191580891609,
      "num_tokens": 19956561.0,
      "step": 9110
    },
    {
      "entropy": 5.604629802703857,
      "epoch": 0.9046248511314013,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.3857,
      "mean_token_accuracy": 0.15528203099966048,
      "num_tokens": 19967775.0,
      "step": 9115
    },
    {
      "entropy": 5.603000783920288,
      "epoch": 0.9051210797935689,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.3751,
      "mean_token_accuracy": 0.15253581255674362,
      "num_tokens": 19978935.0,
      "step": 9120
    },
    {
      "entropy": 5.539563226699829,
      "epoch": 0.9056173084557364,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.2297,
      "mean_token_accuracy": 0.17096547782421112,
      "num_tokens": 19989137.0,
      "step": 9125
    },
    {
      "entropy": 5.501555061340332,
      "epoch": 0.9061135371179039,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16666569113731383,
      "num_tokens": 20000507.0,
      "step": 9130
    },
    {
      "entropy": 5.574728012084961,
      "epoch": 0.9066097657800715,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.3375,
      "mean_token_accuracy": 0.15870152711868285,
      "num_tokens": 20010907.0,
      "step": 9135
    },
    {
      "entropy": 5.629899215698242,
      "epoch": 0.9071059944422389,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.3167,
      "mean_token_accuracy": 0.159600105881691,
      "num_tokens": 20021403.0,
      "step": 9140
    },
    {
      "entropy": 5.573432111740113,
      "epoch": 0.9076022231044065,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.2977,
      "mean_token_accuracy": 0.1586015537381172,
      "num_tokens": 20032409.0,
      "step": 9145
    },
    {
      "entropy": 5.572761297225952,
      "epoch": 0.908098451766574,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.2513,
      "mean_token_accuracy": 0.1626386508345604,
      "num_tokens": 20043197.0,
      "step": 9150
    },
    {
      "entropy": 5.58721604347229,
      "epoch": 0.9085946804287416,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.3784,
      "mean_token_accuracy": 0.1552750751376152,
      "num_tokens": 20054496.0,
      "step": 9155
    },
    {
      "entropy": 5.626295709609986,
      "epoch": 0.9090909090909091,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.4032,
      "mean_token_accuracy": 0.15426109582185746,
      "num_tokens": 20065639.0,
      "step": 9160
    },
    {
      "entropy": 5.641341018676758,
      "epoch": 0.9095871377530766,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.3071,
      "mean_token_accuracy": 0.16040030121803284,
      "num_tokens": 20077354.0,
      "step": 9165
    },
    {
      "entropy": 5.602892160415649,
      "epoch": 0.9100833664152441,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.3337,
      "mean_token_accuracy": 0.15840072482824324,
      "num_tokens": 20089088.0,
      "step": 9170
    },
    {
      "entropy": 5.70202751159668,
      "epoch": 0.9105795950774117,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.4918,
      "mean_token_accuracy": 0.15192114636301995,
      "num_tokens": 20101496.0,
      "step": 9175
    },
    {
      "entropy": 5.621792888641357,
      "epoch": 0.9110758237395792,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000492576237410368,
      "loss": 5.3043,
      "mean_token_accuracy": 0.15790807604789733,
      "num_tokens": 20113682.0,
      "step": 9180
    },
    {
      "entropy": 5.444388818740845,
      "epoch": 0.9115720524017468,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.1765,
      "mean_token_accuracy": 0.17713022530078887,
      "num_tokens": 20123831.0,
      "step": 9185
    },
    {
      "entropy": 5.559485006332397,
      "epoch": 0.9120682810639142,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.3221,
      "mean_token_accuracy": 0.16382687538862228,
      "num_tokens": 20133414.0,
      "step": 9190
    },
    {
      "entropy": 5.5272163391113285,
      "epoch": 0.9125645097260818,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.2653,
      "mean_token_accuracy": 0.16380606591701508,
      "num_tokens": 20145120.0,
      "step": 9195
    },
    {
      "entropy": 5.555481386184693,
      "epoch": 0.9130607383882493,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.2857,
      "mean_token_accuracy": 0.1593401163816452,
      "num_tokens": 20156807.0,
      "step": 9200
    },
    {
      "entropy": 5.607868814468384,
      "epoch": 0.9135569670504168,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.3112,
      "mean_token_accuracy": 0.1611064612865448,
      "num_tokens": 20166775.0,
      "step": 9205
    },
    {
      "entropy": 5.569726419448853,
      "epoch": 0.9140531957125844,
      "grad_norm": 0.78515625,
      "learning_rate": 0.000492521982090256,
      "loss": 5.2695,
      "mean_token_accuracy": 0.16777637898921965,
      "num_tokens": 20178695.0,
      "step": 9210
    },
    {
      "entropy": 5.695722675323486,
      "epoch": 0.9145494243747518,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.3723,
      "mean_token_accuracy": 0.15836737304925919,
      "num_tokens": 20190222.0,
      "step": 9215
    },
    {
      "entropy": 5.610282516479492,
      "epoch": 0.9150456530369194,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.328,
      "mean_token_accuracy": 0.1555210366845131,
      "num_tokens": 20202519.0,
      "step": 9220
    },
    {
      "entropy": 5.583818960189819,
      "epoch": 0.9155418816990869,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.4213,
      "mean_token_accuracy": 0.15217563658952712,
      "num_tokens": 20214685.0,
      "step": 9225
    },
    {
      "entropy": 5.682659673690796,
      "epoch": 0.9160381103612545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049248570399233,
      "loss": 5.3667,
      "mean_token_accuracy": 0.15235009789466858,
      "num_tokens": 20225597.0,
      "step": 9230
    },
    {
      "entropy": 5.603179168701172,
      "epoch": 0.916534339023422,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492476620984344,
      "loss": 5.3443,
      "mean_token_accuracy": 0.16417841464281083,
      "num_tokens": 20237067.0,
      "step": 9235
    },
    {
      "entropy": 5.69936842918396,
      "epoch": 0.9170305676855895,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1566211089491844,
      "num_tokens": 20247587.0,
      "step": 9240
    },
    {
      "entropy": 5.6290325164794925,
      "epoch": 0.917526796347757,
      "grad_norm": 0.875,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.3692,
      "mean_token_accuracy": 0.15845018476247788,
      "num_tokens": 20259094.0,
      "step": 9245
    },
    {
      "entropy": 5.5913111686706545,
      "epoch": 0.9180230250099246,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.2932,
      "mean_token_accuracy": 0.1615595653653145,
      "num_tokens": 20269776.0,
      "step": 9250
    },
    {
      "entropy": 5.59716796875,
      "epoch": 0.9185192536720921,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000492440235025138,
      "loss": 5.3169,
      "mean_token_accuracy": 0.1535188302397728,
      "num_tokens": 20281814.0,
      "step": 9255
    },
    {
      "entropy": 5.5827878475189205,
      "epoch": 0.9190154823342597,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.3749,
      "mean_token_accuracy": 0.1557564079761505,
      "num_tokens": 20293185.0,
      "step": 9260
    },
    {
      "entropy": 5.564189434051514,
      "epoch": 0.9195117109964271,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.2467,
      "mean_token_accuracy": 0.16347389221191405,
      "num_tokens": 20305097.0,
      "step": 9265
    },
    {
      "entropy": 5.576847124099731,
      "epoch": 0.9200079396585947,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.3183,
      "mean_token_accuracy": 0.16169384121894836,
      "num_tokens": 20316257.0,
      "step": 9270
    },
    {
      "entropy": 5.5875959396362305,
      "epoch": 0.9205041683207622,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.2673,
      "mean_token_accuracy": 0.16148687452077864,
      "num_tokens": 20326209.0,
      "step": 9275
    },
    {
      "entropy": 5.635387516021728,
      "epoch": 0.9210003969829297,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.3137,
      "mean_token_accuracy": 0.16084765791893005,
      "num_tokens": 20337665.0,
      "step": 9280
    },
    {
      "entropy": 5.557839250564575,
      "epoch": 0.9214966256450973,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.3496,
      "mean_token_accuracy": 0.15164009630680084,
      "num_tokens": 20350074.0,
      "step": 9285
    },
    {
      "entropy": 5.5968907356262205,
      "epoch": 0.9219928543072647,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.3075,
      "mean_token_accuracy": 0.1545549139380455,
      "num_tokens": 20361058.0,
      "step": 9290
    },
    {
      "entropy": 5.510803318023681,
      "epoch": 0.9224890829694323,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.2643,
      "mean_token_accuracy": 0.16046431362628938,
      "num_tokens": 20372605.0,
      "step": 9295
    },
    {
      "entropy": 5.614958381652832,
      "epoch": 0.9229853116315998,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.3764,
      "mean_token_accuracy": 0.1551586389541626,
      "num_tokens": 20383512.0,
      "step": 9300
    },
    {
      "entropy": 5.639304971694946,
      "epoch": 0.9234815402937674,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.353,
      "mean_token_accuracy": 0.15185174569487572,
      "num_tokens": 20395862.0,
      "step": 9305
    },
    {
      "entropy": 5.477538967132569,
      "epoch": 0.9239777689559349,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.1878,
      "mean_token_accuracy": 0.16464863419532777,
      "num_tokens": 20407349.0,
      "step": 9310
    },
    {
      "entropy": 5.607826232910156,
      "epoch": 0.9244739976181024,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.3928,
      "mean_token_accuracy": 0.1599385693669319,
      "num_tokens": 20419468.0,
      "step": 9315
    },
    {
      "entropy": 5.572306442260742,
      "epoch": 0.9249702262802699,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.2063,
      "mean_token_accuracy": 0.1640086904168129,
      "num_tokens": 20429689.0,
      "step": 9320
    },
    {
      "entropy": 5.5462327003479,
      "epoch": 0.9254664549424375,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.3152,
      "mean_token_accuracy": 0.15812406837940216,
      "num_tokens": 20440232.0,
      "step": 9325
    },
    {
      "entropy": 5.5461005687713625,
      "epoch": 0.925962683604605,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.3024,
      "mean_token_accuracy": 0.15408745259046555,
      "num_tokens": 20451021.0,
      "step": 9330
    },
    {
      "entropy": 5.596753549575806,
      "epoch": 0.9264589122667726,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.2649,
      "mean_token_accuracy": 0.15523408949375153,
      "num_tokens": 20461301.0,
      "step": 9335
    },
    {
      "entropy": 5.599203777313233,
      "epoch": 0.92695514092894,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.3409,
      "mean_token_accuracy": 0.15896366834640502,
      "num_tokens": 20472232.0,
      "step": 9340
    },
    {
      "entropy": 5.604495573043823,
      "epoch": 0.9274513695911076,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.3755,
      "mean_token_accuracy": 0.15549518764019013,
      "num_tokens": 20483063.0,
      "step": 9345
    },
    {
      "entropy": 5.604168367385864,
      "epoch": 0.9279475982532751,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.3255,
      "mean_token_accuracy": 0.1573859289288521,
      "num_tokens": 20493631.0,
      "step": 9350
    },
    {
      "entropy": 5.545047044754028,
      "epoch": 0.9284438269154426,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.3286,
      "mean_token_accuracy": 0.15994380861520768,
      "num_tokens": 20505173.0,
      "step": 9355
    },
    {
      "entropy": 5.589771032333374,
      "epoch": 0.9289400555776102,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.2675,
      "mean_token_accuracy": 0.16269598603248597,
      "num_tokens": 20516354.0,
      "step": 9360
    },
    {
      "entropy": 5.496107959747315,
      "epoch": 0.9294362842397776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1724638894200325,
      "num_tokens": 20528531.0,
      "step": 9365
    },
    {
      "entropy": 5.556259679794311,
      "epoch": 0.9299325129019452,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.2596,
      "mean_token_accuracy": 0.16702480167150496,
      "num_tokens": 20539957.0,
      "step": 9370
    },
    {
      "entropy": 5.540766763687134,
      "epoch": 0.9304287415641127,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.283,
      "mean_token_accuracy": 0.16127182841300963,
      "num_tokens": 20550656.0,
      "step": 9375
    },
    {
      "entropy": 5.57069582939148,
      "epoch": 0.9309249702262803,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.2968,
      "mean_token_accuracy": 0.16210304200649261,
      "num_tokens": 20561025.0,
      "step": 9380
    },
    {
      "entropy": 5.597545480728149,
      "epoch": 0.9314211988884478,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492201624039533,
      "loss": 5.229,
      "mean_token_accuracy": 0.15689384937286377,
      "num_tokens": 20570839.0,
      "step": 9385
    },
    {
      "entropy": 5.551278924942016,
      "epoch": 0.9319174275506154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.2517,
      "mean_token_accuracy": 0.1581224173307419,
      "num_tokens": 20581748.0,
      "step": 9390
    },
    {
      "entropy": 5.545231914520263,
      "epoch": 0.9324136562127828,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.2455,
      "mean_token_accuracy": 0.16763485968112946,
      "num_tokens": 20592639.0,
      "step": 9395
    },
    {
      "entropy": 5.545618057250977,
      "epoch": 0.9329098848749504,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1670430824160576,
      "num_tokens": 20603660.0,
      "step": 9400
    },
    {
      "entropy": 5.6173388957977295,
      "epoch": 0.9334061135371179,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.3622,
      "mean_token_accuracy": 0.15423354655504226,
      "num_tokens": 20614541.0,
      "step": 9405
    },
    {
      "entropy": 5.495154333114624,
      "epoch": 0.9339023421992855,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.2513,
      "mean_token_accuracy": 0.1634796753525734,
      "num_tokens": 20625274.0,
      "step": 9410
    },
    {
      "entropy": 5.563883209228516,
      "epoch": 0.934398570861453,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1603926494717598,
      "num_tokens": 20636282.0,
      "step": 9415
    },
    {
      "entropy": 5.611299180984497,
      "epoch": 0.9348947995236205,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1615723505616188,
      "num_tokens": 20647651.0,
      "step": 9420
    },
    {
      "entropy": 5.569247722625732,
      "epoch": 0.935391028185788,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.2812,
      "mean_token_accuracy": 0.1621583878993988,
      "num_tokens": 20657587.0,
      "step": 9425
    },
    {
      "entropy": 5.555886268615723,
      "epoch": 0.9358872568479555,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.3576,
      "mean_token_accuracy": 0.1605152368545532,
      "num_tokens": 20668672.0,
      "step": 9430
    },
    {
      "entropy": 5.682486724853516,
      "epoch": 0.9363834855101231,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.3623,
      "mean_token_accuracy": 0.16053014099597931,
      "num_tokens": 20680020.0,
      "step": 9435
    },
    {
      "entropy": 5.597968816757202,
      "epoch": 0.9368797141722905,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.342,
      "mean_token_accuracy": 0.15562722757458686,
      "num_tokens": 20690693.0,
      "step": 9440
    },
    {
      "entropy": 5.557184505462646,
      "epoch": 0.9373759428344581,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1651083841919899,
      "num_tokens": 20701943.0,
      "step": 9445
    },
    {
      "entropy": 5.599330186843872,
      "epoch": 0.9378721714966256,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.2655,
      "mean_token_accuracy": 0.16107137501239777,
      "num_tokens": 20711680.0,
      "step": 9450
    },
    {
      "entropy": 5.661785411834717,
      "epoch": 0.9383684001587932,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.3756,
      "mean_token_accuracy": 0.16237849742174149,
      "num_tokens": 20722273.0,
      "step": 9455
    },
    {
      "entropy": 5.526783752441406,
      "epoch": 0.9388646288209607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.2108,
      "mean_token_accuracy": 0.17029204368591308,
      "num_tokens": 20732914.0,
      "step": 9460
    },
    {
      "entropy": 5.5612101554870605,
      "epoch": 0.9393608574831283,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.3203,
      "mean_token_accuracy": 0.15463553071022035,
      "num_tokens": 20743485.0,
      "step": 9465
    },
    {
      "entropy": 5.613198137283325,
      "epoch": 0.9398570861452957,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.3492,
      "mean_token_accuracy": 0.15887228697538375,
      "num_tokens": 20754130.0,
      "step": 9470
    },
    {
      "entropy": 5.570826053619385,
      "epoch": 0.9403533148074633,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000492034298670543,
      "loss": 5.3254,
      "mean_token_accuracy": 0.16322394758462905,
      "num_tokens": 20766181.0,
      "step": 9475
    },
    {
      "entropy": 5.597544622421265,
      "epoch": 0.9408495434696308,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1631805568933487,
      "num_tokens": 20777321.0,
      "step": 9480
    },
    {
      "entropy": 5.549790477752685,
      "epoch": 0.9413457721317984,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.2189,
      "mean_token_accuracy": 0.16411449611186982,
      "num_tokens": 20788133.0,
      "step": 9485
    },
    {
      "entropy": 5.515652990341186,
      "epoch": 0.9418420007939659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.2606,
      "mean_token_accuracy": 0.15775451213121414,
      "num_tokens": 20798019.0,
      "step": 9490
    },
    {
      "entropy": 5.576691484451294,
      "epoch": 0.9423382294561334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.271,
      "mean_token_accuracy": 0.15867859721183777,
      "num_tokens": 20808449.0,
      "step": 9495
    },
    {
      "entropy": 5.509046745300293,
      "epoch": 0.9428344581183009,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.2424,
      "mean_token_accuracy": 0.16577329933643342,
      "num_tokens": 20819365.0,
      "step": 9500
    },
    {
      "entropy": 5.541382312774658,
      "epoch": 0.9433306867804684,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.287,
      "mean_token_accuracy": 0.15975593477487565,
      "num_tokens": 20829988.0,
      "step": 9505
    },
    {
      "entropy": 5.529103469848633,
      "epoch": 0.943826915442636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.2785,
      "mean_token_accuracy": 0.15802611261606217,
      "num_tokens": 20841047.0,
      "step": 9510
    },
    {
      "entropy": 5.532830142974854,
      "epoch": 0.9443231441048034,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000491959372033607,
      "loss": 5.2733,
      "mean_token_accuracy": 0.16636740863323213,
      "num_tokens": 20852652.0,
      "step": 9515
    },
    {
      "entropy": 5.649986839294433,
      "epoch": 0.944819372766971,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.3777,
      "mean_token_accuracy": 0.15682994127273558,
      "num_tokens": 20863558.0,
      "step": 9520
    },
    {
      "entropy": 5.606718921661377,
      "epoch": 0.9453156014291385,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.3215,
      "mean_token_accuracy": 0.1540657788515091,
      "num_tokens": 20873906.0,
      "step": 9525
    },
    {
      "entropy": 5.5851610660552975,
      "epoch": 0.9458118300913061,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491931185766302,
      "loss": 5.3894,
      "mean_token_accuracy": 0.16129270046949387,
      "num_tokens": 20885205.0,
      "step": 9530
    },
    {
      "entropy": 5.611719417572021,
      "epoch": 0.9463080587534736,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.246,
      "mean_token_accuracy": 0.16707937866449357,
      "num_tokens": 20895172.0,
      "step": 9535
    },
    {
      "entropy": 5.636444520950318,
      "epoch": 0.9468042874156412,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.4714,
      "mean_token_accuracy": 0.15052906423807144,
      "num_tokens": 20907454.0,
      "step": 9540
    },
    {
      "entropy": 5.610788631439209,
      "epoch": 0.9473005160778086,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.3514,
      "mean_token_accuracy": 0.1658547952771187,
      "num_tokens": 20919080.0,
      "step": 9545
    },
    {
      "entropy": 5.640568113327026,
      "epoch": 0.9477967447399762,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.2916,
      "mean_token_accuracy": 0.15813243091106416,
      "num_tokens": 20930807.0,
      "step": 9550
    },
    {
      "entropy": 5.59441647529602,
      "epoch": 0.9482929734021437,
      "grad_norm": 0.9375,
      "learning_rate": 0.000491884101068721,
      "loss": 5.2736,
      "mean_token_accuracy": 0.15802280902862548,
      "num_tokens": 20941255.0,
      "step": 9555
    },
    {
      "entropy": 5.530737066268921,
      "epoch": 0.9487892020643113,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.2777,
      "mean_token_accuracy": 0.16303081810474396,
      "num_tokens": 20951008.0,
      "step": 9560
    },
    {
      "entropy": 5.636520051956177,
      "epoch": 0.9492854307264788,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.393,
      "mean_token_accuracy": 0.1571933925151825,
      "num_tokens": 20963296.0,
      "step": 9565
    },
    {
      "entropy": 5.703459167480469,
      "epoch": 0.9497816593886463,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1529134213924408,
      "num_tokens": 20974995.0,
      "step": 9570
    },
    {
      "entropy": 5.6474607467651365,
      "epoch": 0.9502778880508138,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.3498,
      "mean_token_accuracy": 0.15372801423072815,
      "num_tokens": 20986297.0,
      "step": 9575
    },
    {
      "entropy": 5.578103828430176,
      "epoch": 0.9507741167129813,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.2948,
      "mean_token_accuracy": 0.1626210853457451,
      "num_tokens": 20997871.0,
      "step": 9580
    },
    {
      "entropy": 5.5420936107635494,
      "epoch": 0.9512703453751489,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.2992,
      "mean_token_accuracy": 0.1585755854845047,
      "num_tokens": 21010295.0,
      "step": 9585
    },
    {
      "entropy": 5.510882186889648,
      "epoch": 0.9517665740373163,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.2506,
      "mean_token_accuracy": 0.15988150909543036,
      "num_tokens": 21021307.0,
      "step": 9590
    },
    {
      "entropy": 5.5568835735321045,
      "epoch": 0.9522628026994839,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.2573,
      "mean_token_accuracy": 0.1667869657278061,
      "num_tokens": 21032325.0,
      "step": 9595
    },
    {
      "entropy": 5.605762243270874,
      "epoch": 0.9527590313616514,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.2847,
      "mean_token_accuracy": 0.16267508864402772,
      "num_tokens": 21043881.0,
      "step": 9600
    },
    {
      "entropy": 5.538880252838135,
      "epoch": 0.953255260023819,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.2701,
      "mean_token_accuracy": 0.16219368129968642,
      "num_tokens": 21054042.0,
      "step": 9605
    },
    {
      "entropy": 5.5494959354400635,
      "epoch": 0.9537514886859865,
      "grad_norm": 0.875,
      "learning_rate": 0.000491780041488553,
      "loss": 5.362,
      "mean_token_accuracy": 0.1615282714366913,
      "num_tokens": 21064374.0,
      "step": 9610
    },
    {
      "entropy": 5.592586469650269,
      "epoch": 0.9542477173481541,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.3383,
      "mean_token_accuracy": 0.16091350317001343,
      "num_tokens": 21075525.0,
      "step": 9615
    },
    {
      "entropy": 5.598389148712158,
      "epoch": 0.9547439460103215,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.2486,
      "mean_token_accuracy": 0.16410970091819763,
      "num_tokens": 21086259.0,
      "step": 9620
    },
    {
      "entropy": 5.549150848388672,
      "epoch": 0.9552401746724891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.2117,
      "mean_token_accuracy": 0.17003200352191924,
      "num_tokens": 21096987.0,
      "step": 9625
    },
    {
      "entropy": 5.65834527015686,
      "epoch": 0.9557364033346566,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049174204035444,
      "loss": 5.339,
      "mean_token_accuracy": 0.1531968355178833,
      "num_tokens": 21108220.0,
      "step": 9630
    },
    {
      "entropy": 5.615423154830933,
      "epoch": 0.9562326319968242,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.3307,
      "mean_token_accuracy": 0.16099415421485902,
      "num_tokens": 21118679.0,
      "step": 9635
    },
    {
      "entropy": 5.614273309707642,
      "epoch": 0.9567288606589917,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.3132,
      "mean_token_accuracy": 0.16014388352632522,
      "num_tokens": 21129945.0,
      "step": 9640
    },
    {
      "entropy": 5.619975280761719,
      "epoch": 0.9572250893211592,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.3537,
      "mean_token_accuracy": 0.15905492901802062,
      "num_tokens": 21140577.0,
      "step": 9645
    },
    {
      "entropy": 5.580228471755982,
      "epoch": 0.9577213179833267,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.2303,
      "mean_token_accuracy": 0.16249371469020843,
      "num_tokens": 21151041.0,
      "step": 9650
    },
    {
      "entropy": 5.550718545913696,
      "epoch": 0.9582175466454942,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.3235,
      "mean_token_accuracy": 0.15899259448051453,
      "num_tokens": 21161401.0,
      "step": 9655
    },
    {
      "entropy": 5.619397592544556,
      "epoch": 0.9587137753076618,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.3566,
      "mean_token_accuracy": 0.156053064763546,
      "num_tokens": 21170897.0,
      "step": 9660
    },
    {
      "entropy": 5.621097755432129,
      "epoch": 0.9592100039698293,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491675331451519,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15335856303572654,
      "num_tokens": 21183313.0,
      "step": 9665
    },
    {
      "entropy": 5.609352684020996,
      "epoch": 0.9597062326319968,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.2722,
      "mean_token_accuracy": 0.1617807224392891,
      "num_tokens": 21192820.0,
      "step": 9670
    },
    {
      "entropy": 5.572204160690307,
      "epoch": 0.9602024612941643,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.2331,
      "mean_token_accuracy": 0.17172446846961975,
      "num_tokens": 21203493.0,
      "step": 9675
    },
    {
      "entropy": 5.594243764877319,
      "epoch": 0.9606986899563319,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.255,
      "mean_token_accuracy": 0.15715011656284333,
      "num_tokens": 21214345.0,
      "step": 9680
    },
    {
      "entropy": 5.591793775558472,
      "epoch": 0.9611949186184994,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.2803,
      "mean_token_accuracy": 0.16009293347597123,
      "num_tokens": 21224345.0,
      "step": 9685
    },
    {
      "entropy": 5.559533596038818,
      "epoch": 0.961691147280667,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.3837,
      "mean_token_accuracy": 0.15860463976860045,
      "num_tokens": 21235938.0,
      "step": 9690
    },
    {
      "entropy": 5.641740226745606,
      "epoch": 0.9621873759428344,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.3892,
      "mean_token_accuracy": 0.15687987804412842,
      "num_tokens": 21247670.0,
      "step": 9695
    },
    {
      "entropy": 5.659240245819092,
      "epoch": 0.962683604605002,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.2684,
      "mean_token_accuracy": 0.1622430443763733,
      "num_tokens": 21257982.0,
      "step": 9700
    },
    {
      "entropy": 5.5047314167022705,
      "epoch": 0.9631798332671695,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.2253,
      "mean_token_accuracy": 0.16433116793632507,
      "num_tokens": 21268044.0,
      "step": 9705
    },
    {
      "entropy": 5.582844591140747,
      "epoch": 0.9636760619293371,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.2928,
      "mean_token_accuracy": 0.1639973595738411,
      "num_tokens": 21278331.0,
      "step": 9710
    },
    {
      "entropy": 5.617759656906128,
      "epoch": 0.9641722905915046,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.323,
      "mean_token_accuracy": 0.15305657535791398,
      "num_tokens": 21287225.0,
      "step": 9715
    },
    {
      "entropy": 5.610102558135987,
      "epoch": 0.9646685192536721,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000491569971271806,
      "loss": 5.3898,
      "mean_token_accuracy": 0.1510537788271904,
      "num_tokens": 21298607.0,
      "step": 9720
    },
    {
      "entropy": 5.62746753692627,
      "epoch": 0.9651647479158396,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1576084762811661,
      "num_tokens": 21311308.0,
      "step": 9725
    },
    {
      "entropy": 5.608367586135865,
      "epoch": 0.9656609765780071,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.3783,
      "mean_token_accuracy": 0.1556979686021805,
      "num_tokens": 21322370.0,
      "step": 9730
    },
    {
      "entropy": 5.557743501663208,
      "epoch": 0.9661572052401747,
      "grad_norm": 0.859375,
      "learning_rate": 0.000491541123884178,
      "loss": 5.23,
      "mean_token_accuracy": 0.16177124977111818,
      "num_tokens": 21333488.0,
      "step": 9735
    },
    {
      "entropy": 5.6245378971099855,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.4268,
      "mean_token_accuracy": 0.1553826704621315,
      "num_tokens": 21344907.0,
      "step": 9740
    },
    {
      "entropy": 5.66646294593811,
      "epoch": 0.9671496625645097,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.4075,
      "mean_token_accuracy": 0.1596427157521248,
      "num_tokens": 21355983.0,
      "step": 9745
    },
    {
      "entropy": 5.697890186309815,
      "epoch": 0.9676458912266772,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15834713876247405,
      "num_tokens": 21366313.0,
      "step": 9750
    },
    {
      "entropy": 5.65020055770874,
      "epoch": 0.9681421198888448,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.3864,
      "mean_token_accuracy": 0.15253149718046188,
      "num_tokens": 21377239.0,
      "step": 9755
    },
    {
      "entropy": 5.584231901168823,
      "epoch": 0.9686383485510123,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.3817,
      "mean_token_accuracy": 0.15608368515968324,
      "num_tokens": 21387622.0,
      "step": 9760
    },
    {
      "entropy": 5.560739421844483,
      "epoch": 0.9691345772131799,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.2403,
      "mean_token_accuracy": 0.15900204330682755,
      "num_tokens": 21397298.0,
      "step": 9765
    },
    {
      "entropy": 5.542948198318482,
      "epoch": 0.9696308058753473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000491473625382833,
      "loss": 5.2429,
      "mean_token_accuracy": 0.15814067125320436,
      "num_tokens": 21408444.0,
      "step": 9770
    },
    {
      "entropy": 5.638873958587647,
      "epoch": 0.9701270345375149,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491463961265551,
      "loss": 5.3845,
      "mean_token_accuracy": 0.16476238518953323,
      "num_tokens": 21420313.0,
      "step": 9775
    },
    {
      "entropy": 5.600589942932129,
      "epoch": 0.9706232631996824,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.3138,
      "mean_token_accuracy": 0.1586429312825203,
      "num_tokens": 21432068.0,
      "step": 9780
    },
    {
      "entropy": 5.6518560409545895,
      "epoch": 0.97111949186185,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.3904,
      "mean_token_accuracy": 0.15789133459329605,
      "num_tokens": 21442824.0,
      "step": 9785
    },
    {
      "entropy": 5.6646990299224855,
      "epoch": 0.9716157205240175,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.3866,
      "mean_token_accuracy": 0.15178794115781785,
      "num_tokens": 21452991.0,
      "step": 9790
    },
    {
      "entropy": 5.525694751739502,
      "epoch": 0.972111949186185,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.2307,
      "mean_token_accuracy": 0.16449764966964722,
      "num_tokens": 21465610.0,
      "step": 9795
    },
    {
      "entropy": 5.4984368801116945,
      "epoch": 0.9726081778483525,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.2223,
      "mean_token_accuracy": 0.16715873777866364,
      "num_tokens": 21475951.0,
      "step": 9800
    },
    {
      "entropy": 5.602439498901367,
      "epoch": 0.97310440651052,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1557515874505043,
      "num_tokens": 21486101.0,
      "step": 9805
    },
    {
      "entropy": 5.603495454788208,
      "epoch": 0.9736006351726876,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.267,
      "mean_token_accuracy": 0.1645111158490181,
      "num_tokens": 21497494.0,
      "step": 9810
    },
    {
      "entropy": 5.489730596542358,
      "epoch": 0.974096863834855,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.2281,
      "mean_token_accuracy": 0.16488623917102813,
      "num_tokens": 21507840.0,
      "step": 9815
    },
    {
      "entropy": 5.5775861740112305,
      "epoch": 0.9745930924970226,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.2979,
      "mean_token_accuracy": 0.1620860442519188,
      "num_tokens": 21518099.0,
      "step": 9820
    },
    {
      "entropy": 5.551573991775513,
      "epoch": 0.9750893211591901,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.2641,
      "mean_token_accuracy": 0.160970102250576,
      "num_tokens": 21529465.0,
      "step": 9825
    },
    {
      "entropy": 5.546416234970093,
      "epoch": 0.9755855498213577,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.3225,
      "mean_token_accuracy": 0.15884360074996948,
      "num_tokens": 21540206.0,
      "step": 9830
    },
    {
      "entropy": 5.634373950958252,
      "epoch": 0.9760817784835252,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.319,
      "mean_token_accuracy": 0.16149636059999467,
      "num_tokens": 21551055.0,
      "step": 9835
    },
    {
      "entropy": 5.541636037826538,
      "epoch": 0.9765780071456928,
      "grad_norm": 0.875,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.2583,
      "mean_token_accuracy": 0.15903128385543824,
      "num_tokens": 21562001.0,
      "step": 9840
    },
    {
      "entropy": 5.5050465106964115,
      "epoch": 0.9770742358078602,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.2789,
      "mean_token_accuracy": 0.15861424803733826,
      "num_tokens": 21571781.0,
      "step": 9845
    },
    {
      "entropy": 5.5796304702758786,
      "epoch": 0.9775704644700278,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.2701,
      "mean_token_accuracy": 0.16638385653495788,
      "num_tokens": 21581300.0,
      "step": 9850
    },
    {
      "entropy": 5.594467639923096,
      "epoch": 0.9780666931321953,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.3536,
      "mean_token_accuracy": 0.1550612434744835,
      "num_tokens": 21593540.0,
      "step": 9855
    },
    {
      "entropy": 5.596910858154297,
      "epoch": 0.9785629217943629,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.2462,
      "mean_token_accuracy": 0.16831225901842117,
      "num_tokens": 21603395.0,
      "step": 9860
    },
    {
      "entropy": 5.6809454441070555,
      "epoch": 0.9790591504565304,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.421,
      "mean_token_accuracy": 0.15276228934526442,
      "num_tokens": 21615508.0,
      "step": 9865
    },
    {
      "entropy": 5.485996055603027,
      "epoch": 0.9795553791186979,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.216,
      "mean_token_accuracy": 0.16633347868919374,
      "num_tokens": 21625832.0,
      "step": 9870
    },
    {
      "entropy": 5.54438157081604,
      "epoch": 0.9800516077808654,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.2835,
      "mean_token_accuracy": 0.16135922819375992,
      "num_tokens": 21637256.0,
      "step": 9875
    },
    {
      "entropy": 5.632312631607055,
      "epoch": 0.9805478364430329,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.3489,
      "mean_token_accuracy": 0.1508886530995369,
      "num_tokens": 21647944.0,
      "step": 9880
    },
    {
      "entropy": 5.568255805969239,
      "epoch": 0.9810440651052005,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.2088,
      "mean_token_accuracy": 0.16009017825126648,
      "num_tokens": 21658572.0,
      "step": 9885
    },
    {
      "entropy": 5.569843482971192,
      "epoch": 0.981540293767368,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.292,
      "mean_token_accuracy": 0.1592393174767494,
      "num_tokens": 21668782.0,
      "step": 9890
    },
    {
      "entropy": 5.574169921875,
      "epoch": 0.9820365224295355,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.3476,
      "mean_token_accuracy": 0.15564077347517014,
      "num_tokens": 21679820.0,
      "step": 9895
    },
    {
      "entropy": 5.5570581436157225,
      "epoch": 0.982532751091703,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1682171866297722,
      "num_tokens": 21690912.0,
      "step": 9900
    },
    {
      "entropy": 5.514180088043213,
      "epoch": 0.9830289797538706,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.2141,
      "mean_token_accuracy": 0.16139313131570815,
      "num_tokens": 21701653.0,
      "step": 9905
    },
    {
      "entropy": 5.553722858428955,
      "epoch": 0.9835252084160381,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.3364,
      "mean_token_accuracy": 0.15768067985773088,
      "num_tokens": 21713841.0,
      "step": 9910
    },
    {
      "entropy": 5.61244797706604,
      "epoch": 0.9840214370782057,
      "grad_norm": 0.90625,
      "learning_rate": 0.000491191187518024,
      "loss": 5.3094,
      "mean_token_accuracy": 0.1589959070086479,
      "num_tokens": 21724343.0,
      "step": 9915
    },
    {
      "entropy": 5.617506217956543,
      "epoch": 0.9845176657403731,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.3267,
      "mean_token_accuracy": 0.15407982468605042,
      "num_tokens": 21734424.0,
      "step": 9920
    },
    {
      "entropy": 5.515016078948975,
      "epoch": 0.9850138944025407,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.2859,
      "mean_token_accuracy": 0.1618168830871582,
      "num_tokens": 21745001.0,
      "step": 9925
    },
    {
      "entropy": 5.562051248550415,
      "epoch": 0.9855101230647082,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.2565,
      "mean_token_accuracy": 0.16531966030597686,
      "num_tokens": 21755838.0,
      "step": 9930
    },
    {
      "entropy": 5.548488283157349,
      "epoch": 0.9860063517268758,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.2237,
      "mean_token_accuracy": 0.16645463407039643,
      "num_tokens": 21768057.0,
      "step": 9935
    },
    {
      "entropy": 5.580961751937866,
      "epoch": 0.9865025803890433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.3505,
      "mean_token_accuracy": 0.15447235703468323,
      "num_tokens": 21779136.0,
      "step": 9940
    },
    {
      "entropy": 5.6001537322998045,
      "epoch": 0.9869988090512108,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.3088,
      "mean_token_accuracy": 0.16008486822247506,
      "num_tokens": 21790556.0,
      "step": 9945
    },
    {
      "entropy": 5.6215332508087155,
      "epoch": 0.9874950377133783,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.2991,
      "mean_token_accuracy": 0.16555699855089187,
      "num_tokens": 21801931.0,
      "step": 9950
    },
    {
      "entropy": 5.505740690231323,
      "epoch": 0.9879912663755459,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.3419,
      "mean_token_accuracy": 0.15602683201432227,
      "num_tokens": 21813923.0,
      "step": 9955
    },
    {
      "entropy": 5.5032895565032955,
      "epoch": 0.9884874950377134,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.2043,
      "mean_token_accuracy": 0.16960564702749253,
      "num_tokens": 21825230.0,
      "step": 9960
    },
    {
      "entropy": 5.566311168670654,
      "epoch": 0.9889837236998809,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.3308,
      "mean_token_accuracy": 0.15750246196985246,
      "num_tokens": 21835047.0,
      "step": 9965
    },
    {
      "entropy": 5.570126152038574,
      "epoch": 0.9894799523620484,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.2762,
      "mean_token_accuracy": 0.15562335550785064,
      "num_tokens": 21845063.0,
      "step": 9970
    },
    {
      "entropy": 5.591881799697876,
      "epoch": 0.9899761810242159,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.2268,
      "mean_token_accuracy": 0.1651510030031204,
      "num_tokens": 21856283.0,
      "step": 9975
    },
    {
      "entropy": 5.552926158905029,
      "epoch": 0.9904724096863835,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.2817,
      "mean_token_accuracy": 0.16413484513759613,
      "num_tokens": 21867159.0,
      "step": 9980
    },
    {
      "entropy": 5.573501062393189,
      "epoch": 0.990968638348551,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.2772,
      "mean_token_accuracy": 0.16077665388584136,
      "num_tokens": 21877373.0,
      "step": 9985
    },
    {
      "entropy": 5.523486328125,
      "epoch": 0.9914648670107186,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.2948,
      "mean_token_accuracy": 0.1622526913881302,
      "num_tokens": 21889855.0,
      "step": 9990
    },
    {
      "entropy": 5.484282732009888,
      "epoch": 0.991961095672886,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000491033429366081,
      "loss": 5.216,
      "mean_token_accuracy": 0.16845631450414658,
      "num_tokens": 21899522.0,
      "step": 9995
    },
    {
      "entropy": 5.503854465484619,
      "epoch": 0.9924573243350536,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.24,
      "mean_token_accuracy": 0.1672941491007805,
      "num_tokens": 21910578.0,
      "step": 10000
    },
    {
      "entropy": 5.5410528659820555,
      "epoch": 0.9929535529972211,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000491013613147562,
      "loss": 5.2826,
      "mean_token_accuracy": 0.15815844982862473,
      "num_tokens": 21921048.0,
      "step": 10005
    },
    {
      "entropy": 5.4939943790435795,
      "epoch": 0.9934497816593887,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.2762,
      "mean_token_accuracy": 0.15548515766859056,
      "num_tokens": 21930950.0,
      "step": 10010
    },
    {
      "entropy": 5.686042308807373,
      "epoch": 0.9939460103215562,
      "grad_norm": 1.0,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.413,
      "mean_token_accuracy": 0.1515097662806511,
      "num_tokens": 21941333.0,
      "step": 10015
    },
    {
      "entropy": 5.561811923980713,
      "epoch": 0.9944422389837237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1634193703532219,
      "num_tokens": 21951895.0,
      "step": 10020
    },
    {
      "entropy": 5.591481351852417,
      "epoch": 0.9949384676458912,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.321,
      "mean_token_accuracy": 0.15906060338020325,
      "num_tokens": 21963915.0,
      "step": 10025
    },
    {
      "entropy": 5.63429741859436,
      "epoch": 0.9954346963080588,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.3462,
      "mean_token_accuracy": 0.1596025750041008,
      "num_tokens": 21973653.0,
      "step": 10030
    },
    {
      "entropy": 5.633570289611816,
      "epoch": 0.9959309249702263,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.2829,
      "mean_token_accuracy": 0.16097168624401093,
      "num_tokens": 21985042.0,
      "step": 10035
    },
    {
      "entropy": 5.569746303558349,
      "epoch": 0.9964271536323938,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.3427,
      "mean_token_accuracy": 0.160613751411438,
      "num_tokens": 21996276.0,
      "step": 10040
    },
    {
      "entropy": 5.563586139678955,
      "epoch": 0.9969233822945613,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049093413402629,
      "loss": 5.3042,
      "mean_token_accuracy": 0.15861425548791885,
      "num_tokens": 22006774.0,
      "step": 10045
    },
    {
      "entropy": 5.533929872512817,
      "epoch": 0.9974196109567288,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.3239,
      "mean_token_accuracy": 0.1569620743393898,
      "num_tokens": 22017708.0,
      "step": 10050
    },
    {
      "entropy": 5.5235710620880125,
      "epoch": 0.9979158396188964,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16317325383424758,
      "num_tokens": 22030393.0,
      "step": 10055
    },
    {
      "entropy": 5.615251970291138,
      "epoch": 0.9984120682810639,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000490904240996462,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1638371840119362,
      "num_tokens": 22041124.0,
      "step": 10060
    },
    {
      "entropy": 5.568389558792115,
      "epoch": 0.9989082969432315,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1646386206150055,
      "num_tokens": 22052512.0,
      "step": 10065
    },
    {
      "entropy": 5.545720434188842,
      "epoch": 0.9994045256053989,
      "grad_norm": 0.875,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1589846730232239,
      "num_tokens": 22063689.0,
      "step": 10070
    },
    {
      "entropy": 5.519991493225097,
      "epoch": 0.9999007542675665,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.2146,
      "mean_token_accuracy": 0.16673877388238906,
      "num_tokens": 22074375.0,
      "step": 10075
    },
    {
      "entropy": 5.625807857513427,
      "epoch": 1.0003969829297341,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.3051,
      "mean_token_accuracy": 0.16275031715631486,
      "num_tokens": 22084040.0,
      "step": 10080
    },
    {
      "entropy": 5.55686388015747,
      "epoch": 1.0008932115919016,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000490854312204184,
      "loss": 5.1454,
      "mean_token_accuracy": 0.16498544663190842,
      "num_tokens": 22095964.0,
      "step": 10085
    },
    {
      "entropy": 5.4999556064605715,
      "epoch": 1.001389440254069,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.2298,
      "mean_token_accuracy": 0.1654888391494751,
      "num_tokens": 22106395.0,
      "step": 10090
    },
    {
      "entropy": 5.488574409484864,
      "epoch": 1.0018856689162365,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.078,
      "mean_token_accuracy": 0.17335053235292436,
      "num_tokens": 22115903.0,
      "step": 10095
    },
    {
      "entropy": 5.613149166107178,
      "epoch": 1.0023818975784042,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.2685,
      "mean_token_accuracy": 0.1640152931213379,
      "num_tokens": 22127302.0,
      "step": 10100
    },
    {
      "entropy": 5.635703611373901,
      "epoch": 1.0028781262405717,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.2964,
      "mean_token_accuracy": 0.1601553812623024,
      "num_tokens": 22136916.0,
      "step": 10105
    },
    {
      "entropy": 5.533606767654419,
      "epoch": 1.0033743549027392,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.1967,
      "mean_token_accuracy": 0.16874119639396667,
      "num_tokens": 22146857.0,
      "step": 10110
    },
    {
      "entropy": 5.597428655624389,
      "epoch": 1.0038705835649067,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.2282,
      "mean_token_accuracy": 0.1679922416806221,
      "num_tokens": 22158039.0,
      "step": 10115
    },
    {
      "entropy": 5.610249185562134,
      "epoch": 1.0043668122270741,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.2298,
      "mean_token_accuracy": 0.16430157274007798,
      "num_tokens": 22167779.0,
      "step": 10120
    },
    {
      "entropy": 5.509985160827637,
      "epoch": 1.0048630408892418,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.2215,
      "mean_token_accuracy": 0.16015431582927703,
      "num_tokens": 22179405.0,
      "step": 10125
    },
    {
      "entropy": 5.505718755722046,
      "epoch": 1.0053592695514093,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.2261,
      "mean_token_accuracy": 0.16293877214193345,
      "num_tokens": 22189716.0,
      "step": 10130
    },
    {
      "entropy": 5.595527839660645,
      "epoch": 1.0058554982135768,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000490754053196548,
      "loss": 5.1573,
      "mean_token_accuracy": 0.16752440631389617,
      "num_tokens": 22199610.0,
      "step": 10135
    },
    {
      "entropy": 5.606732130050659,
      "epoch": 1.0063517268757443,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1635689213871956,
      "num_tokens": 22210900.0,
      "step": 10140
    },
    {
      "entropy": 5.490957927703858,
      "epoch": 1.006847955537912,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.1675,
      "mean_token_accuracy": 0.17996809631586075,
      "num_tokens": 22222156.0,
      "step": 10145
    },
    {
      "entropy": 5.506067323684692,
      "epoch": 1.0073441842000794,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.1634,
      "mean_token_accuracy": 0.17115239500999452,
      "num_tokens": 22232744.0,
      "step": 10150
    },
    {
      "entropy": 5.5331563472747805,
      "epoch": 1.007840412862247,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16708440482616424,
      "num_tokens": 22242291.0,
      "step": 10155
    },
    {
      "entropy": 5.605658483505249,
      "epoch": 1.0083366415244144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000490703723043345,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16799066215753555,
      "num_tokens": 22253517.0,
      "step": 10160
    },
    {
      "entropy": 5.622158527374268,
      "epoch": 1.008832870186582,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.3033,
      "mean_token_accuracy": 0.15277321636676788,
      "num_tokens": 22265364.0,
      "step": 10165
    },
    {
      "entropy": 5.543975067138672,
      "epoch": 1.0093290988487496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000490683553537403,
      "loss": 5.2717,
      "mean_token_accuracy": 0.16844537556171418,
      "num_tokens": 22276169.0,
      "step": 10170
    },
    {
      "entropy": 5.590293979644775,
      "epoch": 1.009825327510917,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.2646,
      "mean_token_accuracy": 0.1639661341905594,
      "num_tokens": 22287451.0,
      "step": 10175
    },
    {
      "entropy": 5.563012886047363,
      "epoch": 1.0103215561730845,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.2203,
      "mean_token_accuracy": 0.16457379162311553,
      "num_tokens": 22298206.0,
      "step": 10180
    },
    {
      "entropy": 5.588278913497925,
      "epoch": 1.010817784835252,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.1394,
      "mean_token_accuracy": 0.17084644436836244,
      "num_tokens": 22308950.0,
      "step": 10185
    },
    {
      "entropy": 5.552133798599243,
      "epoch": 1.0113140134974197,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.1744,
      "mean_token_accuracy": 0.16630226969718934,
      "num_tokens": 22319875.0,
      "step": 10190
    },
    {
      "entropy": 5.502207660675049,
      "epoch": 1.0118102421595871,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1696072906255722,
      "num_tokens": 22330781.0,
      "step": 10195
    },
    {
      "entropy": 5.583018350601196,
      "epoch": 1.0123064708217546,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.1859,
      "mean_token_accuracy": 0.16928102672100068,
      "num_tokens": 22341459.0,
      "step": 10200
    },
    {
      "entropy": 5.566336917877197,
      "epoch": 1.012802699483922,
      "grad_norm": 0.875,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16433742344379426,
      "num_tokens": 22352360.0,
      "step": 10205
    },
    {
      "entropy": 5.528634929656983,
      "epoch": 1.0132989281460898,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.201,
      "mean_token_accuracy": 0.16199807673692704,
      "num_tokens": 22363027.0,
      "step": 10210
    },
    {
      "entropy": 5.600583839416504,
      "epoch": 1.0137951568082573,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.282,
      "mean_token_accuracy": 0.155699260532856,
      "num_tokens": 22374816.0,
      "step": 10215
    },
    {
      "entropy": 5.631434488296509,
      "epoch": 1.0142913854704247,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000490582385138539,
      "loss": 5.2724,
      "mean_token_accuracy": 0.1594749003648758,
      "num_tokens": 22385294.0,
      "step": 10220
    },
    {
      "entropy": 5.630405139923096,
      "epoch": 1.0147876141325922,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.2403,
      "mean_token_accuracy": 0.15684994012117387,
      "num_tokens": 22396435.0,
      "step": 10225
    },
    {
      "entropy": 5.5204249858856205,
      "epoch": 1.01528384279476,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.2016,
      "mean_token_accuracy": 0.17052147835493087,
      "num_tokens": 22407160.0,
      "step": 10230
    },
    {
      "entropy": 5.544656610488891,
      "epoch": 1.0157800714569274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.2581,
      "mean_token_accuracy": 0.1619655206799507,
      "num_tokens": 22419186.0,
      "step": 10235
    },
    {
      "entropy": 5.502834510803223,
      "epoch": 1.0162763001190949,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1617102101445198,
      "num_tokens": 22430884.0,
      "step": 10240
    },
    {
      "entropy": 5.548295497894287,
      "epoch": 1.0167725287812623,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1638963669538498,
      "num_tokens": 22441494.0,
      "step": 10245
    },
    {
      "entropy": 5.591971921920776,
      "epoch": 1.01726875744343,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.2715,
      "mean_token_accuracy": 0.16176241636276245,
      "num_tokens": 22453127.0,
      "step": 10250
    },
    {
      "entropy": 5.54666109085083,
      "epoch": 1.0177649861055975,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16689993292093278,
      "num_tokens": 22465230.0,
      "step": 10255
    },
    {
      "entropy": 5.581253337860107,
      "epoch": 1.018261214767765,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049050106549698,
      "loss": 5.1813,
      "mean_token_accuracy": 0.1654479295015335,
      "num_tokens": 22476450.0,
      "step": 10260
    },
    {
      "entropy": 5.6126690864562985,
      "epoch": 1.0187574434299325,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.2337,
      "mean_token_accuracy": 0.16068822592496873,
      "num_tokens": 22487557.0,
      "step": 10265
    },
    {
      "entropy": 5.536162996292115,
      "epoch": 1.0192536720921,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.2527,
      "mean_token_accuracy": 0.1694289267063141,
      "num_tokens": 22498893.0,
      "step": 10270
    },
    {
      "entropy": 5.503987264633179,
      "epoch": 1.0197499007542676,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.2187,
      "mean_token_accuracy": 0.1625882178544998,
      "num_tokens": 22509873.0,
      "step": 10275
    },
    {
      "entropy": 5.593803787231446,
      "epoch": 1.020246129416435,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.2183,
      "mean_token_accuracy": 0.16553791761398315,
      "num_tokens": 22521501.0,
      "step": 10280
    },
    {
      "entropy": 5.547065734863281,
      "epoch": 1.0207423580786026,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.226,
      "mean_token_accuracy": 0.16059738546609878,
      "num_tokens": 22532350.0,
      "step": 10285
    },
    {
      "entropy": 5.521934556961059,
      "epoch": 1.02123858674077,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490439851312355,
      "loss": 5.1689,
      "mean_token_accuracy": 0.16977608650922776,
      "num_tokens": 22542632.0,
      "step": 10290
    },
    {
      "entropy": 5.5724938869476315,
      "epoch": 1.0217348154029378,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.2406,
      "mean_token_accuracy": 0.15963857024908065,
      "num_tokens": 22552894.0,
      "step": 10295
    },
    {
      "entropy": 5.608418607711792,
      "epoch": 1.0222310440651052,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.2973,
      "mean_token_accuracy": 0.16497753858566283,
      "num_tokens": 22563279.0,
      "step": 10300
    },
    {
      "entropy": 5.563898038864136,
      "epoch": 1.0227272727272727,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1591017246246338,
      "num_tokens": 22574217.0,
      "step": 10305
    },
    {
      "entropy": 5.580491924285889,
      "epoch": 1.0232235013894402,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.2494,
      "mean_token_accuracy": 0.16106862127780913,
      "num_tokens": 22585001.0,
      "step": 10310
    },
    {
      "entropy": 5.620190954208374,
      "epoch": 1.0237197300516079,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.2696,
      "mean_token_accuracy": 0.15410591363906861,
      "num_tokens": 22596068.0,
      "step": 10315
    },
    {
      "entropy": 5.648692417144775,
      "epoch": 1.0242159587137754,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000490378444799757,
      "loss": 5.2323,
      "mean_token_accuracy": 0.15800314545631408,
      "num_tokens": 22606818.0,
      "step": 10320
    },
    {
      "entropy": 5.542697477340698,
      "epoch": 1.0247121873759428,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.2097,
      "mean_token_accuracy": 0.16680783927440643,
      "num_tokens": 22617429.0,
      "step": 10325
    },
    {
      "entropy": 5.470415782928467,
      "epoch": 1.0252084160381103,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.1609,
      "mean_token_accuracy": 0.1701019749045372,
      "num_tokens": 22628619.0,
      "step": 10330
    },
    {
      "entropy": 5.546237134933472,
      "epoch": 1.0257046447002778,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1673685058951378,
      "num_tokens": 22640753.0,
      "step": 10335
    },
    {
      "entropy": 5.5900781631469725,
      "epoch": 1.0262008733624455,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.3075,
      "mean_token_accuracy": 0.15137375742197037,
      "num_tokens": 22652814.0,
      "step": 10340
    },
    {
      "entropy": 5.640706872940063,
      "epoch": 1.026697102024613,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.2331,
      "mean_token_accuracy": 0.1595412850379944,
      "num_tokens": 22663848.0,
      "step": 10345
    },
    {
      "entropy": 5.5051850318908695,
      "epoch": 1.0271933306867804,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16063161194324493,
      "num_tokens": 22676859.0,
      "step": 10350
    },
    {
      "entropy": 5.549219942092895,
      "epoch": 1.027689559348948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16623704731464387,
      "num_tokens": 22687000.0,
      "step": 10355
    },
    {
      "entropy": 5.593934392929077,
      "epoch": 1.0281857880111156,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000490296270367553,
      "loss": 5.1793,
      "mean_token_accuracy": 0.16315631866455077,
      "num_tokens": 22697440.0,
      "step": 10360
    },
    {
      "entropy": 5.590612936019897,
      "epoch": 1.028682016673283,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.1897,
      "mean_token_accuracy": 0.16783479154109954,
      "num_tokens": 22709031.0,
      "step": 10365
    },
    {
      "entropy": 5.605007553100586,
      "epoch": 1.0291782453354505,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1559052884578705,
      "num_tokens": 22721247.0,
      "step": 10370
    },
    {
      "entropy": 5.5781842231750485,
      "epoch": 1.029674473997618,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.2456,
      "mean_token_accuracy": 0.16868440359830855,
      "num_tokens": 22731784.0,
      "step": 10375
    },
    {
      "entropy": 5.635866641998291,
      "epoch": 1.0301707026597857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.2636,
      "mean_token_accuracy": 0.15868718326091766,
      "num_tokens": 22742715.0,
      "step": 10380
    },
    {
      "entropy": 5.62136869430542,
      "epoch": 1.0306669313219532,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.3349,
      "mean_token_accuracy": 0.16320303231477737,
      "num_tokens": 22752086.0,
      "step": 10385
    },
    {
      "entropy": 5.561475610733032,
      "epoch": 1.0311631599841207,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.2167,
      "mean_token_accuracy": 0.17091167271137236,
      "num_tokens": 22764407.0,
      "step": 10390
    },
    {
      "entropy": 5.54537878036499,
      "epoch": 1.0316593886462881,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000490224087443552,
      "loss": 5.1998,
      "mean_token_accuracy": 0.16494240164756774,
      "num_tokens": 22774457.0,
      "step": 10395
    },
    {
      "entropy": 5.605111932754516,
      "epoch": 1.0321556173084558,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.2423,
      "mean_token_accuracy": 0.16786742508411406,
      "num_tokens": 22786726.0,
      "step": 10400
    },
    {
      "entropy": 5.593416547775268,
      "epoch": 1.0326518459706233,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.2297,
      "mean_token_accuracy": 0.16095243394374847,
      "num_tokens": 22797677.0,
      "step": 10405
    },
    {
      "entropy": 5.5067671775817875,
      "epoch": 1.0331480746327908,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16387224346399307,
      "num_tokens": 22808230.0,
      "step": 10410
    },
    {
      "entropy": 5.583348131179809,
      "epoch": 1.0336443032949583,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.2489,
      "mean_token_accuracy": 0.16171268075704576,
      "num_tokens": 22821111.0,
      "step": 10415
    },
    {
      "entropy": 5.515660953521729,
      "epoch": 1.0341405319571257,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.1651,
      "mean_token_accuracy": 0.16897316873073578,
      "num_tokens": 22832531.0,
      "step": 10420
    },
    {
      "entropy": 5.58558840751648,
      "epoch": 1.0346367606192934,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.2447,
      "mean_token_accuracy": 0.1678275376558304,
      "num_tokens": 22845065.0,
      "step": 10425
    },
    {
      "entropy": 5.556848573684692,
      "epoch": 1.035132989281461,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.2251,
      "mean_token_accuracy": 0.15832116305828095,
      "num_tokens": 22857652.0,
      "step": 10430
    },
    {
      "entropy": 5.528628015518189,
      "epoch": 1.0356292179436284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1657000571489334,
      "num_tokens": 22866827.0,
      "step": 10435
    },
    {
      "entropy": 5.5617283344268795,
      "epoch": 1.0361254466057959,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.2041,
      "mean_token_accuracy": 0.1652309015393257,
      "num_tokens": 22877669.0,
      "step": 10440
    },
    {
      "entropy": 5.564185857772827,
      "epoch": 1.0366216752679636,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16198566555976868,
      "num_tokens": 22888122.0,
      "step": 10445
    },
    {
      "entropy": 5.618378591537476,
      "epoch": 1.037117903930131,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.2647,
      "mean_token_accuracy": 0.16126612275838853,
      "num_tokens": 22898977.0,
      "step": 10450
    },
    {
      "entropy": 5.605776166915893,
      "epoch": 1.0376141325922985,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.2065,
      "mean_token_accuracy": 0.1621193304657936,
      "num_tokens": 22909666.0,
      "step": 10455
    },
    {
      "entropy": 5.555507040023803,
      "epoch": 1.038110361254466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.1917,
      "mean_token_accuracy": 0.15943103432655334,
      "num_tokens": 22920258.0,
      "step": 10460
    },
    {
      "entropy": 5.551351404190063,
      "epoch": 1.0386065899166337,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.3108,
      "mean_token_accuracy": 0.160049869120121,
      "num_tokens": 22931694.0,
      "step": 10465
    },
    {
      "entropy": 5.609781646728516,
      "epoch": 1.0391028185788012,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.2729,
      "mean_token_accuracy": 0.1614201158285141,
      "num_tokens": 22942462.0,
      "step": 10470
    },
    {
      "entropy": 5.545709037780762,
      "epoch": 1.0395990472409686,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1677052780985832,
      "num_tokens": 22954752.0,
      "step": 10475
    },
    {
      "entropy": 5.605498838424682,
      "epoch": 1.040095275903136,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000490047697464287,
      "loss": 5.2876,
      "mean_token_accuracy": 0.16603398248553275,
      "num_tokens": 22965630.0,
      "step": 10480
    },
    {
      "entropy": 5.559588289260864,
      "epoch": 1.0405915045653038,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.189,
      "mean_token_accuracy": 0.16801199018955232,
      "num_tokens": 22976403.0,
      "step": 10485
    },
    {
      "entropy": 5.537876081466675,
      "epoch": 1.0410877332274713,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.1721,
      "mean_token_accuracy": 0.16360948532819747,
      "num_tokens": 22987558.0,
      "step": 10490
    },
    {
      "entropy": 5.555220079421997,
      "epoch": 1.0415839618896388,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000490016409803204,
      "loss": 5.2312,
      "mean_token_accuracy": 0.1657511293888092,
      "num_tokens": 22998340.0,
      "step": 10495
    },
    {
      "entropy": 5.526044130325317,
      "epoch": 1.0420801905518062,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.1839,
      "mean_token_accuracy": 0.1617182105779648,
      "num_tokens": 23010058.0,
      "step": 10500
    },
    {
      "entropy": 5.527077150344849,
      "epoch": 1.0425764192139737,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16861534267663955,
      "num_tokens": 23020777.0,
      "step": 10505
    },
    {
      "entropy": 5.596467351913452,
      "epoch": 1.0430726478761414,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000489985074154629,
      "loss": 5.3478,
      "mean_token_accuracy": 0.15816236138343812,
      "num_tokens": 23033510.0,
      "step": 10510
    },
    {
      "entropy": 5.6106781482696535,
      "epoch": 1.0435688765383089,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16901810020208358,
      "num_tokens": 23044385.0,
      "step": 10515
    },
    {
      "entropy": 5.546474981307983,
      "epoch": 1.0440651052004764,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.1973,
      "mean_token_accuracy": 0.16702024638652802,
      "num_tokens": 23054907.0,
      "step": 10520
    },
    {
      "entropy": 5.530324554443359,
      "epoch": 1.0445613338626438,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.3038,
      "mean_token_accuracy": 0.161403888463974,
      "num_tokens": 23067035.0,
      "step": 10525
    },
    {
      "entropy": 5.576786088943481,
      "epoch": 1.0450575625248115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.2313,
      "mean_token_accuracy": 0.16473373770713806,
      "num_tokens": 23077084.0,
      "step": 10530
    },
    {
      "entropy": 5.597245121002198,
      "epoch": 1.045553791186979,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.2351,
      "mean_token_accuracy": 0.16497762352228165,
      "num_tokens": 23088287.0,
      "step": 10535
    },
    {
      "entropy": 5.544708347320556,
      "epoch": 1.0460500198491465,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16559987291693687,
      "num_tokens": 23099252.0,
      "step": 10540
    },
    {
      "entropy": 5.518348550796508,
      "epoch": 1.046546248511314,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1663445234298706,
      "num_tokens": 23109752.0,
      "step": 10545
    },
    {
      "entropy": 5.5565505027771,
      "epoch": 1.0470424771734816,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.2294,
      "mean_token_accuracy": 0.16092514097690583,
      "num_tokens": 23120218.0,
      "step": 10550
    },
    {
      "entropy": 5.612714672088623,
      "epoch": 1.0475387058356491,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16738139986991882,
      "num_tokens": 23131156.0,
      "step": 10555
    },
    {
      "entropy": 5.577847003936768,
      "epoch": 1.0480349344978166,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.2478,
      "mean_token_accuracy": 0.16274990886449814,
      "num_tokens": 23142248.0,
      "step": 10560
    },
    {
      "entropy": 5.538260555267334,
      "epoch": 1.048531163159984,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1578839972615242,
      "num_tokens": 23153888.0,
      "step": 10565
    },
    {
      "entropy": 5.657569408416748,
      "epoch": 1.0490273918221515,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.2917,
      "mean_token_accuracy": 0.1679407238960266,
      "num_tokens": 23164411.0,
      "step": 10570
    },
    {
      "entropy": 5.534296894073487,
      "epoch": 1.0495236204843192,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.1696,
      "mean_token_accuracy": 0.16798045635223388,
      "num_tokens": 23176008.0,
      "step": 10575
    },
    {
      "entropy": 5.64479694366455,
      "epoch": 1.0500198491464867,
      "grad_norm": 0.875,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.4058,
      "mean_token_accuracy": 0.1526684433221817,
      "num_tokens": 23187163.0,
      "step": 10580
    },
    {
      "entropy": 5.705494165420532,
      "epoch": 1.0505160778086542,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.3062,
      "mean_token_accuracy": 0.15976687967777253,
      "num_tokens": 23197567.0,
      "step": 10585
    },
    {
      "entropy": 5.647280025482178,
      "epoch": 1.0510123064708217,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.263,
      "mean_token_accuracy": 0.16457771360874177,
      "num_tokens": 23208900.0,
      "step": 10590
    },
    {
      "entropy": 5.582605981826783,
      "epoch": 1.0515085351329894,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.266,
      "mean_token_accuracy": 0.16265039891004562,
      "num_tokens": 23220874.0,
      "step": 10595
    },
    {
      "entropy": 5.509483051300049,
      "epoch": 1.0520047637951568,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000489796052918233,
      "loss": 5.1672,
      "mean_token_accuracy": 0.16874666661024093,
      "num_tokens": 23232051.0,
      "step": 10600
    },
    {
      "entropy": 5.615007305145264,
      "epoch": 1.0525009924573243,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.2354,
      "mean_token_accuracy": 0.16529050916433335,
      "num_tokens": 23243159.0,
      "step": 10605
    },
    {
      "entropy": 5.564494323730469,
      "epoch": 1.0529972211194918,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.2257,
      "mean_token_accuracy": 0.16362677067518233,
      "num_tokens": 23255911.0,
      "step": 10610
    },
    {
      "entropy": 5.631772422790528,
      "epoch": 1.0534934497816595,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15403539985418319,
      "num_tokens": 23268773.0,
      "step": 10615
    },
    {
      "entropy": 5.532486438751221,
      "epoch": 1.053989678443827,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000489753813780361,
      "loss": 5.1701,
      "mean_token_accuracy": 0.17065977305173874,
      "num_tokens": 23278980.0,
      "step": 10620
    },
    {
      "entropy": 5.55983600616455,
      "epoch": 1.0544859071059944,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.2771,
      "mean_token_accuracy": 0.15551337450742722,
      "num_tokens": 23288941.0,
      "step": 10625
    },
    {
      "entropy": 5.570149660110474,
      "epoch": 1.054982135768162,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.1895,
      "mean_token_accuracy": 0.16804467290639877,
      "num_tokens": 23298941.0,
      "step": 10630
    },
    {
      "entropy": 5.552684879302978,
      "epoch": 1.0554783644303294,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.1941,
      "mean_token_accuracy": 0.16401765048503875,
      "num_tokens": 23310979.0,
      "step": 10635
    },
    {
      "entropy": 5.5387287616729735,
      "epoch": 1.055974593092497,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1693628802895546,
      "num_tokens": 23321755.0,
      "step": 10640
    },
    {
      "entropy": 5.479111242294311,
      "epoch": 1.0564708217546646,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.1598,
      "mean_token_accuracy": 0.1607544332742691,
      "num_tokens": 23331966.0,
      "step": 10645
    },
    {
      "entropy": 5.5815229415893555,
      "epoch": 1.056967050416832,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.1766,
      "mean_token_accuracy": 0.16756346374750136,
      "num_tokens": 23343866.0,
      "step": 10650
    },
    {
      "entropy": 5.545109367370605,
      "epoch": 1.0574632790789995,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.1289,
      "mean_token_accuracy": 0.16354062408208847,
      "num_tokens": 23354835.0,
      "step": 10655
    },
    {
      "entropy": 5.5583815574646,
      "epoch": 1.0579595077411672,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.1944,
      "mean_token_accuracy": 0.16945532709360123,
      "num_tokens": 23366446.0,
      "step": 10660
    },
    {
      "entropy": 5.54336371421814,
      "epoch": 1.0584557364033347,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.2514,
      "mean_token_accuracy": 0.16323245018720628,
      "num_tokens": 23377971.0,
      "step": 10665
    },
    {
      "entropy": 5.54066948890686,
      "epoch": 1.0589519650655022,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.1236,
      "mean_token_accuracy": 0.17196285277605056,
      "num_tokens": 23388238.0,
      "step": 10670
    },
    {
      "entropy": 5.567670822143555,
      "epoch": 1.0594481937276696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1647180825471878,
      "num_tokens": 23399319.0,
      "step": 10675
    },
    {
      "entropy": 5.53135576248169,
      "epoch": 1.0599444223898373,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.2737,
      "mean_token_accuracy": 0.15930084884166718,
      "num_tokens": 23410882.0,
      "step": 10680
    },
    {
      "entropy": 5.53029727935791,
      "epoch": 1.0604406510520048,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.2046,
      "mean_token_accuracy": 0.16700009256601334,
      "num_tokens": 23421775.0,
      "step": 10685
    },
    {
      "entropy": 5.594577074050903,
      "epoch": 1.0609368797141723,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.2154,
      "mean_token_accuracy": 0.1603095129132271,
      "num_tokens": 23432760.0,
      "step": 10690
    },
    {
      "entropy": 5.5524115562438965,
      "epoch": 1.0614331083763398,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.1462,
      "mean_token_accuracy": 0.16143215000629424,
      "num_tokens": 23442934.0,
      "step": 10695
    },
    {
      "entropy": 5.514578104019165,
      "epoch": 1.0619293370385074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16441091746091843,
      "num_tokens": 23453365.0,
      "step": 10700
    },
    {
      "entropy": 5.55297155380249,
      "epoch": 1.062425565700675,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.1712,
      "mean_token_accuracy": 0.16809804886579513,
      "num_tokens": 23464500.0,
      "step": 10705
    },
    {
      "entropy": 5.540005731582641,
      "epoch": 1.0629217943628424,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.2105,
      "mean_token_accuracy": 0.16483447104692459,
      "num_tokens": 23475702.0,
      "step": 10710
    },
    {
      "entropy": 5.520997190475464,
      "epoch": 1.0634180230250099,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.1962,
      "mean_token_accuracy": 0.16786203682422637,
      "num_tokens": 23487271.0,
      "step": 10715
    },
    {
      "entropy": 5.51843638420105,
      "epoch": 1.0639142516871773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16817058771848678,
      "num_tokens": 23497092.0,
      "step": 10720
    },
    {
      "entropy": 5.495817041397094,
      "epoch": 1.064410480349345,
      "grad_norm": 0.875,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.1246,
      "mean_token_accuracy": 0.16740285009145736,
      "num_tokens": 23507487.0,
      "step": 10725
    },
    {
      "entropy": 5.55347089767456,
      "epoch": 1.0649067090115125,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.1708,
      "mean_token_accuracy": 0.1704006090760231,
      "num_tokens": 23517571.0,
      "step": 10730
    },
    {
      "entropy": 5.6028519630432125,
      "epoch": 1.06540293767368,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.2525,
      "mean_token_accuracy": 0.16175172328948975,
      "num_tokens": 23527723.0,
      "step": 10735
    },
    {
      "entropy": 5.5679717540740965,
      "epoch": 1.0658991663358475,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.2581,
      "mean_token_accuracy": 0.1595466434955597,
      "num_tokens": 23539382.0,
      "step": 10740
    },
    {
      "entropy": 5.553967666625977,
      "epoch": 1.0663953949980152,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1686042383313179,
      "num_tokens": 23550011.0,
      "step": 10745
    },
    {
      "entropy": 5.535721778869629,
      "epoch": 1.0668916236601826,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.2043,
      "mean_token_accuracy": 0.17024645805358887,
      "num_tokens": 23560837.0,
      "step": 10750
    },
    {
      "entropy": 5.605242919921875,
      "epoch": 1.0673878523223501,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.2634,
      "mean_token_accuracy": 0.1615314170718193,
      "num_tokens": 23572083.0,
      "step": 10755
    },
    {
      "entropy": 5.625654697418213,
      "epoch": 1.0678840809845176,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.2536,
      "mean_token_accuracy": 0.1650639757514,
      "num_tokens": 23581799.0,
      "step": 10760
    },
    {
      "entropy": 5.584118747711182,
      "epoch": 1.0683803096466853,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.1925,
      "mean_token_accuracy": 0.16255976408720016,
      "num_tokens": 23592662.0,
      "step": 10765
    },
    {
      "entropy": 5.557419490814209,
      "epoch": 1.0688765383088528,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.2147,
      "mean_token_accuracy": 0.15909711867570878,
      "num_tokens": 23604040.0,
      "step": 10770
    },
    {
      "entropy": 5.496958303451538,
      "epoch": 1.0693727669710202,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.1322,
      "mean_token_accuracy": 0.1683470755815506,
      "num_tokens": 23616182.0,
      "step": 10775
    },
    {
      "entropy": 5.637787389755249,
      "epoch": 1.0698689956331877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.3611,
      "mean_token_accuracy": 0.1534825459122658,
      "num_tokens": 23628184.0,
      "step": 10780
    },
    {
      "entropy": 5.548614263534546,
      "epoch": 1.0703652242953554,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.1681,
      "mean_token_accuracy": 0.17026416957378387,
      "num_tokens": 23638701.0,
      "step": 10785
    },
    {
      "entropy": 5.558590412139893,
      "epoch": 1.0708614529575229,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.2228,
      "mean_token_accuracy": 0.16570426374673844,
      "num_tokens": 23649020.0,
      "step": 10790
    },
    {
      "entropy": 5.5540038585662845,
      "epoch": 1.0713576816196904,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.1682,
      "mean_token_accuracy": 0.17293568029999734,
      "num_tokens": 23660610.0,
      "step": 10795
    },
    {
      "entropy": 5.518844747543335,
      "epoch": 1.0718539102818578,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.168,
      "mean_token_accuracy": 0.16659754365682602,
      "num_tokens": 23671684.0,
      "step": 10800
    },
    {
      "entropy": 5.500228309631348,
      "epoch": 1.0723501389440253,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1672847479581833,
      "num_tokens": 23683577.0,
      "step": 10805
    },
    {
      "entropy": 5.5489821434021,
      "epoch": 1.072846367606193,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1647532105445862,
      "num_tokens": 23695335.0,
      "step": 10810
    },
    {
      "entropy": 5.6645975589752195,
      "epoch": 1.0733425962683605,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.3285,
      "mean_token_accuracy": 0.1543939381837845,
      "num_tokens": 23708017.0,
      "step": 10815
    },
    {
      "entropy": 5.5547504901885985,
      "epoch": 1.073838824930528,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.2019,
      "mean_token_accuracy": 0.16092967689037324,
      "num_tokens": 23719735.0,
      "step": 10820
    },
    {
      "entropy": 5.509702157974243,
      "epoch": 1.0743350535926954,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000489315952904175,
      "loss": 5.1246,
      "mean_token_accuracy": 0.1748264566063881,
      "num_tokens": 23730731.0,
      "step": 10825
    },
    {
      "entropy": 5.586430358886719,
      "epoch": 1.0748312822548631,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.3401,
      "mean_token_accuracy": 0.1619722381234169,
      "num_tokens": 23742588.0,
      "step": 10830
    },
    {
      "entropy": 5.547558212280274,
      "epoch": 1.0753275109170306,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.188,
      "mean_token_accuracy": 0.16561543941497803,
      "num_tokens": 23752965.0,
      "step": 10835
    },
    {
      "entropy": 5.49914517402649,
      "epoch": 1.075823739579198,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.2026,
      "mean_token_accuracy": 0.16218311339616776,
      "num_tokens": 23764756.0,
      "step": 10840
    },
    {
      "entropy": 5.455108118057251,
      "epoch": 1.0763199682413656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.0281,
      "mean_token_accuracy": 0.18681704252958298,
      "num_tokens": 23775595.0,
      "step": 10845
    },
    {
      "entropy": 5.54834771156311,
      "epoch": 1.0768161969035333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1651290997862816,
      "num_tokens": 23786721.0,
      "step": 10850
    },
    {
      "entropy": 5.5173118114471436,
      "epoch": 1.0773124255657007,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.1493,
      "mean_token_accuracy": 0.16126493513584136,
      "num_tokens": 23797440.0,
      "step": 10855
    },
    {
      "entropy": 5.538034152984619,
      "epoch": 1.0778086542278682,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.1985,
      "mean_token_accuracy": 0.159446020424366,
      "num_tokens": 23807807.0,
      "step": 10860
    },
    {
      "entropy": 5.625035524368286,
      "epoch": 1.0783048828900357,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048922947440543,
      "loss": 5.296,
      "mean_token_accuracy": 0.1559945300221443,
      "num_tokens": 23818558.0,
      "step": 10865
    },
    {
      "entropy": 5.585907983779907,
      "epoch": 1.0788011115522032,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.2752,
      "mean_token_accuracy": 0.16363571137189864,
      "num_tokens": 23829228.0,
      "step": 10870
    },
    {
      "entropy": 5.554886341094971,
      "epoch": 1.0792973402143708,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.1465,
      "mean_token_accuracy": 0.1729414477944374,
      "num_tokens": 23840161.0,
      "step": 10875
    },
    {
      "entropy": 5.480851364135742,
      "epoch": 1.0797935688765383,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000489196957305939,
      "loss": 5.094,
      "mean_token_accuracy": 0.16960520148277283,
      "num_tokens": 23850516.0,
      "step": 10880
    },
    {
      "entropy": 5.507450532913208,
      "epoch": 1.0802897975387058,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1664951428771019,
      "num_tokens": 23860589.0,
      "step": 10885
    },
    {
      "entropy": 5.493975734710693,
      "epoch": 1.0807860262008733,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.1828,
      "mean_token_accuracy": 0.16754122078418732,
      "num_tokens": 23871363.0,
      "step": 10890
    },
    {
      "entropy": 5.494058895111084,
      "epoch": 1.081282254863041,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000489164392401842,
      "loss": 5.2363,
      "mean_token_accuracy": 0.16587782502174378,
      "num_tokens": 23882297.0,
      "step": 10895
    },
    {
      "entropy": 5.512590026855468,
      "epoch": 1.0817784835252084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1687490910291672,
      "num_tokens": 23892966.0,
      "step": 10900
    },
    {
      "entropy": 5.557666778564453,
      "epoch": 1.082274712187376,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.1599,
      "mean_token_accuracy": 0.17232990860939026,
      "num_tokens": 23902918.0,
      "step": 10905
    },
    {
      "entropy": 5.637684345245361,
      "epoch": 1.0827709408495434,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1570495694875717,
      "num_tokens": 23913692.0,
      "step": 10910
    },
    {
      "entropy": 5.607086515426635,
      "epoch": 1.083267169511711,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.2782,
      "mean_token_accuracy": 0.16719367057085038,
      "num_tokens": 23923714.0,
      "step": 10915
    },
    {
      "entropy": 5.473653793334961,
      "epoch": 1.0837633981738786,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.1251,
      "mean_token_accuracy": 0.17011374682188035,
      "num_tokens": 23934398.0,
      "step": 10920
    },
    {
      "entropy": 5.52828426361084,
      "epoch": 1.084259626836046,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.2778,
      "mean_token_accuracy": 0.16212295442819596,
      "num_tokens": 23946181.0,
      "step": 10925
    },
    {
      "entropy": 5.574047517776489,
      "epoch": 1.0847558554982135,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.214,
      "mean_token_accuracy": 0.16366038769483565,
      "num_tokens": 23959248.0,
      "step": 10930
    },
    {
      "entropy": 5.544989824295044,
      "epoch": 1.085252084160381,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.1858,
      "mean_token_accuracy": 0.16458184644579887,
      "num_tokens": 23972676.0,
      "step": 10935
    },
    {
      "entropy": 5.5969038009643555,
      "epoch": 1.0857483128225487,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000489066410934646,
      "loss": 5.2817,
      "mean_token_accuracy": 0.1623589038848877,
      "num_tokens": 23983720.0,
      "step": 10940
    },
    {
      "entropy": 5.566885566711425,
      "epoch": 1.0862445414847162,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.2226,
      "mean_token_accuracy": 0.16785094290971755,
      "num_tokens": 23994730.0,
      "step": 10945
    },
    {
      "entropy": 5.591116952896118,
      "epoch": 1.0867407701468836,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.2622,
      "mean_token_accuracy": 0.15998478680849076,
      "num_tokens": 24006022.0,
      "step": 10950
    },
    {
      "entropy": 5.594762182235717,
      "epoch": 1.0872369988090511,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.2915,
      "mean_token_accuracy": 0.16708393692970275,
      "num_tokens": 24016670.0,
      "step": 10955
    },
    {
      "entropy": 5.583100509643555,
      "epoch": 1.0877332274712188,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.2155,
      "mean_token_accuracy": 0.1567889481782913,
      "num_tokens": 24027677.0,
      "step": 10960
    },
    {
      "entropy": 5.534997606277466,
      "epoch": 1.0882294561333863,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.2178,
      "mean_token_accuracy": 0.17140555381774902,
      "num_tokens": 24039191.0,
      "step": 10965
    },
    {
      "entropy": 5.485724639892578,
      "epoch": 1.0887256847955538,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.1081,
      "mean_token_accuracy": 0.1698074996471405,
      "num_tokens": 24049282.0,
      "step": 10970
    },
    {
      "entropy": 5.51558256149292,
      "epoch": 1.0892219134577212,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.2122,
      "mean_token_accuracy": 0.1717247873544693,
      "num_tokens": 24059506.0,
      "step": 10975
    },
    {
      "entropy": 5.4958977699279785,
      "epoch": 1.089718142119889,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.1825,
      "mean_token_accuracy": 0.17010412216186524,
      "num_tokens": 24068973.0,
      "step": 10980
    },
    {
      "entropy": 5.615663242340088,
      "epoch": 1.0902143707820564,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488967999488278,
      "loss": 5.2873,
      "mean_token_accuracy": 0.16835326552391053,
      "num_tokens": 24081012.0,
      "step": 10985
    },
    {
      "entropy": 5.560965824127197,
      "epoch": 1.0907105994442239,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048895703834978,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16774856597185134,
      "num_tokens": 24090540.0,
      "step": 10990
    },
    {
      "entropy": 5.549176788330078,
      "epoch": 1.0912068281063914,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.2078,
      "mean_token_accuracy": 0.16891217231750488,
      "num_tokens": 24101704.0,
      "step": 10995
    },
    {
      "entropy": 5.561347007751465,
      "epoch": 1.091703056768559,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.27,
      "mean_token_accuracy": 0.15854610353708268,
      "num_tokens": 24110685.0,
      "step": 11000
    },
    {
      "entropy": 5.538417196273803,
      "epoch": 1.0921992854307265,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.1702,
      "mean_token_accuracy": 0.16221912652254106,
      "num_tokens": 24121962.0,
      "step": 11005
    },
    {
      "entropy": 5.515323495864868,
      "epoch": 1.092695514092894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000488913140741773,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1734781116247177,
      "num_tokens": 24132279.0,
      "step": 11010
    },
    {
      "entropy": 5.516565656661987,
      "epoch": 1.0931917427550615,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1703609362244606,
      "num_tokens": 24144149.0,
      "step": 11015
    },
    {
      "entropy": 5.614330625534057,
      "epoch": 1.093687971417229,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.2268,
      "mean_token_accuracy": 0.16023587584495544,
      "num_tokens": 24155917.0,
      "step": 11020
    },
    {
      "entropy": 5.553860187530518,
      "epoch": 1.0941842000793967,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.2013,
      "mean_token_accuracy": 0.16211372166872023,
      "num_tokens": 24166214.0,
      "step": 11025
    },
    {
      "entropy": 5.550048017501831,
      "epoch": 1.0946804287415641,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.2538,
      "mean_token_accuracy": 0.16079729944467544,
      "num_tokens": 24177040.0,
      "step": 11030
    },
    {
      "entropy": 5.5392168998718265,
      "epoch": 1.0951766574037316,
      "grad_norm": 1.0,
      "learning_rate": 0.000488858149380651,
      "loss": 5.184,
      "mean_token_accuracy": 0.16641712337732315,
      "num_tokens": 24188130.0,
      "step": 11035
    },
    {
      "entropy": 5.628276824951172,
      "epoch": 1.095672886065899,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000488847135197671,
      "loss": 5.2957,
      "mean_token_accuracy": 0.16604692935943605,
      "num_tokens": 24197393.0,
      "step": 11040
    },
    {
      "entropy": 5.525161361694336,
      "epoch": 1.0961691147280668,
      "grad_norm": 0.875,
      "learning_rate": 0.000488836115711743,
      "loss": 5.2189,
      "mean_token_accuracy": 0.1677385613322258,
      "num_tokens": 24208078.0,
      "step": 11045
    },
    {
      "entropy": 5.606669235229492,
      "epoch": 1.0966653433902342,
      "grad_norm": 0.875,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.2542,
      "mean_token_accuracy": 0.16372596919536592,
      "num_tokens": 24219188.0,
      "step": 11050
    },
    {
      "entropy": 5.588831377029419,
      "epoch": 1.0971615720524017,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.2342,
      "mean_token_accuracy": 0.1582639530301094,
      "num_tokens": 24229315.0,
      "step": 11055
    },
    {
      "entropy": 5.5890861511230465,
      "epoch": 1.0976578007145692,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.3094,
      "mean_token_accuracy": 0.16223085671663284,
      "num_tokens": 24240580.0,
      "step": 11060
    },
    {
      "entropy": 5.608312034606934,
      "epoch": 1.098154029376737,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.255,
      "mean_token_accuracy": 0.16820312291383743,
      "num_tokens": 24252048.0,
      "step": 11065
    },
    {
      "entropy": 5.545646810531617,
      "epoch": 1.0986502580389044,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.1842,
      "mean_token_accuracy": 0.16468918025493623,
      "num_tokens": 24262479.0,
      "step": 11070
    },
    {
      "entropy": 5.504002046585083,
      "epoch": 1.0991464867010718,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16752441078424454,
      "num_tokens": 24272790.0,
      "step": 11075
    },
    {
      "entropy": 5.593094539642334,
      "epoch": 1.0996427153632393,
      "grad_norm": 0.90625,
      "learning_rate": 0.000488758830850669,
      "loss": 5.2254,
      "mean_token_accuracy": 0.16761568933725357,
      "num_tokens": 24284150.0,
      "step": 11080
    },
    {
      "entropy": 5.59558596611023,
      "epoch": 1.100138944025407,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16363905668258666,
      "num_tokens": 24295683.0,
      "step": 11085
    },
    {
      "entropy": 5.500878763198853,
      "epoch": 1.1006351726875745,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.1738,
      "mean_token_accuracy": 0.17295398265123368,
      "num_tokens": 24306087.0,
      "step": 11090
    },
    {
      "entropy": 5.530308103561401,
      "epoch": 1.101131401349742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.2081,
      "mean_token_accuracy": 0.1613711014389992,
      "num_tokens": 24316677.0,
      "step": 11095
    },
    {
      "entropy": 5.557991409301758,
      "epoch": 1.1016276300119094,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488714551450203,
      "loss": 5.1343,
      "mean_token_accuracy": 0.16808457672595978,
      "num_tokens": 24327445.0,
      "step": 11100
    },
    {
      "entropy": 5.587096214294434,
      "epoch": 1.102123858674077,
      "grad_norm": 0.8125,
      "learning_rate": 0.000488703468350249,
      "loss": 5.258,
      "mean_token_accuracy": 0.1658082753419876,
      "num_tokens": 24340231.0,
      "step": 11105
    },
    {
      "entropy": 5.623089551925659,
      "epoch": 1.1026200873362446,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.2784,
      "mean_token_accuracy": 0.15635586082935332,
      "num_tokens": 24352443.0,
      "step": 11110
    },
    {
      "entropy": 5.547194194793701,
      "epoch": 1.103116315998412,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.1763,
      "mean_token_accuracy": 0.16005021929740906,
      "num_tokens": 24364485.0,
      "step": 11115
    },
    {
      "entropy": 5.5828930854797365,
      "epoch": 1.1036125446605796,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.2738,
      "mean_token_accuracy": 0.16197794675827026,
      "num_tokens": 24375943.0,
      "step": 11120
    },
    {
      "entropy": 5.614772605895996,
      "epoch": 1.104108773322747,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.3219,
      "mean_token_accuracy": 0.15372844338417052,
      "num_tokens": 24387703.0,
      "step": 11125
    },
    {
      "entropy": 5.566215848922729,
      "epoch": 1.1046050019849147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.1759,
      "mean_token_accuracy": 0.1636826992034912,
      "num_tokens": 24398888.0,
      "step": 11130
    },
    {
      "entropy": 5.565417194366455,
      "epoch": 1.1051012306470822,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.1397,
      "mean_token_accuracy": 0.1679879069328308,
      "num_tokens": 24409643.0,
      "step": 11135
    },
    {
      "entropy": 5.509988737106323,
      "epoch": 1.1055974593092497,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.2231,
      "mean_token_accuracy": 0.1663526251912117,
      "num_tokens": 24421433.0,
      "step": 11140
    },
    {
      "entropy": 5.554637908935547,
      "epoch": 1.1060936879714172,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000488614612795852,
      "loss": 5.2161,
      "mean_token_accuracy": 0.15888823866844176,
      "num_tokens": 24433649.0,
      "step": 11145
    },
    {
      "entropy": 5.54486927986145,
      "epoch": 1.1065899166335849,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.193,
      "mean_token_accuracy": 0.16530629843473435,
      "num_tokens": 24445216.0,
      "step": 11150
    },
    {
      "entropy": 5.54317307472229,
      "epoch": 1.1070861452957523,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.2441,
      "mean_token_accuracy": 0.16257893517613412,
      "num_tokens": 24456026.0,
      "step": 11155
    },
    {
      "entropy": 5.513910436630249,
      "epoch": 1.1075823739579198,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1670771911740303,
      "num_tokens": 24466623.0,
      "step": 11160
    },
    {
      "entropy": 5.548501682281494,
      "epoch": 1.1080786026200873,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000488570057877465,
      "loss": 5.1577,
      "mean_token_accuracy": 0.16285916566848754,
      "num_tokens": 24476711.0,
      "step": 11165
    },
    {
      "entropy": 5.514219903945923,
      "epoch": 1.108574831282255,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.14,
      "mean_token_accuracy": 0.1683489754796028,
      "num_tokens": 24487987.0,
      "step": 11170
    },
    {
      "entropy": 5.533908748626709,
      "epoch": 1.1090710599444225,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.1291,
      "mean_token_accuracy": 0.17121703922748566,
      "num_tokens": 24498482.0,
      "step": 11175
    },
    {
      "entropy": 5.5430937767028805,
      "epoch": 1.10956728860659,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.2992,
      "mean_token_accuracy": 0.16051233559846878,
      "num_tokens": 24509206.0,
      "step": 11180
    },
    {
      "entropy": 5.544673109054566,
      "epoch": 1.1100635172687574,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1617652952671051,
      "num_tokens": 24519534.0,
      "step": 11185
    },
    {
      "entropy": 5.559151458740234,
      "epoch": 1.1105597459309249,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.223,
      "mean_token_accuracy": 0.15999924540519714,
      "num_tokens": 24531013.0,
      "step": 11190
    },
    {
      "entropy": 5.524998807907105,
      "epoch": 1.1110559745930926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.2166,
      "mean_token_accuracy": 0.17269886732101442,
      "num_tokens": 24540865.0,
      "step": 11195
    },
    {
      "entropy": 5.583633756637573,
      "epoch": 1.11155220325526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1682659462094307,
      "num_tokens": 24550262.0,
      "step": 11200
    },
    {
      "entropy": 5.578919792175293,
      "epoch": 1.1120484319174275,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488480693846703,
      "loss": 5.2428,
      "mean_token_accuracy": 0.15577299892902374,
      "num_tokens": 24561933.0,
      "step": 11205
    },
    {
      "entropy": 5.543815326690674,
      "epoch": 1.112544660579595,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.2097,
      "mean_token_accuracy": 0.1660779595375061,
      "num_tokens": 24572698.0,
      "step": 11210
    },
    {
      "entropy": 5.513328409194946,
      "epoch": 1.1130408892417627,
      "grad_norm": 0.875,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.1417,
      "mean_token_accuracy": 0.17206312268972396,
      "num_tokens": 24581900.0,
      "step": 11215
    },
    {
      "entropy": 5.583297491073608,
      "epoch": 1.1135371179039302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.2135,
      "mean_token_accuracy": 0.16887544989585876,
      "num_tokens": 24592163.0,
      "step": 11220
    },
    {
      "entropy": 5.552220582962036,
      "epoch": 1.1140333465660976,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.1129,
      "mean_token_accuracy": 0.16435604095458983,
      "num_tokens": 24602544.0,
      "step": 11225
    },
    {
      "entropy": 5.569680404663086,
      "epoch": 1.1145295752282651,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1626817688345909,
      "num_tokens": 24613204.0,
      "step": 11230
    },
    {
      "entropy": 5.583879804611206,
      "epoch": 1.1150258038904326,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.2594,
      "mean_token_accuracy": 0.15875246971845627,
      "num_tokens": 24624045.0,
      "step": 11235
    },
    {
      "entropy": 5.634034204483032,
      "epoch": 1.1155220325526003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.2843,
      "mean_token_accuracy": 0.157689693570137,
      "num_tokens": 24635629.0,
      "step": 11240
    },
    {
      "entropy": 5.599589538574219,
      "epoch": 1.1160182612147678,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.274,
      "mean_token_accuracy": 0.16046741008758544,
      "num_tokens": 24646705.0,
      "step": 11245
    },
    {
      "entropy": 5.456070518493652,
      "epoch": 1.1165144898769352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.073,
      "mean_token_accuracy": 0.17442625612020493,
      "num_tokens": 24655969.0,
      "step": 11250
    },
    {
      "entropy": 5.485085344314575,
      "epoch": 1.1170107185391027,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.1233,
      "mean_token_accuracy": 0.16948835104703902,
      "num_tokens": 24666739.0,
      "step": 11255
    },
    {
      "entropy": 5.585072374343872,
      "epoch": 1.1175069472012704,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.2578,
      "mean_token_accuracy": 0.16412586718797684,
      "num_tokens": 24677775.0,
      "step": 11260
    },
    {
      "entropy": 5.551873588562012,
      "epoch": 1.118003175863438,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.161,
      "mean_token_accuracy": 0.16682316213846207,
      "num_tokens": 24688781.0,
      "step": 11265
    },
    {
      "entropy": 5.5695586681365965,
      "epoch": 1.1184994045256054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.1761,
      "mean_token_accuracy": 0.1654194638133049,
      "num_tokens": 24698979.0,
      "step": 11270
    },
    {
      "entropy": 5.574695634841919,
      "epoch": 1.1189956331877728,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488323491623972,
      "loss": 5.1775,
      "mean_token_accuracy": 0.16536389291286469,
      "num_tokens": 24709419.0,
      "step": 11275
    },
    {
      "entropy": 5.5357836246490475,
      "epoch": 1.1194918618499405,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.2566,
      "mean_token_accuracy": 0.16212385445833205,
      "num_tokens": 24720495.0,
      "step": 11280
    },
    {
      "entropy": 5.560220003128052,
      "epoch": 1.119988090512108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.1537,
      "mean_token_accuracy": 0.16680119782686234,
      "num_tokens": 24730714.0,
      "step": 11285
    },
    {
      "entropy": 5.539450311660767,
      "epoch": 1.1204843191742755,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.2031,
      "mean_token_accuracy": 0.16852664798498154,
      "num_tokens": 24741953.0,
      "step": 11290
    },
    {
      "entropy": 5.5231469631195065,
      "epoch": 1.120980547836443,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.2153,
      "mean_token_accuracy": 0.16592925041913986,
      "num_tokens": 24753090.0,
      "step": 11295
    },
    {
      "entropy": 5.5818626403808596,
      "epoch": 1.1214767764986107,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000488267096660026,
      "loss": 5.1883,
      "mean_token_accuracy": 0.16323001831769943,
      "num_tokens": 24763738.0,
      "step": 11300
    },
    {
      "entropy": 5.553966903686524,
      "epoch": 1.1219730051607781,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.2013,
      "mean_token_accuracy": 0.16106340736150743,
      "num_tokens": 24774557.0,
      "step": 11305
    },
    {
      "entropy": 5.53676438331604,
      "epoch": 1.1224692338229456,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.2361,
      "mean_token_accuracy": 0.16456136107444763,
      "num_tokens": 24784815.0,
      "step": 11310
    },
    {
      "entropy": 5.604258823394775,
      "epoch": 1.122965462485113,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.2369,
      "mean_token_accuracy": 0.16451041251420975,
      "num_tokens": 24796149.0,
      "step": 11315
    },
    {
      "entropy": 5.537981748580933,
      "epoch": 1.1234616911472806,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.2746,
      "mean_token_accuracy": 0.16203459203243256,
      "num_tokens": 24806562.0,
      "step": 11320
    },
    {
      "entropy": 5.527090358734131,
      "epoch": 1.1239579198094483,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.1704,
      "mean_token_accuracy": 0.16578106731176376,
      "num_tokens": 24817569.0,
      "step": 11325
    },
    {
      "entropy": 5.54568076133728,
      "epoch": 1.1244541484716157,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.1918,
      "mean_token_accuracy": 0.165217849612236,
      "num_tokens": 24828759.0,
      "step": 11330
    },
    {
      "entropy": 5.5325864315032955,
      "epoch": 1.1249503771337832,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.1664,
      "mean_token_accuracy": 0.16603039652109147,
      "num_tokens": 24840891.0,
      "step": 11335
    },
    {
      "entropy": 5.526506805419922,
      "epoch": 1.1254466057959507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.1721,
      "mean_token_accuracy": 0.17060137540102005,
      "num_tokens": 24851740.0,
      "step": 11340
    },
    {
      "entropy": 5.590481996536255,
      "epoch": 1.1259428344581184,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.1924,
      "mean_token_accuracy": 0.16490679532289504,
      "num_tokens": 24862484.0,
      "step": 11345
    },
    {
      "entropy": 5.613970327377319,
      "epoch": 1.1264390631202859,
      "grad_norm": 0.921875,
      "learning_rate": 0.000488153910124864,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1659662216901779,
      "num_tokens": 24872897.0,
      "step": 11350
    },
    {
      "entropy": 5.656288576126099,
      "epoch": 1.1269352917824533,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000488142562395044,
      "loss": 5.3248,
      "mean_token_accuracy": 0.16699188500642775,
      "num_tokens": 24884318.0,
      "step": 11355
    },
    {
      "entropy": 5.5151939868927,
      "epoch": 1.1274315204446208,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16777215749025345,
      "num_tokens": 24894284.0,
      "step": 11360
    },
    {
      "entropy": 5.548619937896729,
      "epoch": 1.1279277491067885,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.2201,
      "mean_token_accuracy": 0.1656787857413292,
      "num_tokens": 24904971.0,
      "step": 11365
    },
    {
      "entropy": 5.507843828201294,
      "epoch": 1.128423977768956,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.1306,
      "mean_token_accuracy": 0.17277810871601104,
      "num_tokens": 24915650.0,
      "step": 11370
    },
    {
      "entropy": 5.538347578048706,
      "epoch": 1.1289202064311235,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.208,
      "mean_token_accuracy": 0.17010795772075654,
      "num_tokens": 24925430.0,
      "step": 11375
    },
    {
      "entropy": 5.555023908615112,
      "epoch": 1.129416435093291,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1685572847723961,
      "num_tokens": 24935568.0,
      "step": 11380
    },
    {
      "entropy": 5.513298416137696,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.2426,
      "mean_token_accuracy": 0.15762319415807724,
      "num_tokens": 24945761.0,
      "step": 11385
    },
    {
      "entropy": 5.621783113479614,
      "epoch": 1.130408892417626,
      "grad_norm": 0.875,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.3115,
      "mean_token_accuracy": 0.15445470213890075,
      "num_tokens": 24958259.0,
      "step": 11390
    },
    {
      "entropy": 5.5246954441070555,
      "epoch": 1.1309051210797936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17227267771959304,
      "num_tokens": 24967697.0,
      "step": 11395
    },
    {
      "entropy": 5.535379695892334,
      "epoch": 1.131401349741961,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.21,
      "mean_token_accuracy": 0.16070815771818162,
      "num_tokens": 24979010.0,
      "step": 11400
    },
    {
      "entropy": 5.479898595809937,
      "epoch": 1.1318975784041285,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17295579463243485,
      "num_tokens": 24989700.0,
      "step": 11405
    },
    {
      "entropy": 5.465715312957764,
      "epoch": 1.1323938070662962,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17131666094064713,
      "num_tokens": 25000851.0,
      "step": 11410
    },
    {
      "entropy": 5.521159219741821,
      "epoch": 1.1328900357284637,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.1582,
      "mean_token_accuracy": 0.1659923404455185,
      "num_tokens": 25010336.0,
      "step": 11415
    },
    {
      "entropy": 5.545714616775513,
      "epoch": 1.1333862643906312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.1469,
      "mean_token_accuracy": 0.17251309156417846,
      "num_tokens": 25019711.0,
      "step": 11420
    },
    {
      "entropy": 5.52295708656311,
      "epoch": 1.1338824930527986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.1583,
      "mean_token_accuracy": 0.16516998410224915,
      "num_tokens": 25030349.0,
      "step": 11425
    },
    {
      "entropy": 5.477740287780762,
      "epoch": 1.1343787217149663,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1775646448135376,
      "num_tokens": 25041356.0,
      "step": 11430
    },
    {
      "entropy": 5.500832271575928,
      "epoch": 1.1348749503771338,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000487960280084905,
      "loss": 5.1632,
      "mean_token_accuracy": 0.1694568246603012,
      "num_tokens": 25051609.0,
      "step": 11435
    },
    {
      "entropy": 5.549370384216308,
      "epoch": 1.1353711790393013,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.2007,
      "mean_token_accuracy": 0.1605103000998497,
      "num_tokens": 25061813.0,
      "step": 11440
    },
    {
      "entropy": 5.5893495082855225,
      "epoch": 1.1358674077014688,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.1939,
      "mean_token_accuracy": 0.15800101011991502,
      "num_tokens": 25072750.0,
      "step": 11445
    },
    {
      "entropy": 5.613778018951416,
      "epoch": 1.1363636363636362,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.2843,
      "mean_token_accuracy": 0.16158646047115327,
      "num_tokens": 25083735.0,
      "step": 11450
    },
    {
      "entropy": 5.586733293533325,
      "epoch": 1.136859865025804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.2448,
      "mean_token_accuracy": 0.1650795504450798,
      "num_tokens": 25094037.0,
      "step": 11455
    },
    {
      "entropy": 5.510610580444336,
      "epoch": 1.1373560936879714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.1278,
      "mean_token_accuracy": 0.16976490169763564,
      "num_tokens": 25103497.0,
      "step": 11460
    },
    {
      "entropy": 5.529094886779785,
      "epoch": 1.137852322350139,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.1609,
      "mean_token_accuracy": 0.16677828282117843,
      "num_tokens": 25114322.0,
      "step": 11465
    },
    {
      "entropy": 5.535622024536133,
      "epoch": 1.1383485510123066,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16457552462816238,
      "num_tokens": 25125355.0,
      "step": 11470
    },
    {
      "entropy": 5.655879783630371,
      "epoch": 1.138844779674474,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.3324,
      "mean_token_accuracy": 0.153234301507473,
      "num_tokens": 25137264.0,
      "step": 11475
    },
    {
      "entropy": 5.547727584838867,
      "epoch": 1.1393410083366415,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.2046,
      "mean_token_accuracy": 0.16050828844308854,
      "num_tokens": 25148294.0,
      "step": 11480
    },
    {
      "entropy": 5.485212326049805,
      "epoch": 1.139837236998809,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.1785,
      "mean_token_accuracy": 0.17027826607227325,
      "num_tokens": 25159123.0,
      "step": 11485
    },
    {
      "entropy": 5.62592339515686,
      "epoch": 1.1403334656609765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.2848,
      "mean_token_accuracy": 0.16033299565315245,
      "num_tokens": 25168910.0,
      "step": 11490
    },
    {
      "entropy": 5.503257131576538,
      "epoch": 1.1408296943231442,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.0954,
      "mean_token_accuracy": 0.16735006421804427,
      "num_tokens": 25180433.0,
      "step": 11495
    },
    {
      "entropy": 5.544435119628906,
      "epoch": 1.1413259229853117,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.133,
      "mean_token_accuracy": 0.1690497800707817,
      "num_tokens": 25191558.0,
      "step": 11500
    },
    {
      "entropy": 5.460061645507812,
      "epoch": 1.1418221516474791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1632181152701378,
      "num_tokens": 25201241.0,
      "step": 11505
    },
    {
      "entropy": 5.526847505569458,
      "epoch": 1.1423183803096466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1676988646388054,
      "num_tokens": 25212305.0,
      "step": 11510
    },
    {
      "entropy": 5.501944923400879,
      "epoch": 1.1428146089718143,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.1693,
      "mean_token_accuracy": 0.16777824610471725,
      "num_tokens": 25224296.0,
      "step": 11515
    },
    {
      "entropy": 5.551266193389893,
      "epoch": 1.1433108376339818,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.1899,
      "mean_token_accuracy": 0.16450172215700148,
      "num_tokens": 25235730.0,
      "step": 11520
    },
    {
      "entropy": 5.5914887428283695,
      "epoch": 1.1438070662961493,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.2442,
      "mean_token_accuracy": 0.15847379714250565,
      "num_tokens": 25247066.0,
      "step": 11525
    },
    {
      "entropy": 5.613976716995239,
      "epoch": 1.1443032949583167,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.274,
      "mean_token_accuracy": 0.15797610580921173,
      "num_tokens": 25258417.0,
      "step": 11530
    },
    {
      "entropy": 5.572810792922974,
      "epoch": 1.1447995236204842,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.1908,
      "mean_token_accuracy": 0.163433738052845,
      "num_tokens": 25268268.0,
      "step": 11535
    },
    {
      "entropy": 5.5676380634307865,
      "epoch": 1.145295752282652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1695170134305954,
      "num_tokens": 25278437.0,
      "step": 11540
    },
    {
      "entropy": 5.610391855239868,
      "epoch": 1.1457919809448194,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.1929,
      "mean_token_accuracy": 0.16236614286899567,
      "num_tokens": 25290515.0,
      "step": 11545
    },
    {
      "entropy": 5.546370315551758,
      "epoch": 1.1462882096069869,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.1983,
      "mean_token_accuracy": 0.16022950261831284,
      "num_tokens": 25301990.0,
      "step": 11550
    },
    {
      "entropy": 5.525651693344116,
      "epoch": 1.1467844382691545,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16659033447504043,
      "num_tokens": 25311917.0,
      "step": 11555
    },
    {
      "entropy": 5.578454780578613,
      "epoch": 1.147280666931322,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.3009,
      "mean_token_accuracy": 0.16948064267635346,
      "num_tokens": 25322578.0,
      "step": 11560
    },
    {
      "entropy": 5.6128521919250485,
      "epoch": 1.1477768955934895,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17350199371576308,
      "num_tokens": 25332068.0,
      "step": 11565
    },
    {
      "entropy": 5.516817998886109,
      "epoch": 1.148273124255657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.1991,
      "mean_token_accuracy": 0.1615613095462322,
      "num_tokens": 25342511.0,
      "step": 11570
    },
    {
      "entropy": 5.458237409591675,
      "epoch": 1.1487693529178244,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.145,
      "mean_token_accuracy": 0.16616877615451814,
      "num_tokens": 25353602.0,
      "step": 11575
    },
    {
      "entropy": 5.52956600189209,
      "epoch": 1.1492655815799921,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.1439,
      "mean_token_accuracy": 0.16954532712697984,
      "num_tokens": 25364570.0,
      "step": 11580
    },
    {
      "entropy": 5.510851669311523,
      "epoch": 1.1497618102421596,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.1591,
      "mean_token_accuracy": 0.16627704352140427,
      "num_tokens": 25375280.0,
      "step": 11585
    },
    {
      "entropy": 5.592412900924683,
      "epoch": 1.150258038904327,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.2825,
      "mean_token_accuracy": 0.16374307572841645,
      "num_tokens": 25386786.0,
      "step": 11590
    },
    {
      "entropy": 5.495249938964844,
      "epoch": 1.1507542675664946,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.1694,
      "mean_token_accuracy": 0.16802662163972854,
      "num_tokens": 25397867.0,
      "step": 11595
    },
    {
      "entropy": 5.610469913482666,
      "epoch": 1.1512504962286623,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000487580054473603,
      "loss": 5.2741,
      "mean_token_accuracy": 0.1630384221673012,
      "num_tokens": 25408394.0,
      "step": 11600
    },
    {
      "entropy": 5.659369897842407,
      "epoch": 1.1517467248908297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.2877,
      "mean_token_accuracy": 0.15826055258512498,
      "num_tokens": 25418815.0,
      "step": 11605
    },
    {
      "entropy": 5.567052698135376,
      "epoch": 1.1522429535529972,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.2179,
      "mean_token_accuracy": 0.16351694166660308,
      "num_tokens": 25430262.0,
      "step": 11610
    },
    {
      "entropy": 5.564117050170898,
      "epoch": 1.1527391822151647,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.2633,
      "mean_token_accuracy": 0.15664817839860917,
      "num_tokens": 25441082.0,
      "step": 11615
    },
    {
      "entropy": 5.557660007476807,
      "epoch": 1.1532354108773322,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.2345,
      "mean_token_accuracy": 0.16344863474369048,
      "num_tokens": 25452096.0,
      "step": 11620
    },
    {
      "entropy": 5.595453786849975,
      "epoch": 1.1537316395394999,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.2446,
      "mean_token_accuracy": 0.1581490859389305,
      "num_tokens": 25462872.0,
      "step": 11625
    },
    {
      "entropy": 5.558007287979126,
      "epoch": 1.1542278682016673,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.226,
      "mean_token_accuracy": 0.1628376469016075,
      "num_tokens": 25473538.0,
      "step": 11630
    },
    {
      "entropy": 5.5229957580566404,
      "epoch": 1.1547240968638348,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.1552,
      "mean_token_accuracy": 0.17118891775608064,
      "num_tokens": 25484707.0,
      "step": 11635
    },
    {
      "entropy": 5.567117166519165,
      "epoch": 1.1552203255260023,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.1564,
      "mean_token_accuracy": 0.1662668138742447,
      "num_tokens": 25496155.0,
      "step": 11640
    },
    {
      "entropy": 5.533782529830932,
      "epoch": 1.15571655418817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.185,
      "mean_token_accuracy": 0.16940237507224082,
      "num_tokens": 25507146.0,
      "step": 11645
    },
    {
      "entropy": 5.523106527328491,
      "epoch": 1.1562127828503375,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.1557,
      "mean_token_accuracy": 0.16525058150291444,
      "num_tokens": 25519099.0,
      "step": 11650
    },
    {
      "entropy": 5.529668092727661,
      "epoch": 1.156709011512505,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.1249,
      "mean_token_accuracy": 0.17752259522676467,
      "num_tokens": 25529491.0,
      "step": 11655
    },
    {
      "entropy": 5.48610520362854,
      "epoch": 1.1572052401746724,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.095,
      "mean_token_accuracy": 0.1788385421037674,
      "num_tokens": 25540889.0,
      "step": 11660
    },
    {
      "entropy": 5.464457368850708,
      "epoch": 1.15770146883684,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16780339032411576,
      "num_tokens": 25551334.0,
      "step": 11665
    },
    {
      "entropy": 5.53451681137085,
      "epoch": 1.1581976974990076,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.157,
      "mean_token_accuracy": 0.16737764626741408,
      "num_tokens": 25561538.0,
      "step": 11670
    },
    {
      "entropy": 5.472196626663208,
      "epoch": 1.158693926161175,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.1415,
      "mean_token_accuracy": 0.16963538229465486,
      "num_tokens": 25571652.0,
      "step": 11675
    },
    {
      "entropy": 5.524732971191407,
      "epoch": 1.1591901548233425,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.2496,
      "mean_token_accuracy": 0.16036608219146728,
      "num_tokens": 25582799.0,
      "step": 11680
    },
    {
      "entropy": 5.521860647201538,
      "epoch": 1.1596863834855102,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.1776,
      "mean_token_accuracy": 0.16820574998855592,
      "num_tokens": 25594540.0,
      "step": 11685
    },
    {
      "entropy": 5.6300219058990475,
      "epoch": 1.1601826121476777,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1669654741883278,
      "num_tokens": 25604656.0,
      "step": 11690
    },
    {
      "entropy": 5.660798597335815,
      "epoch": 1.1606788408098452,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.2145,
      "mean_token_accuracy": 0.1636493593454361,
      "num_tokens": 25615385.0,
      "step": 11695
    },
    {
      "entropy": 5.529118156433105,
      "epoch": 1.1611750694720127,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.149,
      "mean_token_accuracy": 0.17255172580480577,
      "num_tokens": 25626456.0,
      "step": 11700
    },
    {
      "entropy": 5.533388137817383,
      "epoch": 1.1616712981341801,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.1421,
      "mean_token_accuracy": 0.168950617313385,
      "num_tokens": 25636098.0,
      "step": 11705
    },
    {
      "entropy": 5.501568460464478,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.2012,
      "mean_token_accuracy": 0.1694554165005684,
      "num_tokens": 25646578.0,
      "step": 11710
    },
    {
      "entropy": 5.5583991527557375,
      "epoch": 1.1626637554585153,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.3201,
      "mean_token_accuracy": 0.16465546637773515,
      "num_tokens": 25658587.0,
      "step": 11715
    },
    {
      "entropy": 5.575591993331909,
      "epoch": 1.1631599841206828,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.16,
      "mean_token_accuracy": 0.16496191024780274,
      "num_tokens": 25667745.0,
      "step": 11720
    },
    {
      "entropy": 5.605786466598511,
      "epoch": 1.1636562127828503,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000487288182361241,
      "loss": 5.2733,
      "mean_token_accuracy": 0.16624002754688263,
      "num_tokens": 25679034.0,
      "step": 11725
    },
    {
      "entropy": 5.616104698181152,
      "epoch": 1.164152441445018,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1674956575036049,
      "num_tokens": 25689352.0,
      "step": 11730
    },
    {
      "entropy": 5.526976871490478,
      "epoch": 1.1646486701071854,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.1171,
      "mean_token_accuracy": 0.16972881704568862,
      "num_tokens": 25700425.0,
      "step": 11735
    },
    {
      "entropy": 5.582541561126709,
      "epoch": 1.165144898769353,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.2262,
      "mean_token_accuracy": 0.16485716104507447,
      "num_tokens": 25711641.0,
      "step": 11740
    },
    {
      "entropy": 5.6054810047149655,
      "epoch": 1.1656411274315204,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16996169835329056,
      "num_tokens": 25721725.0,
      "step": 11745
    },
    {
      "entropy": 5.5174681663513185,
      "epoch": 1.1661373560936878,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.2601,
      "mean_token_accuracy": 0.16537000238895416,
      "num_tokens": 25731465.0,
      "step": 11750
    },
    {
      "entropy": 5.53748459815979,
      "epoch": 1.1666335847558555,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.1125,
      "mean_token_accuracy": 0.16587963402271272,
      "num_tokens": 25741864.0,
      "step": 11755
    },
    {
      "entropy": 5.557615900039673,
      "epoch": 1.167129813418023,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.2555,
      "mean_token_accuracy": 0.16198045909404754,
      "num_tokens": 25753486.0,
      "step": 11760
    },
    {
      "entropy": 5.549865198135376,
      "epoch": 1.1676260420801905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.1516,
      "mean_token_accuracy": 0.16764950901269912,
      "num_tokens": 25763361.0,
      "step": 11765
    },
    {
      "entropy": 5.527234172821045,
      "epoch": 1.1681222707423582,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.1575,
      "mean_token_accuracy": 0.16450496017932892,
      "num_tokens": 25772157.0,
      "step": 11770
    },
    {
      "entropy": 5.516959238052368,
      "epoch": 1.1686184994045257,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16618189066648484,
      "num_tokens": 25783948.0,
      "step": 11775
    },
    {
      "entropy": 5.589645624160767,
      "epoch": 1.1691147280666931,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.2136,
      "mean_token_accuracy": 0.1638169452548027,
      "num_tokens": 25794762.0,
      "step": 11780
    },
    {
      "entropy": 5.538985013961792,
      "epoch": 1.1696109567288606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1692478448152542,
      "num_tokens": 25805519.0,
      "step": 11785
    },
    {
      "entropy": 5.613421058654785,
      "epoch": 1.170107185391028,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.2728,
      "mean_token_accuracy": 0.15733546167612075,
      "num_tokens": 25816772.0,
      "step": 11790
    },
    {
      "entropy": 5.535953330993652,
      "epoch": 1.1706034140531958,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.2128,
      "mean_token_accuracy": 0.16337545216083527,
      "num_tokens": 25826649.0,
      "step": 11795
    },
    {
      "entropy": 5.603769969940186,
      "epoch": 1.1710996427153633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.256,
      "mean_token_accuracy": 0.1603723093867302,
      "num_tokens": 25837263.0,
      "step": 11800
    },
    {
      "entropy": 5.6373053073883055,
      "epoch": 1.1715958713775307,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1611332416534424,
      "num_tokens": 25848687.0,
      "step": 11805
    },
    {
      "entropy": 5.564898872375489,
      "epoch": 1.1720921000396982,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.163,
      "mean_token_accuracy": 0.169627645611763,
      "num_tokens": 25857982.0,
      "step": 11810
    },
    {
      "entropy": 5.524759674072266,
      "epoch": 1.172588328701866,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.2361,
      "mean_token_accuracy": 0.16359006315469743,
      "num_tokens": 25869772.0,
      "step": 11815
    },
    {
      "entropy": 5.573530673980713,
      "epoch": 1.1730845573640334,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.2748,
      "mean_token_accuracy": 0.1671598255634308,
      "num_tokens": 25881165.0,
      "step": 11820
    },
    {
      "entropy": 5.543649291992187,
      "epoch": 1.1735807860262009,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.168,
      "mean_token_accuracy": 0.16921925991773606,
      "num_tokens": 25891135.0,
      "step": 11825
    },
    {
      "entropy": 5.5515154838562015,
      "epoch": 1.1740770146883683,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.2016,
      "mean_token_accuracy": 0.16082285940647126,
      "num_tokens": 25901837.0,
      "step": 11830
    },
    {
      "entropy": 5.614464569091797,
      "epoch": 1.1745732433505358,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000487028613117317,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1587645560503006,
      "num_tokens": 25913833.0,
      "step": 11835
    },
    {
      "entropy": 5.574047470092774,
      "epoch": 1.1750694720127035,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.1772,
      "mean_token_accuracy": 0.16558087170124053,
      "num_tokens": 25924803.0,
      "step": 11840
    },
    {
      "entropy": 5.599658870697022,
      "epoch": 1.175565700674871,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.2391,
      "mean_token_accuracy": 0.16776734888553618,
      "num_tokens": 25935582.0,
      "step": 11845
    },
    {
      "entropy": 5.579851818084717,
      "epoch": 1.1760619293370385,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1618258312344551,
      "num_tokens": 25947515.0,
      "step": 11850
    },
    {
      "entropy": 5.50541877746582,
      "epoch": 1.1765581579992062,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.2039,
      "mean_token_accuracy": 0.16348015666007995,
      "num_tokens": 25957238.0,
      "step": 11855
    },
    {
      "entropy": 5.479667091369629,
      "epoch": 1.1770543866613736,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.1746,
      "mean_token_accuracy": 0.168183895945549,
      "num_tokens": 25968315.0,
      "step": 11860
    },
    {
      "entropy": 5.5505918025970455,
      "epoch": 1.177550615323541,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1738766223192215,
      "num_tokens": 25978172.0,
      "step": 11865
    },
    {
      "entropy": 5.541842937469482,
      "epoch": 1.1780468439857086,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.1979,
      "mean_token_accuracy": 0.1644609659910202,
      "num_tokens": 25989720.0,
      "step": 11870
    },
    {
      "entropy": 5.602197694778442,
      "epoch": 1.178543072647876,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1622913137078285,
      "num_tokens": 26000394.0,
      "step": 11875
    },
    {
      "entropy": 5.579372978210449,
      "epoch": 1.1790393013100438,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.2846,
      "mean_token_accuracy": 0.15705638378858566,
      "num_tokens": 26012183.0,
      "step": 11880
    },
    {
      "entropy": 5.554619312286377,
      "epoch": 1.1795355299722112,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.2175,
      "mean_token_accuracy": 0.1665964677929878,
      "num_tokens": 26022711.0,
      "step": 11885
    },
    {
      "entropy": 5.46068058013916,
      "epoch": 1.1800317586343787,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.0833,
      "mean_token_accuracy": 0.1724819377064705,
      "num_tokens": 26033800.0,
      "step": 11890
    },
    {
      "entropy": 5.538817739486694,
      "epoch": 1.1805279872965462,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.2677,
      "mean_token_accuracy": 0.16112611070275307,
      "num_tokens": 26045461.0,
      "step": 11895
    },
    {
      "entropy": 5.598849534988403,
      "epoch": 1.1810242159587139,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.3143,
      "mean_token_accuracy": 0.15816771537065505,
      "num_tokens": 26057674.0,
      "step": 11900
    },
    {
      "entropy": 5.590289163589477,
      "epoch": 1.1815204446208813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000486862107513917,
      "loss": 5.2141,
      "mean_token_accuracy": 0.16285006254911422,
      "num_tokens": 26069995.0,
      "step": 11905
    },
    {
      "entropy": 5.594806480407715,
      "epoch": 1.1820166732830488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16426704674959183,
      "num_tokens": 26079969.0,
      "step": 11910
    },
    {
      "entropy": 5.561361265182495,
      "epoch": 1.1825129019452163,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.2322,
      "mean_token_accuracy": 0.16911201030015946,
      "num_tokens": 26089865.0,
      "step": 11915
    },
    {
      "entropy": 5.524196434020996,
      "epoch": 1.1830091306073838,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.188,
      "mean_token_accuracy": 0.16885925978422164,
      "num_tokens": 26101349.0,
      "step": 11920
    },
    {
      "entropy": 5.574898910522461,
      "epoch": 1.1835053592695515,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.2384,
      "mean_token_accuracy": 0.16266244798898696,
      "num_tokens": 26113933.0,
      "step": 11925
    },
    {
      "entropy": 5.626174736022949,
      "epoch": 1.184001587931719,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.2408,
      "mean_token_accuracy": 0.16621812283992768,
      "num_tokens": 26125106.0,
      "step": 11930
    },
    {
      "entropy": 5.58794264793396,
      "epoch": 1.1844978165938864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.2584,
      "mean_token_accuracy": 0.16033971309661865,
      "num_tokens": 26135131.0,
      "step": 11935
    },
    {
      "entropy": 5.610899257659912,
      "epoch": 1.184994045256054,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.3244,
      "mean_token_accuracy": 0.15547003895044326,
      "num_tokens": 26147053.0,
      "step": 11940
    },
    {
      "entropy": 5.60792350769043,
      "epoch": 1.1854902739182216,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.1961,
      "mean_token_accuracy": 0.16453517079353333,
      "num_tokens": 26157465.0,
      "step": 11945
    },
    {
      "entropy": 5.561259794235229,
      "epoch": 1.185986502580389,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.1736,
      "mean_token_accuracy": 0.16757160574197769,
      "num_tokens": 26168383.0,
      "step": 11950
    },
    {
      "entropy": 5.591293478012085,
      "epoch": 1.1864827312425565,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.2415,
      "mean_token_accuracy": 0.16471500396728517,
      "num_tokens": 26178316.0,
      "step": 11955
    },
    {
      "entropy": 5.528698396682739,
      "epoch": 1.186978959904724,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.139,
      "mean_token_accuracy": 0.1680547133088112,
      "num_tokens": 26189494.0,
      "step": 11960
    },
    {
      "entropy": 5.5247955322265625,
      "epoch": 1.1874751885668917,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.2263,
      "mean_token_accuracy": 0.16259261220693588,
      "num_tokens": 26200865.0,
      "step": 11965
    },
    {
      "entropy": 5.5474440574646,
      "epoch": 1.1879714172290592,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16449961066246033,
      "num_tokens": 26212187.0,
      "step": 11970
    },
    {
      "entropy": 5.5477245330810545,
      "epoch": 1.1884676458912267,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.1971,
      "mean_token_accuracy": 0.16790599822998048,
      "num_tokens": 26222059.0,
      "step": 11975
    },
    {
      "entropy": 5.558330392837524,
      "epoch": 1.1889638745533941,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.2131,
      "mean_token_accuracy": 0.16820578426122665,
      "num_tokens": 26232630.0,
      "step": 11980
    },
    {
      "entropy": 5.5456281185150145,
      "epoch": 1.1894601032155618,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.2231,
      "mean_token_accuracy": 0.1680737778544426,
      "num_tokens": 26244232.0,
      "step": 11985
    },
    {
      "entropy": 5.646167755126953,
      "epoch": 1.1899563318777293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.3123,
      "mean_token_accuracy": 0.16620631888508797,
      "num_tokens": 26254721.0,
      "step": 11990
    },
    {
      "entropy": 5.557544898986817,
      "epoch": 1.1904525605398968,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.2164,
      "mean_token_accuracy": 0.16272016614675522,
      "num_tokens": 26266174.0,
      "step": 11995
    },
    {
      "entropy": 5.562256288528443,
      "epoch": 1.1909487892020643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.2601,
      "mean_token_accuracy": 0.1643774315714836,
      "num_tokens": 26277501.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.388576689398637,
      "eval_loss": 5.310436725616455,
      "eval_mean_token_accuracy": 0.16628264827597564,
      "eval_num_tokens": 26277501.0,
      "eval_runtime": 20.5362,
      "eval_samples_per_second": 1639.492,
      "eval_steps_per_second": 204.955,
      "step": 12000
    },
    {
      "entropy": 5.555384826660156,
      "epoch": 1.1914450178642317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.116,
      "mean_token_accuracy": 0.1720243826508522,
      "num_tokens": 26287291.0,
      "step": 12005
    },
    {
      "entropy": 5.4951423645019535,
      "epoch": 1.1919412465263994,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17727535963058472,
      "num_tokens": 26297259.0,
      "step": 12010
    },
    {
      "entropy": 5.460361528396606,
      "epoch": 1.192437475188567,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.1687,
      "mean_token_accuracy": 0.16923524737358092,
      "num_tokens": 26307873.0,
      "step": 12015
    },
    {
      "entropy": 5.589886331558228,
      "epoch": 1.1929337038507344,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.2164,
      "mean_token_accuracy": 0.1645038478076458,
      "num_tokens": 26320421.0,
      "step": 12020
    },
    {
      "entropy": 5.609117221832276,
      "epoch": 1.1934299325129019,
      "grad_norm": 0.875,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.2279,
      "mean_token_accuracy": 0.16438471227884294,
      "num_tokens": 26330932.0,
      "step": 12025
    },
    {
      "entropy": 5.587629222869873,
      "epoch": 1.1939261611750696,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.227,
      "mean_token_accuracy": 0.15783288478851318,
      "num_tokens": 26342286.0,
      "step": 12030
    },
    {
      "entropy": 5.498705148696899,
      "epoch": 1.194422389837237,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.1098,
      "mean_token_accuracy": 0.16899092644453048,
      "num_tokens": 26352210.0,
      "step": 12035
    },
    {
      "entropy": 5.5282965183258055,
      "epoch": 1.1949186184994045,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.2608,
      "mean_token_accuracy": 0.16518405228853225,
      "num_tokens": 26363630.0,
      "step": 12040
    },
    {
      "entropy": 5.492415475845337,
      "epoch": 1.195414847161572,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000486526007933085,
      "loss": 5.2053,
      "mean_token_accuracy": 0.16128052026033401,
      "num_tokens": 26375723.0,
      "step": 12045
    },
    {
      "entropy": 5.557999324798584,
      "epoch": 1.1959110758237395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.1642,
      "mean_token_accuracy": 0.16735564470291137,
      "num_tokens": 26387551.0,
      "step": 12050
    },
    {
      "entropy": 5.526860141754151,
      "epoch": 1.1964073044859072,
      "grad_norm": 0.8125,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.133,
      "mean_token_accuracy": 0.17189502120018005,
      "num_tokens": 26399015.0,
      "step": 12055
    },
    {
      "entropy": 5.468436002731323,
      "epoch": 1.1969035331480746,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17859987765550614,
      "num_tokens": 26409242.0,
      "step": 12060
    },
    {
      "entropy": 5.532299280166626,
      "epoch": 1.197399761810242,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.2774,
      "mean_token_accuracy": 0.1599874585866928,
      "num_tokens": 26419189.0,
      "step": 12065
    },
    {
      "entropy": 5.530673933029175,
      "epoch": 1.1978959904724098,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1694414123892784,
      "num_tokens": 26429957.0,
      "step": 12070
    },
    {
      "entropy": 5.52723503112793,
      "epoch": 1.1983922191345773,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.1711,
      "mean_token_accuracy": 0.16879020631313324,
      "num_tokens": 26441088.0,
      "step": 12075
    },
    {
      "entropy": 5.5665703296661375,
      "epoch": 1.1988884477967447,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.2259,
      "mean_token_accuracy": 0.16330428570508956,
      "num_tokens": 26450891.0,
      "step": 12080
    },
    {
      "entropy": 5.508478164672852,
      "epoch": 1.1993846764589122,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.1433,
      "mean_token_accuracy": 0.16968042999505997,
      "num_tokens": 26460445.0,
      "step": 12085
    },
    {
      "entropy": 5.518480396270752,
      "epoch": 1.1998809051210797,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1663826271891594,
      "num_tokens": 26472110.0,
      "step": 12090
    },
    {
      "entropy": 5.482213687896729,
      "epoch": 1.2003771337832474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17413381934165956,
      "num_tokens": 26483122.0,
      "step": 12095
    },
    {
      "entropy": 5.494415903091431,
      "epoch": 1.2008733624454149,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.1701,
      "mean_token_accuracy": 0.16388996243476867,
      "num_tokens": 26495541.0,
      "step": 12100
    },
    {
      "entropy": 5.49240779876709,
      "epoch": 1.2013695911075823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.0797,
      "mean_token_accuracy": 0.18035586327314376,
      "num_tokens": 26504904.0,
      "step": 12105
    },
    {
      "entropy": 5.5502636432647705,
      "epoch": 1.2018658197697498,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000486368563405542,
      "loss": 5.2159,
      "mean_token_accuracy": 0.16807728111743928,
      "num_tokens": 26515850.0,
      "step": 12110
    },
    {
      "entropy": 5.587061834335327,
      "epoch": 1.2023620484319175,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1702674448490143,
      "num_tokens": 26526671.0,
      "step": 12115
    },
    {
      "entropy": 5.5095583438873295,
      "epoch": 1.202858277094085,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.2252,
      "mean_token_accuracy": 0.17039599865674973,
      "num_tokens": 26537919.0,
      "step": 12120
    },
    {
      "entropy": 5.588599014282226,
      "epoch": 1.2033545057562525,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1675727277994156,
      "num_tokens": 26547903.0,
      "step": 12125
    },
    {
      "entropy": 5.598510265350342,
      "epoch": 1.20385073441842,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.2998,
      "mean_token_accuracy": 0.1645420640707016,
      "num_tokens": 26559253.0,
      "step": 12130
    },
    {
      "entropy": 5.596667957305908,
      "epoch": 1.2043469630805874,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.1795,
      "mean_token_accuracy": 0.16862173229455948,
      "num_tokens": 26569562.0,
      "step": 12135
    },
    {
      "entropy": 5.518482780456543,
      "epoch": 1.2048431917427551,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.1525,
      "mean_token_accuracy": 0.17335347086191177,
      "num_tokens": 26580701.0,
      "step": 12140
    },
    {
      "entropy": 5.471106243133545,
      "epoch": 1.2053394204049226,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000486283418637006,
      "loss": 5.2231,
      "mean_token_accuracy": 0.1655241459608078,
      "num_tokens": 26592233.0,
      "step": 12145
    },
    {
      "entropy": 5.542800998687744,
      "epoch": 1.20583564906709,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.1804,
      "mean_token_accuracy": 0.1658639669418335,
      "num_tokens": 26603073.0,
      "step": 12150
    },
    {
      "entropy": 5.568694257736206,
      "epoch": 1.2063318777292578,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.1822,
      "mean_token_accuracy": 0.16344834268093109,
      "num_tokens": 26612497.0,
      "step": 12155
    },
    {
      "entropy": 5.555770826339722,
      "epoch": 1.2068281063914252,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.1891,
      "mean_token_accuracy": 0.172037573158741,
      "num_tokens": 26623616.0,
      "step": 12160
    },
    {
      "entropy": 5.561669874191284,
      "epoch": 1.2073243350535927,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.2605,
      "mean_token_accuracy": 0.16387614905834197,
      "num_tokens": 26634571.0,
      "step": 12165
    },
    {
      "entropy": 5.507146406173706,
      "epoch": 1.2078205637157602,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048622244375796,
      "loss": 5.1566,
      "mean_token_accuracy": 0.16821479350328444,
      "num_tokens": 26646635.0,
      "step": 12170
    },
    {
      "entropy": 5.532960367202759,
      "epoch": 1.2083167923779277,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.1874,
      "mean_token_accuracy": 0.16284343004226684,
      "num_tokens": 26657923.0,
      "step": 12175
    },
    {
      "entropy": 5.553115320205689,
      "epoch": 1.2088130210400954,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.1516,
      "mean_token_accuracy": 0.16993044018745423,
      "num_tokens": 26666989.0,
      "step": 12180
    },
    {
      "entropy": 5.517288875579834,
      "epoch": 1.2093092497022628,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.245,
      "mean_token_accuracy": 0.15715040862560273,
      "num_tokens": 26677325.0,
      "step": 12185
    },
    {
      "entropy": 5.5217184066772464,
      "epoch": 1.2098054783644303,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16748420745134354,
      "num_tokens": 26688306.0,
      "step": 12190
    },
    {
      "entropy": 5.533965826034546,
      "epoch": 1.2103017070265978,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.1716,
      "mean_token_accuracy": 0.16325385123491287,
      "num_tokens": 26698059.0,
      "step": 12195
    },
    {
      "entropy": 5.512914848327637,
      "epoch": 1.2107979356887655,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.1988,
      "mean_token_accuracy": 0.1644452303647995,
      "num_tokens": 26708165.0,
      "step": 12200
    },
    {
      "entropy": 5.585209035873413,
      "epoch": 1.211294164350933,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.2817,
      "mean_token_accuracy": 0.1636446863412857,
      "num_tokens": 26720010.0,
      "step": 12205
    },
    {
      "entropy": 5.537531089782715,
      "epoch": 1.2117903930131004,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.2192,
      "mean_token_accuracy": 0.16603850871324538,
      "num_tokens": 26732756.0,
      "step": 12210
    },
    {
      "entropy": 5.580551385879517,
      "epoch": 1.212286621675268,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.2075,
      "mean_token_accuracy": 0.169172440469265,
      "num_tokens": 26744656.0,
      "step": 12215
    },
    {
      "entropy": 5.615132808685303,
      "epoch": 1.2127828503374354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.2209,
      "mean_token_accuracy": 0.16609493941068648,
      "num_tokens": 26756676.0,
      "step": 12220
    },
    {
      "entropy": 5.52651309967041,
      "epoch": 1.213279078999603,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.2089,
      "mean_token_accuracy": 0.16779153794050217,
      "num_tokens": 26767671.0,
      "step": 12225
    },
    {
      "entropy": 5.500944471359253,
      "epoch": 1.2137753076617706,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.153,
      "mean_token_accuracy": 0.17726381123065948,
      "num_tokens": 26779149.0,
      "step": 12230
    },
    {
      "entropy": 5.521598291397095,
      "epoch": 1.214271536323938,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.1462,
      "mean_token_accuracy": 0.16852051615715027,
      "num_tokens": 26789825.0,
      "step": 12235
    },
    {
      "entropy": 5.569707870483398,
      "epoch": 1.2147677649861055,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16673970371484756,
      "num_tokens": 26802145.0,
      "step": 12240
    },
    {
      "entropy": 5.583461141586303,
      "epoch": 1.2152639936482732,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.1963,
      "mean_token_accuracy": 0.16869306713342666,
      "num_tokens": 26811925.0,
      "step": 12245
    },
    {
      "entropy": 5.565780353546143,
      "epoch": 1.2157602223104407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16347622722387314,
      "num_tokens": 26821178.0,
      "step": 12250
    },
    {
      "entropy": 5.49365200996399,
      "epoch": 1.2162564509726081,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16858891695737838,
      "num_tokens": 26831563.0,
      "step": 12255
    },
    {
      "entropy": 5.662156057357788,
      "epoch": 1.2167526796347756,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.2903,
      "mean_token_accuracy": 0.15964837968349457,
      "num_tokens": 26844596.0,
      "step": 12260
    },
    {
      "entropy": 5.604224014282226,
      "epoch": 1.2172489082969433,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.2244,
      "mean_token_accuracy": 0.1661066323518753,
      "num_tokens": 26855149.0,
      "step": 12265
    },
    {
      "entropy": 5.538290357589721,
      "epoch": 1.2177451369591108,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1687716767191887,
      "num_tokens": 26866200.0,
      "step": 12270
    },
    {
      "entropy": 5.487606716156006,
      "epoch": 1.2182413656212783,
      "grad_norm": 0.921875,
      "learning_rate": 0.000485964919771632,
      "loss": 5.1482,
      "mean_token_accuracy": 0.17340244948863984,
      "num_tokens": 26876834.0,
      "step": 12275
    },
    {
      "entropy": 5.545223569869995,
      "epoch": 1.2187375942834457,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.106,
      "mean_token_accuracy": 0.17213979810476304,
      "num_tokens": 26887675.0,
      "step": 12280
    },
    {
      "entropy": 5.509747457504273,
      "epoch": 1.2192338229456134,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.1084,
      "mean_token_accuracy": 0.172177554666996,
      "num_tokens": 26897650.0,
      "step": 12285
    },
    {
      "entropy": 5.475010395050049,
      "epoch": 1.219730051607781,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1632297858595848,
      "num_tokens": 26908404.0,
      "step": 12290
    },
    {
      "entropy": 5.4568071365356445,
      "epoch": 1.2202262802699484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.196,
      "mean_token_accuracy": 0.17165374010801315,
      "num_tokens": 26917184.0,
      "step": 12295
    },
    {
      "entropy": 5.567473411560059,
      "epoch": 1.2207225089321159,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.1732,
      "mean_token_accuracy": 0.1754811331629753,
      "num_tokens": 26927757.0,
      "step": 12300
    },
    {
      "entropy": 5.558986759185791,
      "epoch": 1.2212187375942833,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16697036772966384,
      "num_tokens": 26939036.0,
      "step": 12305
    },
    {
      "entropy": 5.491768074035645,
      "epoch": 1.221714966256451,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.1608,
      "mean_token_accuracy": 0.1709651067852974,
      "num_tokens": 26949637.0,
      "step": 12310
    },
    {
      "entropy": 5.513642168045044,
      "epoch": 1.2222111949186185,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.1928,
      "mean_token_accuracy": 0.16743892729282378,
      "num_tokens": 26961333.0,
      "step": 12315
    },
    {
      "entropy": 5.541067600250244,
      "epoch": 1.222707423580786,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.1559,
      "mean_token_accuracy": 0.16534585654735565,
      "num_tokens": 26971830.0,
      "step": 12320
    },
    {
      "entropy": 5.534152793884277,
      "epoch": 1.2232036522429535,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.196,
      "mean_token_accuracy": 0.16568403989076613,
      "num_tokens": 26982808.0,
      "step": 12325
    },
    {
      "entropy": 5.536502647399902,
      "epoch": 1.2236998809051212,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.1739,
      "mean_token_accuracy": 0.164012111723423,
      "num_tokens": 26992593.0,
      "step": 12330
    },
    {
      "entropy": 5.536544513702393,
      "epoch": 1.2241961095672886,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16600132882595062,
      "num_tokens": 27003107.0,
      "step": 12335
    },
    {
      "entropy": 5.5467430591583256,
      "epoch": 1.2246923382294561,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.1634,
      "mean_token_accuracy": 0.16556430757045745,
      "num_tokens": 27013841.0,
      "step": 12340
    },
    {
      "entropy": 5.591259145736695,
      "epoch": 1.2251885668916236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.1933,
      "mean_token_accuracy": 0.1653233528137207,
      "num_tokens": 27026236.0,
      "step": 12345
    },
    {
      "entropy": 5.555192852020264,
      "epoch": 1.225684795553791,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.195,
      "mean_token_accuracy": 0.1686908036470413,
      "num_tokens": 27037517.0,
      "step": 12350
    },
    {
      "entropy": 5.468619871139526,
      "epoch": 1.2261810242159588,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17396840453147888,
      "num_tokens": 27048974.0,
      "step": 12355
    },
    {
      "entropy": 5.539198970794677,
      "epoch": 1.2266772528781262,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.1583,
      "mean_token_accuracy": 0.1757863223552704,
      "num_tokens": 27059637.0,
      "step": 12360
    },
    {
      "entropy": 5.544380807876587,
      "epoch": 1.2271734815402937,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.157,
      "mean_token_accuracy": 0.16767807751893998,
      "num_tokens": 27069584.0,
      "step": 12365
    },
    {
      "entropy": 5.501195287704467,
      "epoch": 1.2276697102024614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16618892848491668,
      "num_tokens": 27080078.0,
      "step": 12370
    },
    {
      "entropy": 5.540884494781494,
      "epoch": 1.2281659388646289,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16999032348394394,
      "num_tokens": 27090396.0,
      "step": 12375
    },
    {
      "entropy": 5.5559690475463865,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.2031,
      "mean_token_accuracy": 0.17036662846803666,
      "num_tokens": 27101052.0,
      "step": 12380
    },
    {
      "entropy": 5.535504150390625,
      "epoch": 1.2291583961889638,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.2029,
      "mean_token_accuracy": 0.16298374086618422,
      "num_tokens": 27111352.0,
      "step": 12385
    },
    {
      "entropy": 5.537643718719482,
      "epoch": 1.2296546248511313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000485680222670505,
      "loss": 5.1853,
      "mean_token_accuracy": 0.16948407143354416,
      "num_tokens": 27122589.0,
      "step": 12390
    },
    {
      "entropy": 5.508548784255981,
      "epoch": 1.230150853513299,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16873765140771865,
      "num_tokens": 27133445.0,
      "step": 12395
    },
    {
      "entropy": 5.548293113708496,
      "epoch": 1.2306470821754665,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.2085,
      "mean_token_accuracy": 0.16404973715543747,
      "num_tokens": 27144043.0,
      "step": 12400
    },
    {
      "entropy": 5.657059526443481,
      "epoch": 1.231143310837634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.2422,
      "mean_token_accuracy": 0.16029953807592393,
      "num_tokens": 27155235.0,
      "step": 12405
    },
    {
      "entropy": 5.576830434799194,
      "epoch": 1.2316395394998014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.183,
      "mean_token_accuracy": 0.16961089819669722,
      "num_tokens": 27166360.0,
      "step": 12410
    },
    {
      "entropy": 5.493213415145874,
      "epoch": 1.2321357681619691,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.2378,
      "mean_token_accuracy": 0.16184543520212175,
      "num_tokens": 27178018.0,
      "step": 12415
    },
    {
      "entropy": 5.608360528945923,
      "epoch": 1.2326319968241366,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16805819422006607,
      "num_tokens": 27187609.0,
      "step": 12420
    },
    {
      "entropy": 5.491872358322143,
      "epoch": 1.233128225486304,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.1251,
      "mean_token_accuracy": 0.17135262936353685,
      "num_tokens": 27197477.0,
      "step": 12425
    },
    {
      "entropy": 5.514398288726807,
      "epoch": 1.2336244541484715,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.1666,
      "mean_token_accuracy": 0.166526398062706,
      "num_tokens": 27209276.0,
      "step": 12430
    },
    {
      "entropy": 5.589037895202637,
      "epoch": 1.234120682810639,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.2445,
      "mean_token_accuracy": 0.16098628640174867,
      "num_tokens": 27221355.0,
      "step": 12435
    },
    {
      "entropy": 5.572207450866699,
      "epoch": 1.2346169114728067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16190543472766877,
      "num_tokens": 27231408.0,
      "step": 12440
    },
    {
      "entropy": 5.565357780456543,
      "epoch": 1.2351131401349742,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.1552,
      "mean_token_accuracy": 0.16832583546638488,
      "num_tokens": 27242577.0,
      "step": 12445
    },
    {
      "entropy": 5.593330812454224,
      "epoch": 1.2356093687971417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1595939129590988,
      "num_tokens": 27252101.0,
      "step": 12450
    },
    {
      "entropy": 5.522315502166748,
      "epoch": 1.2361055974593094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.2417,
      "mean_token_accuracy": 0.16030385792255403,
      "num_tokens": 27263281.0,
      "step": 12455
    },
    {
      "entropy": 5.4901378631591795,
      "epoch": 1.2366018261214768,
      "grad_norm": 0.875,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.1071,
      "mean_token_accuracy": 0.16943009346723556,
      "num_tokens": 27273747.0,
      "step": 12460
    },
    {
      "entropy": 5.547404527664185,
      "epoch": 1.2370980547836443,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.2992,
      "mean_token_accuracy": 0.161282317340374,
      "num_tokens": 27285282.0,
      "step": 12465
    },
    {
      "entropy": 5.553569078445435,
      "epoch": 1.2375942834458118,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.1404,
      "mean_token_accuracy": 0.17210787087678908,
      "num_tokens": 27297258.0,
      "step": 12470
    },
    {
      "entropy": 5.603307867050171,
      "epoch": 1.2380905121079793,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.1971,
      "mean_token_accuracy": 0.1636384442448616,
      "num_tokens": 27309366.0,
      "step": 12475
    },
    {
      "entropy": 5.4855224132537845,
      "epoch": 1.238586740770147,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17058235108852388,
      "num_tokens": 27319878.0,
      "step": 12480
    },
    {
      "entropy": 5.5055114269256595,
      "epoch": 1.2390829694323144,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.1063,
      "mean_token_accuracy": 0.17339075058698655,
      "num_tokens": 27330056.0,
      "step": 12485
    },
    {
      "entropy": 5.552078199386597,
      "epoch": 1.239579198094482,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.1932,
      "mean_token_accuracy": 0.16657274663448335,
      "num_tokens": 27342364.0,
      "step": 12490
    },
    {
      "entropy": 5.563103485107422,
      "epoch": 1.2400754267566494,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.1241,
      "mean_token_accuracy": 0.16964229196310043,
      "num_tokens": 27352651.0,
      "step": 12495
    },
    {
      "entropy": 5.545586824417114,
      "epoch": 1.240571655418817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.2158,
      "mean_token_accuracy": 0.16983262896537782,
      "num_tokens": 27364161.0,
      "step": 12500
    },
    {
      "entropy": 5.530094337463379,
      "epoch": 1.2410678840809846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.1817,
      "mean_token_accuracy": 0.17422365248203278,
      "num_tokens": 27376461.0,
      "step": 12505
    },
    {
      "entropy": 5.57455792427063,
      "epoch": 1.241564112743152,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.2309,
      "mean_token_accuracy": 0.16471761465072632,
      "num_tokens": 27386312.0,
      "step": 12510
    },
    {
      "entropy": 5.506299209594727,
      "epoch": 1.2420603414053195,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.1754,
      "mean_token_accuracy": 0.16553114205598832,
      "num_tokens": 27397971.0,
      "step": 12515
    },
    {
      "entropy": 5.540783786773682,
      "epoch": 1.242556570067487,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.0931,
      "mean_token_accuracy": 0.164743110537529,
      "num_tokens": 27409942.0,
      "step": 12520
    },
    {
      "entropy": 5.603008031845093,
      "epoch": 1.2430527987296547,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.19,
      "mean_token_accuracy": 0.1655370458960533,
      "num_tokens": 27421828.0,
      "step": 12525
    },
    {
      "entropy": 5.503949880599976,
      "epoch": 1.2435490273918222,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.1126,
      "mean_token_accuracy": 0.17701600193977357,
      "num_tokens": 27433321.0,
      "step": 12530
    },
    {
      "entropy": 5.504921579360962,
      "epoch": 1.2440452560539896,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.1403,
      "mean_token_accuracy": 0.16911133974790574,
      "num_tokens": 27443030.0,
      "step": 12535
    },
    {
      "entropy": 5.50619912147522,
      "epoch": 1.244541484716157,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17548736333847045,
      "num_tokens": 27453920.0,
      "step": 12540
    },
    {
      "entropy": 5.477845573425293,
      "epoch": 1.2450377133783248,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.1658,
      "mean_token_accuracy": 0.17303164452314376,
      "num_tokens": 27464014.0,
      "step": 12545
    },
    {
      "entropy": 5.649668884277344,
      "epoch": 1.2455339420404923,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.2554,
      "mean_token_accuracy": 0.16950362548232079,
      "num_tokens": 27475555.0,
      "step": 12550
    },
    {
      "entropy": 5.527647638320923,
      "epoch": 1.2460301707026598,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.1379,
      "mean_token_accuracy": 0.16662439554929734,
      "num_tokens": 27486624.0,
      "step": 12555
    },
    {
      "entropy": 5.563037014007568,
      "epoch": 1.2465263993648272,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.1873,
      "mean_token_accuracy": 0.16514650136232376,
      "num_tokens": 27497105.0,
      "step": 12560
    },
    {
      "entropy": 5.525526905059815,
      "epoch": 1.247022628026995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.1213,
      "mean_token_accuracy": 0.16782769709825515,
      "num_tokens": 27508157.0,
      "step": 12565
    },
    {
      "entropy": 5.518879079818726,
      "epoch": 1.2475188566891624,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.1411,
      "mean_token_accuracy": 0.16889148205518723,
      "num_tokens": 27519128.0,
      "step": 12570
    },
    {
      "entropy": 5.53438720703125,
      "epoch": 1.2480150853513299,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.1891,
      "mean_token_accuracy": 0.16969461888074874,
      "num_tokens": 27529151.0,
      "step": 12575
    },
    {
      "entropy": 5.53507628440857,
      "epoch": 1.2485113140134974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16908288598060608,
      "num_tokens": 27538822.0,
      "step": 12580
    },
    {
      "entropy": 5.565705490112305,
      "epoch": 1.249007542675665,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.1337,
      "mean_token_accuracy": 0.1720982789993286,
      "num_tokens": 27548779.0,
      "step": 12585
    },
    {
      "entropy": 5.515538549423217,
      "epoch": 1.2495037713378325,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.1339,
      "mean_token_accuracy": 0.17146315276622773,
      "num_tokens": 27559467.0,
      "step": 12590
    },
    {
      "entropy": 5.507602882385254,
      "epoch": 1.25,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1713444024324417,
      "num_tokens": 27569518.0,
      "step": 12595
    },
    {
      "entropy": 5.5363757610321045,
      "epoch": 1.2504962286621675,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.1292,
      "mean_token_accuracy": 0.16960541307926177,
      "num_tokens": 27580926.0,
      "step": 12600
    },
    {
      "entropy": 5.46859073638916,
      "epoch": 1.250992457324335,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1709494858980179,
      "num_tokens": 27591354.0,
      "step": 12605
    },
    {
      "entropy": 5.515297698974609,
      "epoch": 1.2514886859865026,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.1669,
      "mean_token_accuracy": 0.1685708522796631,
      "num_tokens": 27601950.0,
      "step": 12610
    },
    {
      "entropy": 5.519558477401733,
      "epoch": 1.2519849146486701,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.1393,
      "mean_token_accuracy": 0.17526807636022568,
      "num_tokens": 27614215.0,
      "step": 12615
    },
    {
      "entropy": 5.569565916061402,
      "epoch": 1.2524811433108376,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.1774,
      "mean_token_accuracy": 0.1671185851097107,
      "num_tokens": 27625486.0,
      "step": 12620
    },
    {
      "entropy": 5.587026119232178,
      "epoch": 1.2529773719730053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1740003064274788,
      "num_tokens": 27636028.0,
      "step": 12625
    },
    {
      "entropy": 5.573077630996704,
      "epoch": 1.2534736006351728,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.2025,
      "mean_token_accuracy": 0.16014232039451598,
      "num_tokens": 27646368.0,
      "step": 12630
    },
    {
      "entropy": 5.535755825042725,
      "epoch": 1.2539698292973402,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.1546,
      "mean_token_accuracy": 0.1649329975247383,
      "num_tokens": 27656337.0,
      "step": 12635
    },
    {
      "entropy": 5.512619876861573,
      "epoch": 1.2544660579595077,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16593048870563507,
      "num_tokens": 27666174.0,
      "step": 12640
    },
    {
      "entropy": 5.496509552001953,
      "epoch": 1.2549622866216752,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17271596640348436,
      "num_tokens": 27675908.0,
      "step": 12645
    },
    {
      "entropy": 5.581809329986572,
      "epoch": 1.2554585152838427,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.2166,
      "mean_token_accuracy": 0.16627753376960755,
      "num_tokens": 27686746.0,
      "step": 12650
    },
    {
      "entropy": 5.5514781951904295,
      "epoch": 1.2559547439460104,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.169,
      "mean_token_accuracy": 0.17870625257492065,
      "num_tokens": 27699911.0,
      "step": 12655
    },
    {
      "entropy": 5.551044654846192,
      "epoch": 1.2564509726081778,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.1219,
      "mean_token_accuracy": 0.16754639148712158,
      "num_tokens": 27709572.0,
      "step": 12660
    },
    {
      "entropy": 5.492155265808106,
      "epoch": 1.2569472012703453,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.1054,
      "mean_token_accuracy": 0.1664474055171013,
      "num_tokens": 27720357.0,
      "step": 12665
    },
    {
      "entropy": 5.531010532379151,
      "epoch": 1.257443429932513,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.18,
      "mean_token_accuracy": 0.16535889059305192,
      "num_tokens": 27731284.0,
      "step": 12670
    },
    {
      "entropy": 5.582598114013672,
      "epoch": 1.2579396585946805,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.2653,
      "mean_token_accuracy": 0.16176071017980576,
      "num_tokens": 27743619.0,
      "step": 12675
    },
    {
      "entropy": 5.539963817596435,
      "epoch": 1.258435887256848,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.1699,
      "mean_token_accuracy": 0.17338959872722626,
      "num_tokens": 27753440.0,
      "step": 12680
    },
    {
      "entropy": 5.542337322235108,
      "epoch": 1.2589321159190154,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.1949,
      "mean_token_accuracy": 0.1703419178724289,
      "num_tokens": 27764765.0,
      "step": 12685
    },
    {
      "entropy": 5.577678585052491,
      "epoch": 1.259428344581183,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000484924535331964,
      "loss": 5.2393,
      "mean_token_accuracy": 0.16041475534439087,
      "num_tokens": 27776817.0,
      "step": 12690
    },
    {
      "entropy": 5.575383472442627,
      "epoch": 1.2599245732433506,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.1812,
      "mean_token_accuracy": 0.169460666179657,
      "num_tokens": 27788276.0,
      "step": 12695
    },
    {
      "entropy": 5.449568223953247,
      "epoch": 1.260420801905518,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17595661282539368,
      "num_tokens": 27798493.0,
      "step": 12700
    },
    {
      "entropy": 5.497296380996704,
      "epoch": 1.2609170305676856,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.1477,
      "mean_token_accuracy": 0.16749653220176697,
      "num_tokens": 27809908.0,
      "step": 12705
    },
    {
      "entropy": 5.490494585037231,
      "epoch": 1.261413259229853,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.1394,
      "mean_token_accuracy": 0.17174632400274276,
      "num_tokens": 27820954.0,
      "step": 12710
    },
    {
      "entropy": 5.578171634674073,
      "epoch": 1.2619094878920207,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.2232,
      "mean_token_accuracy": 0.16557825803756715,
      "num_tokens": 27833645.0,
      "step": 12715
    },
    {
      "entropy": 5.525861501693726,
      "epoch": 1.2624057165541882,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.1087,
      "mean_token_accuracy": 0.17195527553558348,
      "num_tokens": 27844532.0,
      "step": 12720
    },
    {
      "entropy": 5.559710454940796,
      "epoch": 1.2629019452163557,
      "grad_norm": 0.90625,
      "learning_rate": 0.000484835149776263,
      "loss": 5.1969,
      "mean_token_accuracy": 0.17081396281719208,
      "num_tokens": 27855832.0,
      "step": 12725
    },
    {
      "entropy": 5.539887857437134,
      "epoch": 1.2633981738785232,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.2552,
      "mean_token_accuracy": 0.163968925178051,
      "num_tokens": 27867087.0,
      "step": 12730
    },
    {
      "entropy": 5.551989936828614,
      "epoch": 1.2638944025406906,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.1502,
      "mean_token_accuracy": 0.16953234672546386,
      "num_tokens": 27877576.0,
      "step": 12735
    },
    {
      "entropy": 5.481462669372559,
      "epoch": 1.2643906312028583,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17596686631441116,
      "num_tokens": 27888998.0,
      "step": 12740
    },
    {
      "entropy": 5.489080715179443,
      "epoch": 1.2648868598650258,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.0801,
      "mean_token_accuracy": 0.17206472009420395,
      "num_tokens": 27899611.0,
      "step": 12745
    },
    {
      "entropy": 5.512057304382324,
      "epoch": 1.2653830885271933,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1725588172674179,
      "num_tokens": 27911787.0,
      "step": 12750
    },
    {
      "entropy": 5.6380391120910645,
      "epoch": 1.265879317189361,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.2711,
      "mean_token_accuracy": 0.15513214841485023,
      "num_tokens": 27922729.0,
      "step": 12755
    },
    {
      "entropy": 5.51535325050354,
      "epoch": 1.2663755458515285,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1736914947628975,
      "num_tokens": 27934357.0,
      "step": 12760
    },
    {
      "entropy": 5.58016448020935,
      "epoch": 1.266871774513696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.1944,
      "mean_token_accuracy": 0.16450904458761215,
      "num_tokens": 27945355.0,
      "step": 12765
    },
    {
      "entropy": 5.541463232040405,
      "epoch": 1.2673680031758634,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.1552,
      "mean_token_accuracy": 0.17368416041135787,
      "num_tokens": 27956426.0,
      "step": 12770
    },
    {
      "entropy": 5.592769002914428,
      "epoch": 1.2678642318380309,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.2077,
      "mean_token_accuracy": 0.16972106397151948,
      "num_tokens": 27968431.0,
      "step": 12775
    },
    {
      "entropy": 5.587922286987305,
      "epoch": 1.2683604605001986,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.3138,
      "mean_token_accuracy": 0.16311698406934738,
      "num_tokens": 27978903.0,
      "step": 12780
    },
    {
      "entropy": 5.513173151016235,
      "epoch": 1.268856689162366,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.1122,
      "mean_token_accuracy": 0.172031532227993,
      "num_tokens": 27990791.0,
      "step": 12785
    },
    {
      "entropy": 5.538331031799316,
      "epoch": 1.2693529178245335,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.1931,
      "mean_token_accuracy": 0.17001333385705947,
      "num_tokens": 28000534.0,
      "step": 12790
    },
    {
      "entropy": 5.514766979217529,
      "epoch": 1.269849146486701,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000484655613865303,
      "loss": 5.1661,
      "mean_token_accuracy": 0.16812191307544708,
      "num_tokens": 28010573.0,
      "step": 12795
    },
    {
      "entropy": 5.5464653968811035,
      "epoch": 1.2703453751488687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.1771,
      "mean_token_accuracy": 0.16642674654722214,
      "num_tokens": 28021255.0,
      "step": 12800
    },
    {
      "entropy": 5.548657178878784,
      "epoch": 1.2708416038110362,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.2148,
      "mean_token_accuracy": 0.16340156346559526,
      "num_tokens": 28032529.0,
      "step": 12805
    },
    {
      "entropy": 5.575164127349853,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.1384,
      "mean_token_accuracy": 0.16534623503684998,
      "num_tokens": 28044225.0,
      "step": 12810
    },
    {
      "entropy": 5.478167247772217,
      "epoch": 1.2718340611353711,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.1359,
      "mean_token_accuracy": 0.17592095583677292,
      "num_tokens": 28056175.0,
      "step": 12815
    },
    {
      "entropy": 5.550609159469604,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.2484,
      "mean_token_accuracy": 0.1623075932264328,
      "num_tokens": 28067496.0,
      "step": 12820
    },
    {
      "entropy": 5.564299535751343,
      "epoch": 1.2728265184597063,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.1915,
      "mean_token_accuracy": 0.16229915469884873,
      "num_tokens": 28078140.0,
      "step": 12825
    },
    {
      "entropy": 5.52310962677002,
      "epoch": 1.2733227471218738,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.141,
      "mean_token_accuracy": 0.17125664353370668,
      "num_tokens": 28089612.0,
      "step": 12830
    },
    {
      "entropy": 5.450892734527588,
      "epoch": 1.2738189757840412,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.0319,
      "mean_token_accuracy": 0.16823037564754487,
      "num_tokens": 28100537.0,
      "step": 12835
    },
    {
      "entropy": 5.577935075759887,
      "epoch": 1.274315204446209,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.3172,
      "mean_token_accuracy": 0.16093576177954674,
      "num_tokens": 28113317.0,
      "step": 12840
    },
    {
      "entropy": 5.485065889358521,
      "epoch": 1.2748114331083764,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17268689274787902,
      "num_tokens": 28125098.0,
      "step": 12845
    },
    {
      "entropy": 5.600786018371582,
      "epoch": 1.2753076617705439,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.2341,
      "mean_token_accuracy": 0.17117495089769363,
      "num_tokens": 28135099.0,
      "step": 12850
    },
    {
      "entropy": 5.570370531082153,
      "epoch": 1.2758038904327114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.1996,
      "mean_token_accuracy": 0.1667718842625618,
      "num_tokens": 28146470.0,
      "step": 12855
    },
    {
      "entropy": 5.52872142791748,
      "epoch": 1.2763001190948788,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.1267,
      "mean_token_accuracy": 0.1706656575202942,
      "num_tokens": 28156886.0,
      "step": 12860
    },
    {
      "entropy": 5.502231121063232,
      "epoch": 1.2767963477570463,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.1172,
      "mean_token_accuracy": 0.1716299369931221,
      "num_tokens": 28166907.0,
      "step": 12865
    },
    {
      "entropy": 5.531640672683716,
      "epoch": 1.277292576419214,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.1277,
      "mean_token_accuracy": 0.17296555638313293,
      "num_tokens": 28178015.0,
      "step": 12870
    },
    {
      "entropy": 5.584946632385254,
      "epoch": 1.2777888050813815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.1844,
      "mean_token_accuracy": 0.1660143867135048,
      "num_tokens": 28188518.0,
      "step": 12875
    },
    {
      "entropy": 5.533895206451416,
      "epoch": 1.278285033743549,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.1509,
      "mean_token_accuracy": 0.16561569571495055,
      "num_tokens": 28199782.0,
      "step": 12880
    },
    {
      "entropy": 5.513859033584595,
      "epoch": 1.2787812624057167,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.1691,
      "mean_token_accuracy": 0.16753800660371782,
      "num_tokens": 28211455.0,
      "step": 12885
    },
    {
      "entropy": 5.534774112701416,
      "epoch": 1.2792774910678841,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16421716958284377,
      "num_tokens": 28222599.0,
      "step": 12890
    },
    {
      "entropy": 5.565543937683105,
      "epoch": 1.2797737197300516,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.1797,
      "mean_token_accuracy": 0.16656164526939393,
      "num_tokens": 28233129.0,
      "step": 12895
    },
    {
      "entropy": 5.5354594707489015,
      "epoch": 1.280269948392219,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.1655,
      "mean_token_accuracy": 0.166197969019413,
      "num_tokens": 28243603.0,
      "step": 12900
    },
    {
      "entropy": 5.584839534759522,
      "epoch": 1.2807661770543866,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.1754,
      "mean_token_accuracy": 0.162259042263031,
      "num_tokens": 28254695.0,
      "step": 12905
    },
    {
      "entropy": 5.567954444885254,
      "epoch": 1.2812624057165543,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.1679,
      "mean_token_accuracy": 0.16583823561668395,
      "num_tokens": 28266486.0,
      "step": 12910
    },
    {
      "entropy": 5.508651494979858,
      "epoch": 1.2817586343787217,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.1634,
      "mean_token_accuracy": 0.16874209195375442,
      "num_tokens": 28277931.0,
      "step": 12915
    },
    {
      "entropy": 5.518090105056762,
      "epoch": 1.2822548630408892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.2824,
      "mean_token_accuracy": 0.16032982468605042,
      "num_tokens": 28288807.0,
      "step": 12920
    },
    {
      "entropy": 5.5501586437225345,
      "epoch": 1.282751091703057,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1656548261642456,
      "num_tokens": 28299577.0,
      "step": 12925
    },
    {
      "entropy": 5.5841637134552,
      "epoch": 1.2832473203652244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.1884,
      "mean_token_accuracy": 0.16593901664018632,
      "num_tokens": 28310645.0,
      "step": 12930
    },
    {
      "entropy": 5.551185274124146,
      "epoch": 1.2837435490273919,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.1708,
      "mean_token_accuracy": 0.16917659640312194,
      "num_tokens": 28321634.0,
      "step": 12935
    },
    {
      "entropy": 5.558014822006226,
      "epoch": 1.2842397776895593,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.1124,
      "mean_token_accuracy": 0.16836437433958054,
      "num_tokens": 28333325.0,
      "step": 12940
    },
    {
      "entropy": 5.545534706115722,
      "epoch": 1.2847360063517268,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.1806,
      "mean_token_accuracy": 0.17124726325273515,
      "num_tokens": 28344163.0,
      "step": 12945
    },
    {
      "entropy": 5.582979917526245,
      "epoch": 1.2852322350138943,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.2382,
      "mean_token_accuracy": 0.17085533142089843,
      "num_tokens": 28355820.0,
      "step": 12950
    },
    {
      "entropy": 5.604668045043946,
      "epoch": 1.285728463676062,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.1887,
      "mean_token_accuracy": 0.16978720873594283,
      "num_tokens": 28367850.0,
      "step": 12955
    },
    {
      "entropy": 5.6112336158752445,
      "epoch": 1.2862246923382294,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.269,
      "mean_token_accuracy": 0.1582994282245636,
      "num_tokens": 28379013.0,
      "step": 12960
    },
    {
      "entropy": 5.5953107357025145,
      "epoch": 1.286720921000397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.2114,
      "mean_token_accuracy": 0.16867242157459258,
      "num_tokens": 28389649.0,
      "step": 12965
    },
    {
      "entropy": 5.448366594314575,
      "epoch": 1.2872171496625646,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17946554869413375,
      "num_tokens": 28400239.0,
      "step": 12970
    },
    {
      "entropy": 5.505300855636596,
      "epoch": 1.287713378324732,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.13,
      "mean_token_accuracy": 0.1717386171221733,
      "num_tokens": 28410504.0,
      "step": 12975
    },
    {
      "entropy": 5.518002939224243,
      "epoch": 1.2882096069868996,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.1599,
      "mean_token_accuracy": 0.16856123358011246,
      "num_tokens": 28422549.0,
      "step": 12980
    },
    {
      "entropy": 5.559256553649902,
      "epoch": 1.288705835649067,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.1597,
      "mean_token_accuracy": 0.16783968210220337,
      "num_tokens": 28433019.0,
      "step": 12985
    },
    {
      "entropy": 5.523057126998902,
      "epoch": 1.2892020643112345,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1634242281317711,
      "num_tokens": 28443814.0,
      "step": 12990
    },
    {
      "entropy": 5.5545103549957275,
      "epoch": 1.2896982929734022,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.101,
      "mean_token_accuracy": 0.1752532795071602,
      "num_tokens": 28453825.0,
      "step": 12995
    },
    {
      "entropy": 5.5747072219848635,
      "epoch": 1.2901945216355697,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.2433,
      "mean_token_accuracy": 0.16972378492355347,
      "num_tokens": 28464583.0,
      "step": 13000
    },
    {
      "entropy": 5.60496883392334,
      "epoch": 1.2906907502977372,
      "grad_norm": 1.0,
      "learning_rate": 0.000484110895617326,
      "loss": 5.2192,
      "mean_token_accuracy": 0.159957717359066,
      "num_tokens": 28475764.0,
      "step": 13005
    },
    {
      "entropy": 5.623859882354736,
      "epoch": 1.2911869789599046,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.1774,
      "mean_token_accuracy": 0.16501082628965377,
      "num_tokens": 28486451.0,
      "step": 13010
    },
    {
      "entropy": 5.565261602401733,
      "epoch": 1.2916832076220723,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.2173,
      "mean_token_accuracy": 0.1681195616722107,
      "num_tokens": 28498012.0,
      "step": 13015
    },
    {
      "entropy": 5.506037759780884,
      "epoch": 1.2921794362842398,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048407163686642,
      "loss": 5.1736,
      "mean_token_accuracy": 0.16674817800521852,
      "num_tokens": 28509058.0,
      "step": 13020
    },
    {
      "entropy": 5.5992053031921385,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000484058540246633,
      "loss": 5.2036,
      "mean_token_accuracy": 0.17202324122190477,
      "num_tokens": 28519001.0,
      "step": 13025
    },
    {
      "entropy": 5.561823511123658,
      "epoch": 1.2931718936085748,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.2001,
      "mean_token_accuracy": 0.16990649700164795,
      "num_tokens": 28529606.0,
      "step": 13030
    },
    {
      "entropy": 5.558184194564819,
      "epoch": 1.2936681222707422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.2049,
      "mean_token_accuracy": 0.16960095465183259,
      "num_tokens": 28540723.0,
      "step": 13035
    },
    {
      "entropy": 5.564636754989624,
      "epoch": 1.29416435093291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.0904,
      "mean_token_accuracy": 0.170186348259449,
      "num_tokens": 28549817.0,
      "step": 13040
    },
    {
      "entropy": 5.540435314178467,
      "epoch": 1.2946605795950774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.1797,
      "mean_token_accuracy": 0.1692092925310135,
      "num_tokens": 28560519.0,
      "step": 13045
    },
    {
      "entropy": 5.517555284500122,
      "epoch": 1.2951568082572449,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.2024,
      "mean_token_accuracy": 0.17040927559137345,
      "num_tokens": 28573234.0,
      "step": 13050
    },
    {
      "entropy": 5.5716958999633786,
      "epoch": 1.2956530369194126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.2839,
      "mean_token_accuracy": 0.15764139518141745,
      "num_tokens": 28582954.0,
      "step": 13055
    },
    {
      "entropy": 5.567650413513183,
      "epoch": 1.29614926558158,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.049,
      "mean_token_accuracy": 0.17597985714673997,
      "num_tokens": 28593704.0,
      "step": 13060
    },
    {
      "entropy": 5.525706958770752,
      "epoch": 1.2966454942437475,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.1398,
      "mean_token_accuracy": 0.16784396171569824,
      "num_tokens": 28604023.0,
      "step": 13065
    },
    {
      "entropy": 5.573412561416626,
      "epoch": 1.297141722905915,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.2343,
      "mean_token_accuracy": 0.16627237796783448,
      "num_tokens": 28614634.0,
      "step": 13070
    },
    {
      "entropy": 5.496887874603272,
      "epoch": 1.2976379515680825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000483927288971341,
      "loss": 5.1272,
      "mean_token_accuracy": 0.16839791983366012,
      "num_tokens": 28624436.0,
      "step": 13075
    },
    {
      "entropy": 5.559484338760376,
      "epoch": 1.2981341802302502,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16687052994966506,
      "num_tokens": 28636372.0,
      "step": 13080
    },
    {
      "entropy": 5.57750301361084,
      "epoch": 1.2986304088924177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16700341403484345,
      "num_tokens": 28647495.0,
      "step": 13085
    },
    {
      "entropy": 5.552853155136108,
      "epoch": 1.2991266375545851,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16853172779083253,
      "num_tokens": 28657692.0,
      "step": 13090
    },
    {
      "entropy": 5.560763597488403,
      "epoch": 1.2996228662167526,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.1984,
      "mean_token_accuracy": 0.16682953983545304,
      "num_tokens": 28668580.0,
      "step": 13095
    },
    {
      "entropy": 5.687523365020752,
      "epoch": 1.3001190948789203,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1612253099679947,
      "num_tokens": 28680274.0,
      "step": 13100
    },
    {
      "entropy": 5.5614227771759035,
      "epoch": 1.3006153235410878,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.2024,
      "mean_token_accuracy": 0.17255071699619293,
      "num_tokens": 28691572.0,
      "step": 13105
    },
    {
      "entropy": 5.575030517578125,
      "epoch": 1.3011115522032553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.2421,
      "mean_token_accuracy": 0.16237685531377793,
      "num_tokens": 28702153.0,
      "step": 13110
    },
    {
      "entropy": 5.560045576095581,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.0,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1659469038248062,
      "num_tokens": 28712626.0,
      "step": 13115
    },
    {
      "entropy": 5.542477416992187,
      "epoch": 1.3021040095275902,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.2222,
      "mean_token_accuracy": 0.16378988474607467,
      "num_tokens": 28723912.0,
      "step": 13120
    },
    {
      "entropy": 5.551003408432007,
      "epoch": 1.302600238189758,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.1298,
      "mean_token_accuracy": 0.17073115408420564,
      "num_tokens": 28734675.0,
      "step": 13125
    },
    {
      "entropy": 5.470095825195313,
      "epoch": 1.3030964668519254,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.1407,
      "mean_token_accuracy": 0.17050999850034715,
      "num_tokens": 28746470.0,
      "step": 13130
    },
    {
      "entropy": 5.5134937286376955,
      "epoch": 1.3035926955140928,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.1357,
      "mean_token_accuracy": 0.17432697862386703,
      "num_tokens": 28757825.0,
      "step": 13135
    },
    {
      "entropy": 5.575321960449219,
      "epoch": 1.3040889241762605,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.195,
      "mean_token_accuracy": 0.1664292722940445,
      "num_tokens": 28768305.0,
      "step": 13140
    },
    {
      "entropy": 5.567330932617187,
      "epoch": 1.304585152838428,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.2312,
      "mean_token_accuracy": 0.16575057059526443,
      "num_tokens": 28779294.0,
      "step": 13145
    },
    {
      "entropy": 5.522053050994873,
      "epoch": 1.3050813815005955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.1712,
      "mean_token_accuracy": 0.160385625064373,
      "num_tokens": 28789911.0,
      "step": 13150
    },
    {
      "entropy": 5.600982427597046,
      "epoch": 1.305577610162763,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.2569,
      "mean_token_accuracy": 0.16020378321409226,
      "num_tokens": 28800705.0,
      "step": 13155
    },
    {
      "entropy": 5.52875280380249,
      "epoch": 1.3060738388249304,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17439010739326477,
      "num_tokens": 28812453.0,
      "step": 13160
    },
    {
      "entropy": 5.571789169311524,
      "epoch": 1.306570067487098,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.1368,
      "mean_token_accuracy": 0.16728490442037583,
      "num_tokens": 28823257.0,
      "step": 13165
    },
    {
      "entropy": 5.5006379127502445,
      "epoch": 1.3070662961492656,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17571962922811507,
      "num_tokens": 28834344.0,
      "step": 13170
    },
    {
      "entropy": 5.519805765151977,
      "epoch": 1.307562524811433,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.13,
      "mean_token_accuracy": 0.17511018365621567,
      "num_tokens": 28844372.0,
      "step": 13175
    },
    {
      "entropy": 5.500526666641235,
      "epoch": 1.3080587534736006,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.1217,
      "mean_token_accuracy": 0.17420229762792588,
      "num_tokens": 28855384.0,
      "step": 13180
    },
    {
      "entropy": 5.513567209243774,
      "epoch": 1.3085549821357683,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.2234,
      "mean_token_accuracy": 0.16271271854639052,
      "num_tokens": 28865170.0,
      "step": 13185
    },
    {
      "entropy": 5.568775272369384,
      "epoch": 1.3090512107979357,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.2243,
      "mean_token_accuracy": 0.16743958443403245,
      "num_tokens": 28876681.0,
      "step": 13190
    },
    {
      "entropy": 5.613617706298828,
      "epoch": 1.3095474394601032,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.2226,
      "mean_token_accuracy": 0.17012421786785126,
      "num_tokens": 28887033.0,
      "step": 13195
    },
    {
      "entropy": 5.569585704803467,
      "epoch": 1.3100436681222707,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1673176497220993,
      "num_tokens": 28898009.0,
      "step": 13200
    },
    {
      "entropy": 5.5226109504699705,
      "epoch": 1.3105398967844382,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.1525,
      "mean_token_accuracy": 0.17193772047758102,
      "num_tokens": 28909261.0,
      "step": 13205
    },
    {
      "entropy": 5.566798448562622,
      "epoch": 1.3110361254466059,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16502459943294526,
      "num_tokens": 28920789.0,
      "step": 13210
    },
    {
      "entropy": 5.597118520736695,
      "epoch": 1.3115323541087733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.2142,
      "mean_token_accuracy": 0.16848329305648804,
      "num_tokens": 28931964.0,
      "step": 13215
    },
    {
      "entropy": 5.52571291923523,
      "epoch": 1.3120285827709408,
      "grad_norm": 0.875,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17093655318021775,
      "num_tokens": 28942961.0,
      "step": 13220
    },
    {
      "entropy": 5.556482553482056,
      "epoch": 1.3125248114331085,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.1977,
      "mean_token_accuracy": 0.17444772124290467,
      "num_tokens": 28953449.0,
      "step": 13225
    },
    {
      "entropy": 5.555399942398071,
      "epoch": 1.313021040095276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.1923,
      "mean_token_accuracy": 0.1689928114414215,
      "num_tokens": 28963261.0,
      "step": 13230
    },
    {
      "entropy": 5.5225386142730715,
      "epoch": 1.3135172687574435,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.2211,
      "mean_token_accuracy": 0.1661134511232376,
      "num_tokens": 28973611.0,
      "step": 13235
    },
    {
      "entropy": 5.602791023254395,
      "epoch": 1.314013497419611,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.2636,
      "mean_token_accuracy": 0.1636491000652313,
      "num_tokens": 28983502.0,
      "step": 13240
    },
    {
      "entropy": 5.629878425598145,
      "epoch": 1.3145097260817784,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.2366,
      "mean_token_accuracy": 0.16701366007328033,
      "num_tokens": 28994586.0,
      "step": 13245
    },
    {
      "entropy": 5.569770574569702,
      "epoch": 1.3150059547439459,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.135,
      "mean_token_accuracy": 0.16785843372344972,
      "num_tokens": 29005316.0,
      "step": 13250
    },
    {
      "entropy": 5.500804376602173,
      "epoch": 1.3155021834061136,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.1562,
      "mean_token_accuracy": 0.17088977992534637,
      "num_tokens": 29016007.0,
      "step": 13255
    },
    {
      "entropy": 5.554137134552002,
      "epoch": 1.315998412068281,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1614915430545807,
      "num_tokens": 29025256.0,
      "step": 13260
    },
    {
      "entropy": 5.589080476760865,
      "epoch": 1.3164946407304485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.2147,
      "mean_token_accuracy": 0.16728618741035461,
      "num_tokens": 29035271.0,
      "step": 13265
    },
    {
      "entropy": 5.6102643489837645,
      "epoch": 1.3169908693926162,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.2226,
      "mean_token_accuracy": 0.1683186650276184,
      "num_tokens": 29045151.0,
      "step": 13270
    },
    {
      "entropy": 5.534548044204712,
      "epoch": 1.3174870980547837,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.1162,
      "mean_token_accuracy": 0.1743042379617691,
      "num_tokens": 29056346.0,
      "step": 13275
    },
    {
      "entropy": 5.506335496902466,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.0911,
      "mean_token_accuracy": 0.17658344507217408,
      "num_tokens": 29065609.0,
      "step": 13280
    },
    {
      "entropy": 5.513218641281128,
      "epoch": 1.3184795553791187,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.1727,
      "mean_token_accuracy": 0.1704512983560562,
      "num_tokens": 29075308.0,
      "step": 13285
    },
    {
      "entropy": 5.619879961013794,
      "epoch": 1.3189757840412861,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.1967,
      "mean_token_accuracy": 0.16141437739133835,
      "num_tokens": 29086488.0,
      "step": 13290
    },
    {
      "entropy": 5.4979170799255375,
      "epoch": 1.3194720127034538,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.124,
      "mean_token_accuracy": 0.16508664935827255,
      "num_tokens": 29098502.0,
      "step": 13295
    },
    {
      "entropy": 5.485031270980835,
      "epoch": 1.3199682413656213,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.1016,
      "mean_token_accuracy": 0.17273401319980622,
      "num_tokens": 29108991.0,
      "step": 13300
    },
    {
      "entropy": 5.5234699726104735,
      "epoch": 1.3204644700277888,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.1733,
      "mean_token_accuracy": 0.16842518746852875,
      "num_tokens": 29119436.0,
      "step": 13305
    },
    {
      "entropy": 5.553693675994873,
      "epoch": 1.3209606986899565,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.2077,
      "mean_token_accuracy": 0.16521011739969255,
      "num_tokens": 29130627.0,
      "step": 13310
    },
    {
      "entropy": 5.486120510101318,
      "epoch": 1.321456927352124,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.1216,
      "mean_token_accuracy": 0.16930748373270035,
      "num_tokens": 29141386.0,
      "step": 13315
    },
    {
      "entropy": 5.549582481384277,
      "epoch": 1.3219531560142914,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16977623850107193,
      "num_tokens": 29151562.0,
      "step": 13320
    },
    {
      "entropy": 5.545405149459839,
      "epoch": 1.322449384676459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.139,
      "mean_token_accuracy": 0.17153622955083847,
      "num_tokens": 29161571.0,
      "step": 13325
    },
    {
      "entropy": 5.577347040176392,
      "epoch": 1.3229456133386264,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.1541,
      "mean_token_accuracy": 0.170501272380352,
      "num_tokens": 29172221.0,
      "step": 13330
    },
    {
      "entropy": 5.527881669998169,
      "epoch": 1.3234418420007938,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.1828,
      "mean_token_accuracy": 0.168036986887455,
      "num_tokens": 29182434.0,
      "step": 13335
    },
    {
      "entropy": 5.512160539627075,
      "epoch": 1.3239380706629615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17696915566921234,
      "num_tokens": 29191614.0,
      "step": 13340
    },
    {
      "entropy": 5.524117660522461,
      "epoch": 1.324434299325129,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.0799,
      "mean_token_accuracy": 0.17533450871706008,
      "num_tokens": 29203105.0,
      "step": 13345
    },
    {
      "entropy": 5.535234212875366,
      "epoch": 1.3249305279872965,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.1382,
      "mean_token_accuracy": 0.17198466360569,
      "num_tokens": 29213074.0,
      "step": 13350
    },
    {
      "entropy": 5.50702223777771,
      "epoch": 1.3254267566494642,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.1902,
      "mean_token_accuracy": 0.17295391261577606,
      "num_tokens": 29223421.0,
      "step": 13355
    },
    {
      "entropy": 5.50650987625122,
      "epoch": 1.3259229853116317,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.0847,
      "mean_token_accuracy": 0.17830655127763748,
      "num_tokens": 29233330.0,
      "step": 13360
    },
    {
      "entropy": 5.552057456970215,
      "epoch": 1.3264192139737991,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1644572377204895,
      "num_tokens": 29244679.0,
      "step": 13365
    },
    {
      "entropy": 5.561604166030884,
      "epoch": 1.3269154426359666,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17359780222177507,
      "num_tokens": 29254840.0,
      "step": 13370
    },
    {
      "entropy": 5.533254766464234,
      "epoch": 1.327411671298134,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.1191,
      "mean_token_accuracy": 0.16333618760108948,
      "num_tokens": 29265794.0,
      "step": 13375
    },
    {
      "entropy": 5.569819450378418,
      "epoch": 1.3279078999603018,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.1384,
      "mean_token_accuracy": 0.17112235575914383,
      "num_tokens": 29275675.0,
      "step": 13380
    },
    {
      "entropy": 5.5448524951934814,
      "epoch": 1.3284041286224693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483101979887967,
      "loss": 5.2117,
      "mean_token_accuracy": 0.1674923598766327,
      "num_tokens": 29286809.0,
      "step": 13385
    },
    {
      "entropy": 5.56242995262146,
      "epoch": 1.3289003572846367,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17136719822883606,
      "num_tokens": 29297097.0,
      "step": 13390
    },
    {
      "entropy": 5.525705099105835,
      "epoch": 1.3293965859468042,
      "grad_norm": 0.84375,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.1661,
      "mean_token_accuracy": 0.16873336881399154,
      "num_tokens": 29309148.0,
      "step": 13395
    },
    {
      "entropy": 5.6179179668426515,
      "epoch": 1.329892814608972,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.2896,
      "mean_token_accuracy": 0.16274576038122177,
      "num_tokens": 29320461.0,
      "step": 13400
    },
    {
      "entropy": 5.524901342391968,
      "epoch": 1.3303890432711394,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17343089133501052,
      "num_tokens": 29330442.0,
      "step": 13405
    },
    {
      "entropy": 5.445257329940796,
      "epoch": 1.3308852719333069,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1734909877181053,
      "num_tokens": 29340401.0,
      "step": 13410
    },
    {
      "entropy": 5.551399612426758,
      "epoch": 1.3313815005954743,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.2174,
      "mean_token_accuracy": 0.1672436848282814,
      "num_tokens": 29351481.0,
      "step": 13415
    },
    {
      "entropy": 5.604001712799072,
      "epoch": 1.3318777292576418,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.1895,
      "mean_token_accuracy": 0.16770726293325425,
      "num_tokens": 29361642.0,
      "step": 13420
    },
    {
      "entropy": 5.543595314025879,
      "epoch": 1.3323739579198095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.145,
      "mean_token_accuracy": 0.16316841244697572,
      "num_tokens": 29373210.0,
      "step": 13425
    },
    {
      "entropy": 5.538039302825927,
      "epoch": 1.332870186581977,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.1597,
      "mean_token_accuracy": 0.16898218989372255,
      "num_tokens": 29382772.0,
      "step": 13430
    },
    {
      "entropy": 5.578719615936279,
      "epoch": 1.3333664152441445,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.1922,
      "mean_token_accuracy": 0.1643725261092186,
      "num_tokens": 29393909.0,
      "step": 13435
    },
    {
      "entropy": 5.558147192001343,
      "epoch": 1.3338626439063122,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.1674,
      "mean_token_accuracy": 0.16847074329853057,
      "num_tokens": 29406046.0,
      "step": 13440
    },
    {
      "entropy": 5.5348169803619385,
      "epoch": 1.3343588725684796,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.1475,
      "mean_token_accuracy": 0.17249502539634703,
      "num_tokens": 29415785.0,
      "step": 13445
    },
    {
      "entropy": 5.472344970703125,
      "epoch": 1.334855101230647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17436260730028152,
      "num_tokens": 29426083.0,
      "step": 13450
    },
    {
      "entropy": 5.4834678173065186,
      "epoch": 1.3353513298928146,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.127,
      "mean_token_accuracy": 0.1682866632938385,
      "num_tokens": 29437342.0,
      "step": 13455
    },
    {
      "entropy": 5.556680011749267,
      "epoch": 1.335847558554982,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.1525,
      "mean_token_accuracy": 0.17037528455257417,
      "num_tokens": 29448973.0,
      "step": 13460
    },
    {
      "entropy": 5.578584623336792,
      "epoch": 1.3363437872171495,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.1871,
      "mean_token_accuracy": 0.1718019500374794,
      "num_tokens": 29460308.0,
      "step": 13465
    },
    {
      "entropy": 5.595936012268067,
      "epoch": 1.3368400158793172,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.1723,
      "mean_token_accuracy": 0.16832878738641738,
      "num_tokens": 29470245.0,
      "step": 13470
    },
    {
      "entropy": 5.529517126083374,
      "epoch": 1.3373362445414847,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.1816,
      "mean_token_accuracy": 0.16819744408130646,
      "num_tokens": 29480072.0,
      "step": 13475
    },
    {
      "entropy": 5.5282127380371096,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.1278,
      "mean_token_accuracy": 0.1712896019220352,
      "num_tokens": 29489724.0,
      "step": 13480
    },
    {
      "entropy": 5.554466485977173,
      "epoch": 1.3383287018658199,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.1379,
      "mean_token_accuracy": 0.17158496826887132,
      "num_tokens": 29500845.0,
      "step": 13485
    },
    {
      "entropy": 5.562434959411621,
      "epoch": 1.3388249305279873,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1709467276930809,
      "num_tokens": 29511721.0,
      "step": 13490
    },
    {
      "entropy": 5.552387237548828,
      "epoch": 1.3393211591901548,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.1197,
      "mean_token_accuracy": 0.16905124336481095,
      "num_tokens": 29522236.0,
      "step": 13495
    },
    {
      "entropy": 5.5260392189025875,
      "epoch": 1.3398173878523223,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16730691194534303,
      "num_tokens": 29533929.0,
      "step": 13500
    },
    {
      "entropy": 5.494289445877075,
      "epoch": 1.3403136165144898,
      "grad_norm": 0.875,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.1338,
      "mean_token_accuracy": 0.16802367120981215,
      "num_tokens": 29544805.0,
      "step": 13505
    },
    {
      "entropy": 5.613139343261719,
      "epoch": 1.3408098451766575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.2545,
      "mean_token_accuracy": 0.16205059736967087,
      "num_tokens": 29555334.0,
      "step": 13510
    },
    {
      "entropy": 5.540959310531616,
      "epoch": 1.341306073838825,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1677289918065071,
      "num_tokens": 29567681.0,
      "step": 13515
    },
    {
      "entropy": 5.603194236755371,
      "epoch": 1.3418023025009924,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.1961,
      "mean_token_accuracy": 0.16010899990797042,
      "num_tokens": 29579626.0,
      "step": 13520
    },
    {
      "entropy": 5.486583185195923,
      "epoch": 1.3422985311631601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18082302659749985,
      "num_tokens": 29590312.0,
      "step": 13525
    },
    {
      "entropy": 5.501360034942627,
      "epoch": 1.3427947598253276,
      "grad_norm": 1.0,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.1333,
      "mean_token_accuracy": 0.1746755436062813,
      "num_tokens": 29599846.0,
      "step": 13530
    },
    {
      "entropy": 5.515166854858398,
      "epoch": 1.343290988487495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.1576,
      "mean_token_accuracy": 0.16754832863807678,
      "num_tokens": 29610777.0,
      "step": 13535
    },
    {
      "entropy": 5.578885698318482,
      "epoch": 1.3437872171496625,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.2356,
      "mean_token_accuracy": 0.1633896306157112,
      "num_tokens": 29623622.0,
      "step": 13540
    },
    {
      "entropy": 5.600388145446777,
      "epoch": 1.34428344581183,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1657892182469368,
      "num_tokens": 29635390.0,
      "step": 13545
    },
    {
      "entropy": 5.6239570617675785,
      "epoch": 1.3447796744739975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.2529,
      "mean_token_accuracy": 0.1677103668451309,
      "num_tokens": 29645443.0,
      "step": 13550
    },
    {
      "entropy": 5.761407089233399,
      "epoch": 1.3452759031361652,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.3989,
      "mean_token_accuracy": 0.15726639330387115,
      "num_tokens": 29658027.0,
      "step": 13555
    },
    {
      "entropy": 5.5045383930206295,
      "epoch": 1.3457721317983327,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.1233,
      "mean_token_accuracy": 0.16963659077882767,
      "num_tokens": 29668809.0,
      "step": 13560
    },
    {
      "entropy": 5.543357276916504,
      "epoch": 1.3462683604605001,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.253,
      "mean_token_accuracy": 0.16238752603530884,
      "num_tokens": 29680036.0,
      "step": 13565
    },
    {
      "entropy": 5.543824958801269,
      "epoch": 1.3467645891226678,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000482600000263264,
      "loss": 5.1745,
      "mean_token_accuracy": 0.16550751626491547,
      "num_tokens": 29692014.0,
      "step": 13570
    },
    {
      "entropy": 5.533044242858887,
      "epoch": 1.3472608177848353,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.1851,
      "mean_token_accuracy": 0.16674253344535828,
      "num_tokens": 29703168.0,
      "step": 13575
    },
    {
      "entropy": 5.53781909942627,
      "epoch": 1.3477570464470028,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.1745,
      "mean_token_accuracy": 0.16644856035709382,
      "num_tokens": 29715532.0,
      "step": 13580
    },
    {
      "entropy": 5.548588705062866,
      "epoch": 1.3482532751091703,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.0788,
      "mean_token_accuracy": 0.1703122928738594,
      "num_tokens": 29725697.0,
      "step": 13585
    },
    {
      "entropy": 5.5659037113189695,
      "epoch": 1.3487495037713377,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.2082,
      "mean_token_accuracy": 0.16039585322141647,
      "num_tokens": 29736870.0,
      "step": 13590
    },
    {
      "entropy": 5.576161479949951,
      "epoch": 1.3492457324335054,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.2887,
      "mean_token_accuracy": 0.16417292580008508,
      "num_tokens": 29747504.0,
      "step": 13595
    },
    {
      "entropy": 5.577278327941895,
      "epoch": 1.349741961095673,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.1545,
      "mean_token_accuracy": 0.16756544262170792,
      "num_tokens": 29759182.0,
      "step": 13600
    },
    {
      "entropy": 5.547258758544922,
      "epoch": 1.3502381897578404,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17269942164421082,
      "num_tokens": 29770430.0,
      "step": 13605
    },
    {
      "entropy": 5.572694158554077,
      "epoch": 1.350734418420008,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.2154,
      "mean_token_accuracy": 0.16683990210294725,
      "num_tokens": 29782196.0,
      "step": 13610
    },
    {
      "entropy": 5.551163673400879,
      "epoch": 1.3512306470821756,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.1171,
      "mean_token_accuracy": 0.1730942040681839,
      "num_tokens": 29792739.0,
      "step": 13615
    },
    {
      "entropy": 5.5702437400817875,
      "epoch": 1.351726875744343,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1714291051030159,
      "num_tokens": 29804384.0,
      "step": 13620
    },
    {
      "entropy": 5.558650398254395,
      "epoch": 1.3522231044065105,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.2294,
      "mean_token_accuracy": 0.1652764141559601,
      "num_tokens": 29815817.0,
      "step": 13625
    },
    {
      "entropy": 5.565615797042847,
      "epoch": 1.352719333068678,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.3057,
      "mean_token_accuracy": 0.15636627972126008,
      "num_tokens": 29827811.0,
      "step": 13630
    },
    {
      "entropy": 5.6078346252441404,
      "epoch": 1.3532155617308455,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.076,
      "mean_token_accuracy": 0.17374033629894256,
      "num_tokens": 29836776.0,
      "step": 13635
    },
    {
      "entropy": 5.478469324111939,
      "epoch": 1.3537117903930131,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1762598305940628,
      "num_tokens": 29847127.0,
      "step": 13640
    },
    {
      "entropy": 5.485668659210205,
      "epoch": 1.3542080190551806,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.1504,
      "mean_token_accuracy": 0.16748057454824447,
      "num_tokens": 29859056.0,
      "step": 13645
    },
    {
      "entropy": 5.5766771793365475,
      "epoch": 1.354704247717348,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17200551629066468,
      "num_tokens": 29868468.0,
      "step": 13650
    },
    {
      "entropy": 5.545187187194824,
      "epoch": 1.3552004763795158,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.18,
      "mean_token_accuracy": 0.17424165457487106,
      "num_tokens": 29878653.0,
      "step": 13655
    },
    {
      "entropy": 5.488528680801392,
      "epoch": 1.3556967050416833,
      "grad_norm": 0.9375,
      "learning_rate": 0.000482353244626462,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17318879663944245,
      "num_tokens": 29889298.0,
      "step": 13660
    },
    {
      "entropy": 5.530117321014404,
      "epoch": 1.3561929337038507,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.1524,
      "mean_token_accuracy": 0.16843051761388778,
      "num_tokens": 29900712.0,
      "step": 13665
    },
    {
      "entropy": 5.554743909835816,
      "epoch": 1.3566891623660182,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.1113,
      "mean_token_accuracy": 0.175951187312603,
      "num_tokens": 29910528.0,
      "step": 13670
    },
    {
      "entropy": 5.599310111999512,
      "epoch": 1.3571853910281857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.1518,
      "mean_token_accuracy": 0.17019815593957902,
      "num_tokens": 29920832.0,
      "step": 13675
    },
    {
      "entropy": 5.511434650421142,
      "epoch": 1.3576816196903534,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16991495043039323,
      "num_tokens": 29930692.0,
      "step": 13680
    },
    {
      "entropy": 5.602510070800781,
      "epoch": 1.3581778483525209,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16615368276834488,
      "num_tokens": 29941381.0,
      "step": 13685
    },
    {
      "entropy": 5.644968795776367,
      "epoch": 1.3586740770146883,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.1979,
      "mean_token_accuracy": 0.16925547569990157,
      "num_tokens": 29952297.0,
      "step": 13690
    },
    {
      "entropy": 5.525482320785523,
      "epoch": 1.3591703056768558,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.1292,
      "mean_token_accuracy": 0.17372094988822936,
      "num_tokens": 29963011.0,
      "step": 13695
    },
    {
      "entropy": 5.49485936164856,
      "epoch": 1.3596665343390235,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.1321,
      "mean_token_accuracy": 0.16794464290142058,
      "num_tokens": 29975575.0,
      "step": 13700
    },
    {
      "entropy": 5.60940899848938,
      "epoch": 1.360162763001191,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.2558,
      "mean_token_accuracy": 0.16936997920274735,
      "num_tokens": 29987088.0,
      "step": 13705
    },
    {
      "entropy": 5.50831937789917,
      "epoch": 1.3606589916633585,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.0415,
      "mean_token_accuracy": 0.17687755078077316,
      "num_tokens": 29996471.0,
      "step": 13710
    },
    {
      "entropy": 5.572545385360717,
      "epoch": 1.361155220325526,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.2621,
      "mean_token_accuracy": 0.16609625220298768,
      "num_tokens": 30007342.0,
      "step": 13715
    },
    {
      "entropy": 5.558862209320068,
      "epoch": 1.3616514489876934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000482187815201822,
      "loss": 5.1396,
      "mean_token_accuracy": 0.16971559226512908,
      "num_tokens": 30018043.0,
      "step": 13720
    },
    {
      "entropy": 5.474953269958496,
      "epoch": 1.362147677649861,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17067516446113587,
      "num_tokens": 30029033.0,
      "step": 13725
    },
    {
      "entropy": 5.527052974700927,
      "epoch": 1.3626439063120286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.1574,
      "mean_token_accuracy": 0.1745412230491638,
      "num_tokens": 30039460.0,
      "step": 13730
    },
    {
      "entropy": 5.521802568435669,
      "epoch": 1.363140134974196,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.142,
      "mean_token_accuracy": 0.16839467585086823,
      "num_tokens": 30049337.0,
      "step": 13735
    },
    {
      "entropy": 5.516915512084961,
      "epoch": 1.3636363636363638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1666078194975853,
      "num_tokens": 30060852.0,
      "step": 13740
    },
    {
      "entropy": 5.563281679153443,
      "epoch": 1.3641325922985312,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.1513,
      "mean_token_accuracy": 0.17537148743867875,
      "num_tokens": 30071202.0,
      "step": 13745
    },
    {
      "entropy": 5.575348997116089,
      "epoch": 1.3646288209606987,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.1607,
      "mean_token_accuracy": 0.17102163136005402,
      "num_tokens": 30082523.0,
      "step": 13750
    },
    {
      "entropy": 5.543136882781982,
      "epoch": 1.3651250496228662,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.1893,
      "mean_token_accuracy": 0.172026889026165,
      "num_tokens": 30093521.0,
      "step": 13755
    },
    {
      "entropy": 5.569887495040893,
      "epoch": 1.3656212782850337,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.1883,
      "mean_token_accuracy": 0.16554393023252487,
      "num_tokens": 30104085.0,
      "step": 13760
    },
    {
      "entropy": 5.541377067565918,
      "epoch": 1.3661175069472011,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.0904,
      "mean_token_accuracy": 0.1783705472946167,
      "num_tokens": 30114566.0,
      "step": 13765
    },
    {
      "entropy": 5.55091028213501,
      "epoch": 1.3666137356093688,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16803389936685562,
      "num_tokens": 30124643.0,
      "step": 13770
    },
    {
      "entropy": 5.538756561279297,
      "epoch": 1.3671099642715363,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.1517,
      "mean_token_accuracy": 0.17117187678813933,
      "num_tokens": 30135657.0,
      "step": 13775
    },
    {
      "entropy": 5.557026624679565,
      "epoch": 1.3676061929337038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.1195,
      "mean_token_accuracy": 0.17205474823713302,
      "num_tokens": 30146751.0,
      "step": 13780
    },
    {
      "entropy": 5.548298597335815,
      "epoch": 1.3681024215958715,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.2178,
      "mean_token_accuracy": 0.15964044332504274,
      "num_tokens": 30157854.0,
      "step": 13785
    },
    {
      "entropy": 5.581375598907471,
      "epoch": 1.368598650258039,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16787209063768388,
      "num_tokens": 30169726.0,
      "step": 13790
    },
    {
      "entropy": 5.560645055770874,
      "epoch": 1.3690948789202064,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.234,
      "mean_token_accuracy": 0.17157268524169922,
      "num_tokens": 30182006.0,
      "step": 13795
    },
    {
      "entropy": 5.548013305664062,
      "epoch": 1.369591107582374,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1594647780060768,
      "num_tokens": 30192975.0,
      "step": 13800
    },
    {
      "entropy": 5.629910182952881,
      "epoch": 1.3700873362445414,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.2629,
      "mean_token_accuracy": 0.16525213420391083,
      "num_tokens": 30205742.0,
      "step": 13805
    },
    {
      "entropy": 5.5705420017242435,
      "epoch": 1.370583564906709,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.1501,
      "mean_token_accuracy": 0.1765996500849724,
      "num_tokens": 30215850.0,
      "step": 13810
    },
    {
      "entropy": 5.522428417205811,
      "epoch": 1.3710797935688765,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.1143,
      "mean_token_accuracy": 0.1652095079421997,
      "num_tokens": 30227206.0,
      "step": 13815
    },
    {
      "entropy": 5.479555988311768,
      "epoch": 1.371576022231044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17728174924850465,
      "num_tokens": 30238946.0,
      "step": 13820
    },
    {
      "entropy": 5.5256468772888185,
      "epoch": 1.3720722508932117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.1744,
      "mean_token_accuracy": 0.16606214344501496,
      "num_tokens": 30249381.0,
      "step": 13825
    },
    {
      "entropy": 5.550516843795776,
      "epoch": 1.3725684795553792,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.1958,
      "mean_token_accuracy": 0.17347440719604493,
      "num_tokens": 30259877.0,
      "step": 13830
    },
    {
      "entropy": 5.553769207000732,
      "epoch": 1.3730647082175467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.2186,
      "mean_token_accuracy": 0.15944106131792068,
      "num_tokens": 30270401.0,
      "step": 13835
    },
    {
      "entropy": 5.62949161529541,
      "epoch": 1.3735609368797141,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.0996,
      "mean_token_accuracy": 0.18166178166866304,
      "num_tokens": 30281173.0,
      "step": 13840
    },
    {
      "entropy": 5.539551830291748,
      "epoch": 1.3740571655418816,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17502882480621337,
      "num_tokens": 30291792.0,
      "step": 13845
    },
    {
      "entropy": 5.598294830322265,
      "epoch": 1.374553394204049,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.1939,
      "mean_token_accuracy": 0.17207622081041335,
      "num_tokens": 30302830.0,
      "step": 13850
    },
    {
      "entropy": 5.559185647964478,
      "epoch": 1.3750496228662168,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.1505,
      "mean_token_accuracy": 0.16774093210697175,
      "num_tokens": 30314888.0,
      "step": 13855
    },
    {
      "entropy": 5.540045118331909,
      "epoch": 1.3755458515283843,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.165,
      "mean_token_accuracy": 0.17189027816057206,
      "num_tokens": 30325324.0,
      "step": 13860
    },
    {
      "entropy": 5.506457042694092,
      "epoch": 1.3760420801905517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.1382,
      "mean_token_accuracy": 0.17098618447780609,
      "num_tokens": 30336012.0,
      "step": 13865
    },
    {
      "entropy": 5.638138198852539,
      "epoch": 1.3765383088527194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.3226,
      "mean_token_accuracy": 0.15983682572841645,
      "num_tokens": 30346983.0,
      "step": 13870
    },
    {
      "entropy": 5.584801149368286,
      "epoch": 1.377034537514887,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.1133,
      "mean_token_accuracy": 0.17557553797960282,
      "num_tokens": 30357522.0,
      "step": 13875
    },
    {
      "entropy": 5.541017246246338,
      "epoch": 1.3775307661770544,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.1386,
      "mean_token_accuracy": 0.16518948078155518,
      "num_tokens": 30368778.0,
      "step": 13880
    },
    {
      "entropy": 5.536572265625,
      "epoch": 1.3780269948392219,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.1763,
      "mean_token_accuracy": 0.1727789282798767,
      "num_tokens": 30379651.0,
      "step": 13885
    },
    {
      "entropy": 5.564053153991699,
      "epoch": 1.3785232235013893,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.1753,
      "mean_token_accuracy": 0.16870670318603515,
      "num_tokens": 30391288.0,
      "step": 13890
    },
    {
      "entropy": 5.550861644744873,
      "epoch": 1.379019452163557,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16775654554367064,
      "num_tokens": 30400672.0,
      "step": 13895
    },
    {
      "entropy": 5.591609001159668,
      "epoch": 1.3795156808257245,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.2297,
      "mean_token_accuracy": 0.16892174482345582,
      "num_tokens": 30411953.0,
      "step": 13900
    },
    {
      "entropy": 5.618505477905273,
      "epoch": 1.380011909487892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.2558,
      "mean_token_accuracy": 0.16382078528404237,
      "num_tokens": 30422813.0,
      "step": 13905
    },
    {
      "entropy": 5.491104030609131,
      "epoch": 1.3805081381500597,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17484752237796783,
      "num_tokens": 30434088.0,
      "step": 13910
    },
    {
      "entropy": 5.589675951004028,
      "epoch": 1.3810043668122272,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.2678,
      "mean_token_accuracy": 0.16643016412854195,
      "num_tokens": 30446167.0,
      "step": 13915
    },
    {
      "entropy": 5.532016563415527,
      "epoch": 1.3815005954743946,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.0437,
      "mean_token_accuracy": 0.17747886776924132,
      "num_tokens": 30456713.0,
      "step": 13920
    },
    {
      "entropy": 5.539466667175293,
      "epoch": 1.381996824136562,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.1293,
      "mean_token_accuracy": 0.16992588341236115,
      "num_tokens": 30468183.0,
      "step": 13925
    },
    {
      "entropy": 5.519047260284424,
      "epoch": 1.3824930527987296,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.1093,
      "mean_token_accuracy": 0.17551571279764175,
      "num_tokens": 30478545.0,
      "step": 13930
    },
    {
      "entropy": 5.541820478439331,
      "epoch": 1.382989281460897,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.2423,
      "mean_token_accuracy": 0.16112829595804215,
      "num_tokens": 30490843.0,
      "step": 13935
    },
    {
      "entropy": 5.605935716629029,
      "epoch": 1.3834855101230648,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.209,
      "mean_token_accuracy": 0.16607875674962996,
      "num_tokens": 30502301.0,
      "step": 13940
    },
    {
      "entropy": 5.5813775062561035,
      "epoch": 1.3839817387852322,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.248,
      "mean_token_accuracy": 0.1634121522307396,
      "num_tokens": 30513842.0,
      "step": 13945
    },
    {
      "entropy": 5.6588640213012695,
      "epoch": 1.3844779674473997,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.2235,
      "mean_token_accuracy": 0.16514208614826204,
      "num_tokens": 30526867.0,
      "step": 13950
    },
    {
      "entropy": 5.528075265884399,
      "epoch": 1.3849741961095674,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.2476,
      "mean_token_accuracy": 0.16353198289871215,
      "num_tokens": 30539493.0,
      "step": 13955
    },
    {
      "entropy": 5.5318748474121096,
      "epoch": 1.3854704247717349,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.1902,
      "mean_token_accuracy": 0.16850481182336807,
      "num_tokens": 30550978.0,
      "step": 13960
    },
    {
      "entropy": 5.601880073547363,
      "epoch": 1.3859666534339024,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1670283183455467,
      "num_tokens": 30561873.0,
      "step": 13965
    },
    {
      "entropy": 5.550959300994873,
      "epoch": 1.3864628820960698,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.0941,
      "mean_token_accuracy": 0.17828556448221206,
      "num_tokens": 30573053.0,
      "step": 13970
    },
    {
      "entropy": 5.507180690765381,
      "epoch": 1.3869591107582373,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.0781,
      "mean_token_accuracy": 0.177254880964756,
      "num_tokens": 30583455.0,
      "step": 13975
    },
    {
      "entropy": 5.5062010288238525,
      "epoch": 1.387455339420405,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.1583,
      "mean_token_accuracy": 0.1736188679933548,
      "num_tokens": 30593908.0,
      "step": 13980
    },
    {
      "entropy": 5.528103065490723,
      "epoch": 1.3879515680825725,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.1472,
      "mean_token_accuracy": 0.16579075753688813,
      "num_tokens": 30604870.0,
      "step": 13985
    },
    {
      "entropy": 5.555803012847901,
      "epoch": 1.38844779674474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17214082330465316,
      "num_tokens": 30615048.0,
      "step": 13990
    },
    {
      "entropy": 5.576190042495727,
      "epoch": 1.3889440254069074,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.208,
      "mean_token_accuracy": 0.1675828292965889,
      "num_tokens": 30626652.0,
      "step": 13995
    },
    {
      "entropy": 5.5062154769897464,
      "epoch": 1.3894402540690751,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16833097338676453,
      "num_tokens": 30637890.0,
      "step": 14000
    },
    {
      "entropy": 5.592779636383057,
      "epoch": 1.3899364827312426,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.1712,
      "mean_token_accuracy": 0.16752881109714507,
      "num_tokens": 30649267.0,
      "step": 14005
    },
    {
      "entropy": 5.6088659286499025,
      "epoch": 1.39043271139341,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.1677,
      "mean_token_accuracy": 0.17113473862409592,
      "num_tokens": 30661003.0,
      "step": 14010
    },
    {
      "entropy": 5.51440396308899,
      "epoch": 1.3909289400555775,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.1365,
      "mean_token_accuracy": 0.1742216929793358,
      "num_tokens": 30671895.0,
      "step": 14015
    },
    {
      "entropy": 5.529660272598266,
      "epoch": 1.391425168717745,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.1192,
      "mean_token_accuracy": 0.16932023614645003,
      "num_tokens": 30683246.0,
      "step": 14020
    },
    {
      "entropy": 5.645393371582031,
      "epoch": 1.3919213973799127,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.2726,
      "mean_token_accuracy": 0.16719429492950438,
      "num_tokens": 30694078.0,
      "step": 14025
    },
    {
      "entropy": 5.569662284851074,
      "epoch": 1.3924176260420802,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.1961,
      "mean_token_accuracy": 0.16607121378183365,
      "num_tokens": 30705270.0,
      "step": 14030
    },
    {
      "entropy": 5.527068710327148,
      "epoch": 1.3929138547042477,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.1282,
      "mean_token_accuracy": 0.16805757731199264,
      "num_tokens": 30716482.0,
      "step": 14035
    },
    {
      "entropy": 5.552965450286865,
      "epoch": 1.3934100833664154,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000481293044312513,
      "loss": 5.1731,
      "mean_token_accuracy": 0.1602078139781952,
      "num_tokens": 30727257.0,
      "step": 14040
    },
    {
      "entropy": 5.526531982421875,
      "epoch": 1.3939063120285828,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.1494,
      "mean_token_accuracy": 0.16963769495487213,
      "num_tokens": 30739100.0,
      "step": 14045
    },
    {
      "entropy": 5.54159026145935,
      "epoch": 1.3944025406907503,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.0601,
      "mean_token_accuracy": 0.17354130148887634,
      "num_tokens": 30749002.0,
      "step": 14050
    },
    {
      "entropy": 5.552857875823975,
      "epoch": 1.3948987693529178,
      "grad_norm": 0.875,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.1685,
      "mean_token_accuracy": 0.17579727470874787,
      "num_tokens": 30760056.0,
      "step": 14055
    },
    {
      "entropy": 5.550336837768555,
      "epoch": 1.3953949980150853,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16178627610206603,
      "num_tokens": 30771409.0,
      "step": 14060
    },
    {
      "entropy": 5.4688544273376465,
      "epoch": 1.395891226677253,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.1562,
      "mean_token_accuracy": 0.17047780007123947,
      "num_tokens": 30782976.0,
      "step": 14065
    },
    {
      "entropy": 5.5276198387146,
      "epoch": 1.3963874553394204,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.0861,
      "mean_token_accuracy": 0.1763361781835556,
      "num_tokens": 30794745.0,
      "step": 14070
    },
    {
      "entropy": 5.582569599151611,
      "epoch": 1.396883684001588,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.1556,
      "mean_token_accuracy": 0.16592856198549272,
      "num_tokens": 30807034.0,
      "step": 14075
    },
    {
      "entropy": 5.583014297485351,
      "epoch": 1.3973799126637554,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.1833,
      "mean_token_accuracy": 0.16272977143526077,
      "num_tokens": 30818117.0,
      "step": 14080
    },
    {
      "entropy": 5.479445934295654,
      "epoch": 1.397876141325923,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.0723,
      "mean_token_accuracy": 0.17805538922548295,
      "num_tokens": 30829222.0,
      "step": 14085
    },
    {
      "entropy": 5.511166334152222,
      "epoch": 1.3983723699880906,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1689010366797447,
      "num_tokens": 30839766.0,
      "step": 14090
    },
    {
      "entropy": 5.555560731887818,
      "epoch": 1.398868598650258,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.1569,
      "mean_token_accuracy": 0.17226345837116241,
      "num_tokens": 30850381.0,
      "step": 14095
    },
    {
      "entropy": 5.524771404266358,
      "epoch": 1.3993648273124255,
      "grad_norm": 0.875,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.123,
      "mean_token_accuracy": 0.17094842195510865,
      "num_tokens": 30860851.0,
      "step": 14100
    },
    {
      "entropy": 5.554582405090332,
      "epoch": 1.399861055974593,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.1244,
      "mean_token_accuracy": 0.1617887794971466,
      "num_tokens": 30871246.0,
      "step": 14105
    },
    {
      "entropy": 5.540473175048828,
      "epoch": 1.4003572846367607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.195,
      "mean_token_accuracy": 0.17187904417514802,
      "num_tokens": 30883092.0,
      "step": 14110
    },
    {
      "entropy": 5.575849199295044,
      "epoch": 1.4008535132989282,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.2012,
      "mean_token_accuracy": 0.1710610181093216,
      "num_tokens": 30895848.0,
      "step": 14115
    },
    {
      "entropy": 5.57394061088562,
      "epoch": 1.4013497419610956,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.1676,
      "mean_token_accuracy": 0.16563466638326646,
      "num_tokens": 30907085.0,
      "step": 14120
    },
    {
      "entropy": 5.570524835586548,
      "epoch": 1.4018459706232633,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16442003697156907,
      "num_tokens": 30917735.0,
      "step": 14125
    },
    {
      "entropy": 5.5918830871582035,
      "epoch": 1.4023421992854308,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.1037,
      "mean_token_accuracy": 0.1694877862930298,
      "num_tokens": 30927736.0,
      "step": 14130
    },
    {
      "entropy": 5.524030065536499,
      "epoch": 1.4028384279475983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.1394,
      "mean_token_accuracy": 0.1686388924717903,
      "num_tokens": 30938224.0,
      "step": 14135
    },
    {
      "entropy": 5.488293790817261,
      "epoch": 1.4033346566097658,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.1231,
      "mean_token_accuracy": 0.17574870884418486,
      "num_tokens": 30948537.0,
      "step": 14140
    },
    {
      "entropy": 5.545676422119141,
      "epoch": 1.4038308852719332,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1682628035545349,
      "num_tokens": 30959545.0,
      "step": 14145
    },
    {
      "entropy": 5.506363010406494,
      "epoch": 1.4043271139341007,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1745196223258972,
      "num_tokens": 30970026.0,
      "step": 14150
    },
    {
      "entropy": 5.563494110107422,
      "epoch": 1.4048233425962684,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.1368,
      "mean_token_accuracy": 0.1677560657262802,
      "num_tokens": 30981191.0,
      "step": 14155
    },
    {
      "entropy": 5.5054826736450195,
      "epoch": 1.4053195712584359,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17431530952453614,
      "num_tokens": 30992427.0,
      "step": 14160
    },
    {
      "entropy": 5.552686738967895,
      "epoch": 1.4058157999206033,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.2377,
      "mean_token_accuracy": 0.16442525684833526,
      "num_tokens": 31004972.0,
      "step": 14165
    },
    {
      "entropy": 5.54848165512085,
      "epoch": 1.406312028582771,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.198,
      "mean_token_accuracy": 0.16510843634605407,
      "num_tokens": 31015043.0,
      "step": 14170
    },
    {
      "entropy": 5.535326194763184,
      "epoch": 1.4068082572449385,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.1221,
      "mean_token_accuracy": 0.1781321495771408,
      "num_tokens": 31026241.0,
      "step": 14175
    },
    {
      "entropy": 5.522850370407104,
      "epoch": 1.407304485907106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.113,
      "mean_token_accuracy": 0.16710906028747557,
      "num_tokens": 31036548.0,
      "step": 14180
    },
    {
      "entropy": 5.535211038589478,
      "epoch": 1.4078007145692735,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.1012,
      "mean_token_accuracy": 0.16418777108192445,
      "num_tokens": 31046948.0,
      "step": 14185
    },
    {
      "entropy": 5.568821716308594,
      "epoch": 1.408296943231441,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.1795,
      "mean_token_accuracy": 0.17108646780252457,
      "num_tokens": 31057679.0,
      "step": 14190
    },
    {
      "entropy": 5.552938938140869,
      "epoch": 1.4087931718936086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.1571,
      "mean_token_accuracy": 0.1684094563126564,
      "num_tokens": 31068108.0,
      "step": 14195
    },
    {
      "entropy": 5.556671476364135,
      "epoch": 1.4092894005557761,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17133490294218062,
      "num_tokens": 31079229.0,
      "step": 14200
    },
    {
      "entropy": 5.51038613319397,
      "epoch": 1.4097856292179436,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.1357,
      "mean_token_accuracy": 0.16862695962190627,
      "num_tokens": 31090581.0,
      "step": 14205
    },
    {
      "entropy": 5.6055100440979,
      "epoch": 1.4102818578801113,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.1969,
      "mean_token_accuracy": 0.1629125878214836,
      "num_tokens": 31102904.0,
      "step": 14210
    },
    {
      "entropy": 5.640309286117554,
      "epoch": 1.4107780865422788,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.1838,
      "mean_token_accuracy": 0.17239202409982682,
      "num_tokens": 31115123.0,
      "step": 14215
    },
    {
      "entropy": 5.542236471176148,
      "epoch": 1.4112743152044462,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.2135,
      "mean_token_accuracy": 0.16888072490692138,
      "num_tokens": 31126374.0,
      "step": 14220
    },
    {
      "entropy": 5.541384267807007,
      "epoch": 1.4117705438666137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.1829,
      "mean_token_accuracy": 0.17218050956726075,
      "num_tokens": 31137699.0,
      "step": 14225
    },
    {
      "entropy": 5.5834540843963625,
      "epoch": 1.4122667725287812,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.1737,
      "mean_token_accuracy": 0.1736535385251045,
      "num_tokens": 31149202.0,
      "step": 14230
    },
    {
      "entropy": 5.5271200180053714,
      "epoch": 1.4127630011909487,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.1611,
      "mean_token_accuracy": 0.17102789729833603,
      "num_tokens": 31162624.0,
      "step": 14235
    },
    {
      "entropy": 5.501592683792114,
      "epoch": 1.4132592298531164,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.1045,
      "mean_token_accuracy": 0.17532749474048615,
      "num_tokens": 31172604.0,
      "step": 14240
    },
    {
      "entropy": 5.58136305809021,
      "epoch": 1.4137554585152838,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1678159147500992,
      "num_tokens": 31181899.0,
      "step": 14245
    },
    {
      "entropy": 5.559296798706055,
      "epoch": 1.4142516871774513,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.1664,
      "mean_token_accuracy": 0.16486681550741195,
      "num_tokens": 31193110.0,
      "step": 14250
    },
    {
      "entropy": 5.544984149932861,
      "epoch": 1.414747915839619,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.0787,
      "mean_token_accuracy": 0.17500608712434768,
      "num_tokens": 31202901.0,
      "step": 14255
    },
    {
      "entropy": 5.515003776550293,
      "epoch": 1.4152441445017865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1716371700167656,
      "num_tokens": 31213371.0,
      "step": 14260
    },
    {
      "entropy": 5.588757467269898,
      "epoch": 1.415740373163954,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.1629,
      "mean_token_accuracy": 0.16924041211605073,
      "num_tokens": 31224043.0,
      "step": 14265
    },
    {
      "entropy": 5.590748405456543,
      "epoch": 1.4162366018261214,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.21,
      "mean_token_accuracy": 0.1624375104904175,
      "num_tokens": 31236459.0,
      "step": 14270
    },
    {
      "entropy": 5.5237668514251705,
      "epoch": 1.416732830488289,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.1166,
      "mean_token_accuracy": 0.16987296491861342,
      "num_tokens": 31247669.0,
      "step": 14275
    },
    {
      "entropy": 5.538008308410644,
      "epoch": 1.4172290591504566,
      "grad_norm": 1.0,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.1486,
      "mean_token_accuracy": 0.16687043756246567,
      "num_tokens": 31258886.0,
      "step": 14280
    },
    {
      "entropy": 5.632862567901611,
      "epoch": 1.417725287812624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1670844942331314,
      "num_tokens": 31271798.0,
      "step": 14285
    },
    {
      "entropy": 5.585988140106201,
      "epoch": 1.4182215164747916,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.1654,
      "mean_token_accuracy": 0.17095526158809662,
      "num_tokens": 31283075.0,
      "step": 14290
    },
    {
      "entropy": 5.5972963809967045,
      "epoch": 1.418717745136959,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.1796,
      "mean_token_accuracy": 0.1700448900461197,
      "num_tokens": 31292791.0,
      "step": 14295
    },
    {
      "entropy": 5.568958902359009,
      "epoch": 1.4192139737991267,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000480550610125198,
      "loss": 5.1796,
      "mean_token_accuracy": 0.16730034053325654,
      "num_tokens": 31302803.0,
      "step": 14300
    },
    {
      "entropy": 5.548325300216675,
      "epoch": 1.4197102024612942,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004805361973083066,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17574991136789322,
      "num_tokens": 31314620.0,
      "step": 14305
    },
    {
      "entropy": 5.549931621551513,
      "epoch": 1.4202064311234617,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.1446,
      "mean_token_accuracy": 0.16668355762958526,
      "num_tokens": 31324861.0,
      "step": 14310
    },
    {
      "entropy": 5.55963888168335,
      "epoch": 1.4207026597856292,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.1458,
      "mean_token_accuracy": 0.16731596291065215,
      "num_tokens": 31337264.0,
      "step": 14315
    },
    {
      "entropy": 5.517970848083496,
      "epoch": 1.4211988884477966,
      "grad_norm": 0.875,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17248744070529937,
      "num_tokens": 31347584.0,
      "step": 14320
    },
    {
      "entropy": 5.634580612182617,
      "epoch": 1.4216951171099643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.2894,
      "mean_token_accuracy": 0.17099623680114745,
      "num_tokens": 31358871.0,
      "step": 14325
    },
    {
      "entropy": 5.5880820751190186,
      "epoch": 1.4221913457721318,
      "grad_norm": 0.875,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16274407953023912,
      "num_tokens": 31371088.0,
      "step": 14330
    },
    {
      "entropy": 5.527859544754028,
      "epoch": 1.4226875744342993,
      "grad_norm": 0.875,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1664094388484955,
      "num_tokens": 31382608.0,
      "step": 14335
    },
    {
      "entropy": 5.486207628250122,
      "epoch": 1.423183803096467,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1722821459174156,
      "num_tokens": 31394098.0,
      "step": 14340
    },
    {
      "entropy": 5.6375110149383545,
      "epoch": 1.4236800317586344,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16454959213733672,
      "num_tokens": 31403927.0,
      "step": 14345
    },
    {
      "entropy": 5.500016403198242,
      "epoch": 1.424176260420802,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.1598,
      "mean_token_accuracy": 0.16966182440519334,
      "num_tokens": 31413922.0,
      "step": 14350
    },
    {
      "entropy": 5.632697057723999,
      "epoch": 1.4246724890829694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.3177,
      "mean_token_accuracy": 0.15894487127661705,
      "num_tokens": 31425020.0,
      "step": 14355
    },
    {
      "entropy": 5.536109066009521,
      "epoch": 1.4251687177451369,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.1033,
      "mean_token_accuracy": 0.17127999812364578,
      "num_tokens": 31435763.0,
      "step": 14360
    },
    {
      "entropy": 5.540956974029541,
      "epoch": 1.4256649464073046,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.1058,
      "mean_token_accuracy": 0.17396148592233657,
      "num_tokens": 31445112.0,
      "step": 14365
    },
    {
      "entropy": 5.575214815139771,
      "epoch": 1.426161175069472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.1768,
      "mean_token_accuracy": 0.1708570197224617,
      "num_tokens": 31456380.0,
      "step": 14370
    },
    {
      "entropy": 5.587115859985351,
      "epoch": 1.4266574037316395,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000480333882864924,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1623685359954834,
      "num_tokens": 31468764.0,
      "step": 14375
    },
    {
      "entropy": 5.4520583152771,
      "epoch": 1.427153632393807,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004803193936328092,
      "loss": 4.9887,
      "mean_token_accuracy": 0.17766078263521196,
      "num_tokens": 31478982.0,
      "step": 14380
    },
    {
      "entropy": 5.599137878417968,
      "epoch": 1.4276498610559747,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.2607,
      "mean_token_accuracy": 0.1591508701443672,
      "num_tokens": 31489615.0,
      "step": 14385
    },
    {
      "entropy": 5.578116035461425,
      "epoch": 1.4281460897181422,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17569950819015503,
      "num_tokens": 31500688.0,
      "step": 14390
    },
    {
      "entropy": 5.504024410247803,
      "epoch": 1.4286423183803096,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.1122,
      "mean_token_accuracy": 0.17218586951494216,
      "num_tokens": 31512044.0,
      "step": 14395
    },
    {
      "entropy": 5.543915843963623,
      "epoch": 1.4291385470424771,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.1832,
      "mean_token_accuracy": 0.17169902175664903,
      "num_tokens": 31523786.0,
      "step": 14400
    },
    {
      "entropy": 5.5271241664886475,
      "epoch": 1.4296347757046446,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.1715,
      "mean_token_accuracy": 0.16955926269292831,
      "num_tokens": 31534799.0,
      "step": 14405
    },
    {
      "entropy": 5.549414825439453,
      "epoch": 1.4301310043668123,
      "grad_norm": 1.015625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1740756630897522,
      "num_tokens": 31544607.0,
      "step": 14410
    },
    {
      "entropy": 5.534007883071899,
      "epoch": 1.4306272330289798,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.1176,
      "mean_token_accuracy": 0.1702655926346779,
      "num_tokens": 31556173.0,
      "step": 14415
    },
    {
      "entropy": 5.481973028182983,
      "epoch": 1.4311234616911472,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.0906,
      "mean_token_accuracy": 0.17108992785215377,
      "num_tokens": 31568206.0,
      "step": 14420
    },
    {
      "entropy": 5.607807922363281,
      "epoch": 1.431619690353315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000480188761470399,
      "loss": 5.2266,
      "mean_token_accuracy": 0.158329339325428,
      "num_tokens": 31580021.0,
      "step": 14425
    },
    {
      "entropy": 5.521157884597779,
      "epoch": 1.4321159190154824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17361550778150558,
      "num_tokens": 31590343.0,
      "step": 14430
    },
    {
      "entropy": 5.518551921844482,
      "epoch": 1.4326121476776499,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.1431,
      "mean_token_accuracy": 0.16944821178913116,
      "num_tokens": 31603387.0,
      "step": 14435
    },
    {
      "entropy": 5.503916835784912,
      "epoch": 1.4331083763398174,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.0902,
      "mean_token_accuracy": 0.17408197671175002,
      "num_tokens": 31613640.0,
      "step": 14440
    },
    {
      "entropy": 5.5594704151153564,
      "epoch": 1.4336046050019848,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000480130570421727,
      "loss": 5.1543,
      "mean_token_accuracy": 0.17374491244554519,
      "num_tokens": 31624299.0,
      "step": 14445
    },
    {
      "entropy": 5.610043287277222,
      "epoch": 1.4341008336641523,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.1896,
      "mean_token_accuracy": 0.1688566818833351,
      "num_tokens": 31635878.0,
      "step": 14450
    },
    {
      "entropy": 5.526292037963867,
      "epoch": 1.43459706232632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.072,
      "mean_token_accuracy": 0.17360926568508148,
      "num_tokens": 31648132.0,
      "step": 14455
    },
    {
      "entropy": 5.505326652526856,
      "epoch": 1.4350932909884875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.1566,
      "mean_token_accuracy": 0.17275945991277694,
      "num_tokens": 31657791.0,
      "step": 14460
    },
    {
      "entropy": 5.500456380844116,
      "epoch": 1.435589519650655,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.068,
      "mean_token_accuracy": 0.17608202397823333,
      "num_tokens": 31669561.0,
      "step": 14465
    },
    {
      "entropy": 5.527431058883667,
      "epoch": 1.4360857483128227,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.1089,
      "mean_token_accuracy": 0.17133782655000687,
      "num_tokens": 31680099.0,
      "step": 14470
    },
    {
      "entropy": 5.540792512893677,
      "epoch": 1.4365819769749901,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1732966959476471,
      "num_tokens": 31692047.0,
      "step": 14475
    },
    {
      "entropy": 5.573910713195801,
      "epoch": 1.4370782056371576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.1896,
      "mean_token_accuracy": 0.16715002357959746,
      "num_tokens": 31703376.0,
      "step": 14480
    },
    {
      "entropy": 5.53965539932251,
      "epoch": 1.437574434299325,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.1946,
      "mean_token_accuracy": 0.16239404752850534,
      "num_tokens": 31714201.0,
      "step": 14485
    },
    {
      "entropy": 5.617218017578125,
      "epoch": 1.4380706629614926,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.2133,
      "mean_token_accuracy": 0.16210467219352723,
      "num_tokens": 31725357.0,
      "step": 14490
    },
    {
      "entropy": 5.615300226211548,
      "epoch": 1.4385668916236602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000479984736774914,
      "loss": 5.2403,
      "mean_token_accuracy": 0.168412746489048,
      "num_tokens": 31736311.0,
      "step": 14495
    },
    {
      "entropy": 5.541957855224609,
      "epoch": 1.4390631202858277,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.0983,
      "mean_token_accuracy": 0.17428984940052034,
      "num_tokens": 31747446.0,
      "step": 14500
    },
    {
      "entropy": 5.5835151195526125,
      "epoch": 1.4395593489479952,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.286,
      "mean_token_accuracy": 0.16198173612356187,
      "num_tokens": 31758176.0,
      "step": 14505
    },
    {
      "entropy": 5.5542320728302,
      "epoch": 1.440055577610163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17282420694828032,
      "num_tokens": 31767597.0,
      "step": 14510
    },
    {
      "entropy": 5.584510421752929,
      "epoch": 1.4405518062723304,
      "grad_norm": 1.0,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.23,
      "mean_token_accuracy": 0.16832464933395386,
      "num_tokens": 31779521.0,
      "step": 14515
    },
    {
      "entropy": 5.564969491958618,
      "epoch": 1.4410480349344978,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.131,
      "mean_token_accuracy": 0.17719514220952987,
      "num_tokens": 31791780.0,
      "step": 14520
    },
    {
      "entropy": 5.45128812789917,
      "epoch": 1.4415442635966653,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1774008169770241,
      "num_tokens": 31802280.0,
      "step": 14525
    },
    {
      "entropy": 5.491111993789673,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17564256191253663,
      "num_tokens": 31812515.0,
      "step": 14530
    },
    {
      "entropy": 5.547967100143433,
      "epoch": 1.4425367209210003,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.1931,
      "mean_token_accuracy": 0.16749614328145981,
      "num_tokens": 31823483.0,
      "step": 14535
    },
    {
      "entropy": 5.612756824493408,
      "epoch": 1.443032949583168,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.2699,
      "mean_token_accuracy": 0.16320828050374986,
      "num_tokens": 31834978.0,
      "step": 14540
    },
    {
      "entropy": 5.6302001953125,
      "epoch": 1.4435291782453354,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17441036999225618,
      "num_tokens": 31845375.0,
      "step": 14545
    },
    {
      "entropy": 5.564217042922974,
      "epoch": 1.444025406907503,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.201,
      "mean_token_accuracy": 0.17071913480758666,
      "num_tokens": 31856682.0,
      "step": 14550
    },
    {
      "entropy": 5.5191174983978275,
      "epoch": 1.4445216355696706,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.1185,
      "mean_token_accuracy": 0.17646508663892746,
      "num_tokens": 31867904.0,
      "step": 14555
    },
    {
      "entropy": 5.615478515625,
      "epoch": 1.445017864231838,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.2234,
      "mean_token_accuracy": 0.1688385710120201,
      "num_tokens": 31878929.0,
      "step": 14560
    },
    {
      "entropy": 5.5964220523834225,
      "epoch": 1.4455140928940056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1728791579604149,
      "num_tokens": 31890117.0,
      "step": 14565
    },
    {
      "entropy": 5.579601240158081,
      "epoch": 1.446010321556173,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.1902,
      "mean_token_accuracy": 0.1654447540640831,
      "num_tokens": 31901669.0,
      "step": 14570
    },
    {
      "entropy": 5.500247049331665,
      "epoch": 1.4465065502183405,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.1518,
      "mean_token_accuracy": 0.1711816370487213,
      "num_tokens": 31913774.0,
      "step": 14575
    },
    {
      "entropy": 5.552488899230957,
      "epoch": 1.4470027788805082,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000479735653180094,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17066141813993455,
      "num_tokens": 31924435.0,
      "step": 14580
    },
    {
      "entropy": 5.537312126159668,
      "epoch": 1.4474990075426757,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16235086768865586,
      "num_tokens": 31935359.0,
      "step": 14585
    },
    {
      "entropy": 5.571514892578125,
      "epoch": 1.4479952362048432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.1852,
      "mean_token_accuracy": 0.16805281788110732,
      "num_tokens": 31946224.0,
      "step": 14590
    },
    {
      "entropy": 5.569959354400635,
      "epoch": 1.4484914648670106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.1672,
      "mean_token_accuracy": 0.17326754480600357,
      "num_tokens": 31956458.0,
      "step": 14595
    },
    {
      "entropy": 5.622171115875244,
      "epoch": 1.4489876935291783,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.1799,
      "mean_token_accuracy": 0.16824268251657487,
      "num_tokens": 31966984.0,
      "step": 14600
    },
    {
      "entropy": 5.544392204284668,
      "epoch": 1.4494839221913458,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.1613,
      "mean_token_accuracy": 0.16902875155210495,
      "num_tokens": 31978058.0,
      "step": 14605
    },
    {
      "entropy": 5.558837938308716,
      "epoch": 1.4499801508535133,
      "grad_norm": 0.984375,
      "learning_rate": 0.000479647390917855,
      "loss": 5.1912,
      "mean_token_accuracy": 0.16001046299934388,
      "num_tokens": 31987968.0,
      "step": 14610
    },
    {
      "entropy": 5.5057532787323,
      "epoch": 1.4504763795156808,
      "grad_norm": 0.9375,
      "learning_rate": 0.000479632662776836,
      "loss": 5.1322,
      "mean_token_accuracy": 0.17000224739313125,
      "num_tokens": 31999005.0,
      "step": 14615
    },
    {
      "entropy": 5.606134939193725,
      "epoch": 1.4509726081778482,
      "grad_norm": 0.890625,
      "learning_rate": 0.000479617929561368,
      "loss": 5.2328,
      "mean_token_accuracy": 0.1606484219431877,
      "num_tokens": 32011997.0,
      "step": 14620
    },
    {
      "entropy": 5.629475545883179,
      "epoch": 1.451468836840016,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.2277,
      "mean_token_accuracy": 0.16799581944942474,
      "num_tokens": 32024429.0,
      "step": 14625
    },
    {
      "entropy": 5.54275279045105,
      "epoch": 1.4519650655021834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17372212260961534,
      "num_tokens": 32034176.0,
      "step": 14630
    },
    {
      "entropy": 5.479806900024414,
      "epoch": 1.4524612941643509,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.0864,
      "mean_token_accuracy": 0.17401088029146194,
      "num_tokens": 32045482.0,
      "step": 14635
    },
    {
      "entropy": 5.514632654190064,
      "epoch": 1.4529575228265186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17061083167791366,
      "num_tokens": 32056890.0,
      "step": 14640
    },
    {
      "entropy": 5.5025553703308105,
      "epoch": 1.453453751488686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.0533,
      "mean_token_accuracy": 0.1823858454823494,
      "num_tokens": 32067915.0,
      "step": 14645
    },
    {
      "entropy": 5.5020341873168945,
      "epoch": 1.4539499801508535,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1796295866370201,
      "num_tokens": 32080026.0,
      "step": 14650
    },
    {
      "entropy": 5.4978984832763675,
      "epoch": 1.454446208813021,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.0975,
      "mean_token_accuracy": 0.17726409137248994,
      "num_tokens": 32090150.0,
      "step": 14655
    },
    {
      "entropy": 5.6345724105834964,
      "epoch": 1.4549424374751885,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.2219,
      "mean_token_accuracy": 0.1666557937860489,
      "num_tokens": 32101858.0,
      "step": 14660
    },
    {
      "entropy": 5.475343799591064,
      "epoch": 1.4554386661373562,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17893314212560654,
      "num_tokens": 32112078.0,
      "step": 14665
    },
    {
      "entropy": 5.5465490341186525,
      "epoch": 1.4559348947995236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.1663,
      "mean_token_accuracy": 0.17334366887807845,
      "num_tokens": 32123539.0,
      "step": 14670
    },
    {
      "entropy": 5.621039724349975,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1706137701869011,
      "num_tokens": 32135063.0,
      "step": 14675
    },
    {
      "entropy": 5.564453363418579,
      "epoch": 1.4569273521238586,
      "grad_norm": 0.921875,
      "learning_rate": 0.000479440735301836,
      "loss": 5.2033,
      "mean_token_accuracy": 0.16485894620418548,
      "num_tokens": 32145407.0,
      "step": 14680
    },
    {
      "entropy": 5.520057344436646,
      "epoch": 1.4574235807860263,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.1491,
      "mean_token_accuracy": 0.17391877472400666,
      "num_tokens": 32156340.0,
      "step": 14685
    },
    {
      "entropy": 5.506275367736817,
      "epoch": 1.4579198094481938,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17207629829645157,
      "num_tokens": 32168627.0,
      "step": 14690
    },
    {
      "entropy": 5.534217739105225,
      "epoch": 1.4584160381103612,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.1377,
      "mean_token_accuracy": 0.16507538557052612,
      "num_tokens": 32179770.0,
      "step": 14695
    },
    {
      "entropy": 5.619875717163086,
      "epoch": 1.4589122667725287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1733524203300476,
      "num_tokens": 32190309.0,
      "step": 14700
    },
    {
      "entropy": 5.575609254837036,
      "epoch": 1.4594084954346962,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.1424,
      "mean_token_accuracy": 0.17015122771263122,
      "num_tokens": 32201113.0,
      "step": 14705
    },
    {
      "entropy": 5.544891214370727,
      "epoch": 1.459904724096864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.1708,
      "mean_token_accuracy": 0.1722333014011383,
      "num_tokens": 32211412.0,
      "step": 14710
    },
    {
      "entropy": 5.619385671615601,
      "epoch": 1.4604009527590314,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.1426,
      "mean_token_accuracy": 0.17304779589176178,
      "num_tokens": 32222992.0,
      "step": 14715
    },
    {
      "entropy": 5.540670490264892,
      "epoch": 1.4608971814211988,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.1144,
      "mean_token_accuracy": 0.17296387404203414,
      "num_tokens": 32234600.0,
      "step": 14720
    },
    {
      "entropy": 5.604022645950318,
      "epoch": 1.4613934100833665,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1629785805940628,
      "num_tokens": 32245835.0,
      "step": 14725
    },
    {
      "entropy": 5.544327926635742,
      "epoch": 1.461889638745534,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.1963,
      "mean_token_accuracy": 0.17332597225904464,
      "num_tokens": 32257006.0,
      "step": 14730
    },
    {
      "entropy": 5.580705165863037,
      "epoch": 1.4623858674077015,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.1569,
      "mean_token_accuracy": 0.16713878959417344,
      "num_tokens": 32268049.0,
      "step": 14735
    },
    {
      "entropy": 5.523055553436279,
      "epoch": 1.462882096069869,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.1253,
      "mean_token_accuracy": 0.17413712292909622,
      "num_tokens": 32278035.0,
      "step": 14740
    },
    {
      "entropy": 5.571798324584961,
      "epoch": 1.4633783247320364,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.1468,
      "mean_token_accuracy": 0.16694454252719879,
      "num_tokens": 32289029.0,
      "step": 14745
    },
    {
      "entropy": 5.558476877212525,
      "epoch": 1.463874553394204,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.1405,
      "mean_token_accuracy": 0.16665222346782685,
      "num_tokens": 32299656.0,
      "step": 14750
    },
    {
      "entropy": 5.573389625549316,
      "epoch": 1.4643707820563716,
      "grad_norm": 0.828125,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.1165,
      "mean_token_accuracy": 0.1667672410607338,
      "num_tokens": 32309394.0,
      "step": 14755
    },
    {
      "entropy": 5.580714464187622,
      "epoch": 1.464867010718539,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1686982750892639,
      "num_tokens": 32319908.0,
      "step": 14760
    },
    {
      "entropy": 5.50615963935852,
      "epoch": 1.4653632393807066,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.0844,
      "mean_token_accuracy": 0.17705117166042328,
      "num_tokens": 32331190.0,
      "step": 14765
    },
    {
      "entropy": 5.545786046981812,
      "epoch": 1.4658594680428743,
      "grad_norm": 0.875,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17662478387355804,
      "num_tokens": 32342150.0,
      "step": 14770
    },
    {
      "entropy": 5.590291166305542,
      "epoch": 1.4663556967050417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.1938,
      "mean_token_accuracy": 0.16811473220586776,
      "num_tokens": 32353480.0,
      "step": 14775
    },
    {
      "entropy": 5.602324199676514,
      "epoch": 1.4668519253672092,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1737433224916458,
      "num_tokens": 32366365.0,
      "step": 14780
    },
    {
      "entropy": 5.604742908477784,
      "epoch": 1.4673481540293767,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.2116,
      "mean_token_accuracy": 0.16815920323133468,
      "num_tokens": 32377186.0,
      "step": 14785
    },
    {
      "entropy": 5.554545450210571,
      "epoch": 1.4678443826915442,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1664778172969818,
      "num_tokens": 32389207.0,
      "step": 14790
    },
    {
      "entropy": 5.562849521636963,
      "epoch": 1.4683406113537119,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.1928,
      "mean_token_accuracy": 0.171696475148201,
      "num_tokens": 32400278.0,
      "step": 14795
    },
    {
      "entropy": 5.608977317810059,
      "epoch": 1.4688368400158793,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.2102,
      "mean_token_accuracy": 0.16652225106954574,
      "num_tokens": 32412237.0,
      "step": 14800
    },
    {
      "entropy": 5.585179090499878,
      "epoch": 1.4693330686780468,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16812522262334822,
      "num_tokens": 32423236.0,
      "step": 14805
    },
    {
      "entropy": 5.577445840835571,
      "epoch": 1.4698292973402145,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000479054310827975,
      "loss": 5.1482,
      "mean_token_accuracy": 0.16989711821079254,
      "num_tokens": 32434296.0,
      "step": 14810
    },
    {
      "entropy": 5.560256576538086,
      "epoch": 1.470325526002382,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.1243,
      "mean_token_accuracy": 0.17324079275131227,
      "num_tokens": 32444934.0,
      "step": 14815
    },
    {
      "entropy": 5.48416051864624,
      "epoch": 1.4708217546645495,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17098385840654373,
      "num_tokens": 32454157.0,
      "step": 14820
    },
    {
      "entropy": 5.5118402481079105,
      "epoch": 1.471317983326717,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.1187,
      "mean_token_accuracy": 0.17230488061904908,
      "num_tokens": 32464586.0,
      "step": 14825
    },
    {
      "entropy": 5.585790491104126,
      "epoch": 1.4718142119888844,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1670299157500267,
      "num_tokens": 32474121.0,
      "step": 14830
    },
    {
      "entropy": 5.563994455337524,
      "epoch": 1.4723104406510519,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.193,
      "mean_token_accuracy": 0.16661125048995018,
      "num_tokens": 32485539.0,
      "step": 14835
    },
    {
      "entropy": 5.536301326751709,
      "epoch": 1.4728066693132196,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.0785,
      "mean_token_accuracy": 0.17559134662151338,
      "num_tokens": 32495970.0,
      "step": 14840
    },
    {
      "entropy": 5.534909105300903,
      "epoch": 1.473302897975387,
      "grad_norm": 1.0,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.0931,
      "mean_token_accuracy": 0.1729356214404106,
      "num_tokens": 32506401.0,
      "step": 14845
    },
    {
      "entropy": 5.53083348274231,
      "epoch": 1.4737991266375545,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.127,
      "mean_token_accuracy": 0.17543739080429077,
      "num_tokens": 32518033.0,
      "step": 14850
    },
    {
      "entropy": 5.57931809425354,
      "epoch": 1.4742953552997222,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047891975121368,
      "loss": 5.1864,
      "mean_token_accuracy": 0.16746432036161424,
      "num_tokens": 32529225.0,
      "step": 14855
    },
    {
      "entropy": 5.582863616943359,
      "epoch": 1.4747915839618897,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16609462946653367,
      "num_tokens": 32540246.0,
      "step": 14860
    },
    {
      "entropy": 5.514565420150757,
      "epoch": 1.4752878126240572,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.1594,
      "mean_token_accuracy": 0.1693726062774658,
      "num_tokens": 32551001.0,
      "step": 14865
    },
    {
      "entropy": 5.592730712890625,
      "epoch": 1.4757840412862246,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.1427,
      "mean_token_accuracy": 0.16571438908576966,
      "num_tokens": 32561338.0,
      "step": 14870
    },
    {
      "entropy": 5.557177114486694,
      "epoch": 1.4762802699483921,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.1474,
      "mean_token_accuracy": 0.17144385129213333,
      "num_tokens": 32570919.0,
      "step": 14875
    },
    {
      "entropy": 5.536419010162353,
      "epoch": 1.4767764986105598,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1707430213689804,
      "num_tokens": 32581361.0,
      "step": 14880
    },
    {
      "entropy": 5.559890699386597,
      "epoch": 1.4772727272727273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.1912,
      "mean_token_accuracy": 0.16955806612968444,
      "num_tokens": 32592533.0,
      "step": 14885
    },
    {
      "entropy": 5.585031461715698,
      "epoch": 1.4777689559348948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.0771,
      "mean_token_accuracy": 0.1779586210846901,
      "num_tokens": 32602207.0,
      "step": 14890
    },
    {
      "entropy": 5.550985670089721,
      "epoch": 1.4782651845970625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.0811,
      "mean_token_accuracy": 0.17342452704906464,
      "num_tokens": 32611288.0,
      "step": 14895
    },
    {
      "entropy": 5.586190414428711,
      "epoch": 1.47876141325923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.1566,
      "mean_token_accuracy": 0.16284559220075606,
      "num_tokens": 32622078.0,
      "step": 14900
    },
    {
      "entropy": 5.579289865493775,
      "epoch": 1.4792576419213974,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.2625,
      "mean_token_accuracy": 0.16423314213752746,
      "num_tokens": 32634182.0,
      "step": 14905
    },
    {
      "entropy": 5.5943114280700685,
      "epoch": 1.479753870583565,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.2284,
      "mean_token_accuracy": 0.16788498759269715,
      "num_tokens": 32646245.0,
      "step": 14910
    },
    {
      "entropy": 5.468972444534302,
      "epoch": 1.4802500992457324,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.1157,
      "mean_token_accuracy": 0.16907917857170104,
      "num_tokens": 32657521.0,
      "step": 14915
    },
    {
      "entropy": 5.569070100784302,
      "epoch": 1.4807463279078998,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1691214397549629,
      "num_tokens": 32668740.0,
      "step": 14920
    },
    {
      "entropy": 5.6110718727111815,
      "epoch": 1.4812425565700675,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.1794,
      "mean_token_accuracy": 0.16729661971330642,
      "num_tokens": 32680933.0,
      "step": 14925
    },
    {
      "entropy": 5.512780046463012,
      "epoch": 1.481738785232235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.0922,
      "mean_token_accuracy": 0.17291189283132552,
      "num_tokens": 32691690.0,
      "step": 14930
    },
    {
      "entropy": 5.490570783615112,
      "epoch": 1.4822350138944025,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1754784807562828,
      "num_tokens": 32703500.0,
      "step": 14935
    },
    {
      "entropy": 5.5505259990692135,
      "epoch": 1.4827312425565702,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.1668,
      "mean_token_accuracy": 0.16830314546823502,
      "num_tokens": 32714662.0,
      "step": 14940
    },
    {
      "entropy": 5.498032808303833,
      "epoch": 1.4832274712187377,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17432281523942947,
      "num_tokens": 32725995.0,
      "step": 14945
    },
    {
      "entropy": 5.523089075088501,
      "epoch": 1.4837236998809051,
      "grad_norm": 0.953125,
      "learning_rate": 0.000478634336145889,
      "loss": 5.1381,
      "mean_token_accuracy": 0.17249958217144012,
      "num_tokens": 32736510.0,
      "step": 14950
    },
    {
      "entropy": 5.5677539825439455,
      "epoch": 1.4842199285430726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.171,
      "mean_token_accuracy": 0.16978151798248292,
      "num_tokens": 32747032.0,
      "step": 14955
    },
    {
      "entropy": 5.514405345916748,
      "epoch": 1.48471615720524,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.1265,
      "mean_token_accuracy": 0.17562112659215928,
      "num_tokens": 32757199.0,
      "step": 14960
    },
    {
      "entropy": 5.535347509384155,
      "epoch": 1.4852123858674078,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.1129,
      "mean_token_accuracy": 0.1768953323364258,
      "num_tokens": 32767326.0,
      "step": 14965
    },
    {
      "entropy": 5.64764666557312,
      "epoch": 1.4857086145295753,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000478574016391523,
      "loss": 5.2195,
      "mean_token_accuracy": 0.16260064244270325,
      "num_tokens": 32777621.0,
      "step": 14970
    },
    {
      "entropy": 5.500770616531372,
      "epoch": 1.4862048431917427,
      "grad_norm": 1.03125,
      "learning_rate": 0.000478558923831504,
      "loss": 5.1001,
      "mean_token_accuracy": 0.17178967893123626,
      "num_tokens": 32787938.0,
      "step": 14975
    },
    {
      "entropy": 5.482077884674072,
      "epoch": 1.4867010718539102,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18001521974802018,
      "num_tokens": 32798047.0,
      "step": 14980
    },
    {
      "entropy": 5.557400369644165,
      "epoch": 1.487197300516078,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.1803,
      "mean_token_accuracy": 0.1724357858300209,
      "num_tokens": 32809338.0,
      "step": 14985
    },
    {
      "entropy": 5.5663323402404785,
      "epoch": 1.4876935291782454,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17325010597705842,
      "num_tokens": 32820555.0,
      "step": 14990
    },
    {
      "entropy": 5.591356945037842,
      "epoch": 1.4881897578404129,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.1616,
      "mean_token_accuracy": 0.16253309249877929,
      "num_tokens": 32830832.0,
      "step": 14995
    },
    {
      "entropy": 5.508796834945679,
      "epoch": 1.4886859865025803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1731090873479843,
      "num_tokens": 32840534.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.396138208443208,
      "eval_loss": 5.2559638023376465,
      "eval_mean_token_accuracy": 0.1709773288267501,
      "eval_num_tokens": 32840534.0,
      "eval_runtime": 20.3836,
      "eval_samples_per_second": 1651.77,
      "eval_steps_per_second": 206.49,
      "step": 15000
    },
    {
      "entropy": 5.551625680923462,
      "epoch": 1.4891822151647478,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.1049,
      "mean_token_accuracy": 0.17302472293376922,
      "num_tokens": 32851109.0,
      "step": 15005
    },
    {
      "entropy": 5.60433931350708,
      "epoch": 1.4896784438269155,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.255,
      "mean_token_accuracy": 0.16964273303747177,
      "num_tokens": 32862491.0,
      "step": 15010
    },
    {
      "entropy": 5.502454233169556,
      "epoch": 1.490174672489083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.0504,
      "mean_token_accuracy": 0.17719168663024903,
      "num_tokens": 32873831.0,
      "step": 15015
    },
    {
      "entropy": 5.617833805084229,
      "epoch": 1.4906709011512504,
      "grad_norm": 0.953125,
      "learning_rate": 0.000478422863684265,
      "loss": 5.1756,
      "mean_token_accuracy": 0.16890789568424225,
      "num_tokens": 32884288.0,
      "step": 15020
    },
    {
      "entropy": 5.516927766799927,
      "epoch": 1.4911671298134181,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.1061,
      "mean_token_accuracy": 0.16945798546075821,
      "num_tokens": 32894867.0,
      "step": 15025
    },
    {
      "entropy": 5.45184817314148,
      "epoch": 1.4916633584755856,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.1377,
      "mean_token_accuracy": 0.16919610947370528,
      "num_tokens": 32906627.0,
      "step": 15030
    },
    {
      "entropy": 5.5427632331848145,
      "epoch": 1.492159587137753,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.1309,
      "mean_token_accuracy": 0.16880033165216446,
      "num_tokens": 32916245.0,
      "step": 15035
    },
    {
      "entropy": 5.600205278396606,
      "epoch": 1.4926558157999206,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.2195,
      "mean_token_accuracy": 0.17301256954669952,
      "num_tokens": 32927663.0,
      "step": 15040
    },
    {
      "entropy": 5.617769765853882,
      "epoch": 1.493152044462088,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16266803294420243,
      "num_tokens": 32939106.0,
      "step": 15045
    },
    {
      "entropy": 5.511027431488037,
      "epoch": 1.4936482731242555,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16959837526082994,
      "num_tokens": 32949628.0,
      "step": 15050
    },
    {
      "entropy": 5.535074853897095,
      "epoch": 1.4941445017864232,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.1857,
      "mean_token_accuracy": 0.16560948640108109,
      "num_tokens": 32959908.0,
      "step": 15055
    },
    {
      "entropy": 5.554882383346557,
      "epoch": 1.4946407304485907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1740995317697525,
      "num_tokens": 32970312.0,
      "step": 15060
    },
    {
      "entropy": 5.551369667053223,
      "epoch": 1.4951369591107582,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.1245,
      "mean_token_accuracy": 0.16919137239456178,
      "num_tokens": 32982876.0,
      "step": 15065
    },
    {
      "entropy": 5.561912393569946,
      "epoch": 1.4956331877729259,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.1602,
      "mean_token_accuracy": 0.16315172612667084,
      "num_tokens": 32994223.0,
      "step": 15070
    },
    {
      "entropy": 5.651793622970581,
      "epoch": 1.4961294164350933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.2299,
      "mean_token_accuracy": 0.1724542498588562,
      "num_tokens": 33006282.0,
      "step": 15075
    },
    {
      "entropy": 5.601793909072876,
      "epoch": 1.4966256450972608,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.2164,
      "mean_token_accuracy": 0.16755670309066772,
      "num_tokens": 33016823.0,
      "step": 15080
    },
    {
      "entropy": 5.563657617568969,
      "epoch": 1.4971218737594283,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.1658,
      "mean_token_accuracy": 0.16771779358386993,
      "num_tokens": 33027963.0,
      "step": 15085
    },
    {
      "entropy": 5.57607684135437,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.015625,
      "learning_rate": 0.000478210402511098,
      "loss": 5.213,
      "mean_token_accuracy": 0.16735675930976868,
      "num_tokens": 33038711.0,
      "step": 15090
    },
    {
      "entropy": 5.544452857971192,
      "epoch": 1.4981143310837635,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.1336,
      "mean_token_accuracy": 0.17236829847097396,
      "num_tokens": 33048847.0,
      "step": 15095
    },
    {
      "entropy": 5.565048456192017,
      "epoch": 1.498610559745931,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.1644,
      "mean_token_accuracy": 0.16963993459939958,
      "num_tokens": 33059775.0,
      "step": 15100
    },
    {
      "entropy": 5.667967176437378,
      "epoch": 1.4991067884080984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.2674,
      "mean_token_accuracy": 0.1687661662697792,
      "num_tokens": 33071543.0,
      "step": 15105
    },
    {
      "entropy": 5.616515588760376,
      "epoch": 1.499603017070266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.2099,
      "mean_token_accuracy": 0.16770666539669038,
      "num_tokens": 33082487.0,
      "step": 15110
    },
    {
      "entropy": 5.593296098709106,
      "epoch": 1.5000992457324336,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.1819,
      "mean_token_accuracy": 0.166685052216053,
      "num_tokens": 33093381.0,
      "step": 15115
    },
    {
      "entropy": 5.546697950363159,
      "epoch": 1.500595474394601,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.0905,
      "mean_token_accuracy": 0.16974967867136,
      "num_tokens": 33104122.0,
      "step": 15120
    },
    {
      "entropy": 5.579365825653076,
      "epoch": 1.5010917030567685,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.2509,
      "mean_token_accuracy": 0.1709030017256737,
      "num_tokens": 33115541.0,
      "step": 15125
    },
    {
      "entropy": 5.470343112945557,
      "epoch": 1.501587931718936,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.0165,
      "mean_token_accuracy": 0.1760121613740921,
      "num_tokens": 33125002.0,
      "step": 15130
    },
    {
      "entropy": 5.550478792190551,
      "epoch": 1.5020841603811035,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.2065,
      "mean_token_accuracy": 0.17085357159376144,
      "num_tokens": 33135733.0,
      "step": 15135
    },
    {
      "entropy": 5.562398433685303,
      "epoch": 1.5025803890432712,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.141,
      "mean_token_accuracy": 0.1671392872929573,
      "num_tokens": 33146514.0,
      "step": 15140
    },
    {
      "entropy": 5.615274095535279,
      "epoch": 1.5030766177054387,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.1357,
      "mean_token_accuracy": 0.1644637703895569,
      "num_tokens": 33157070.0,
      "step": 15145
    },
    {
      "entropy": 5.508978843688965,
      "epoch": 1.5035728463676064,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.1316,
      "mean_token_accuracy": 0.17476713806390762,
      "num_tokens": 33170311.0,
      "step": 15150
    },
    {
      "entropy": 5.5463995933532715,
      "epoch": 1.5040690750297738,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.1488,
      "mean_token_accuracy": 0.17475616335868835,
      "num_tokens": 33180510.0,
      "step": 15155
    },
    {
      "entropy": 5.548846960067749,
      "epoch": 1.5045653036919413,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.1334,
      "mean_token_accuracy": 0.1713654324412346,
      "num_tokens": 33192265.0,
      "step": 15160
    },
    {
      "entropy": 5.596387720108032,
      "epoch": 1.5050615323541088,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.1319,
      "mean_token_accuracy": 0.1696853294968605,
      "num_tokens": 33204080.0,
      "step": 15165
    },
    {
      "entropy": 5.547253751754761,
      "epoch": 1.5055577610162763,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16693242639303207,
      "num_tokens": 33216090.0,
      "step": 15170
    },
    {
      "entropy": 5.595953941345215,
      "epoch": 1.5060539896784437,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.2104,
      "mean_token_accuracy": 0.16534783244132994,
      "num_tokens": 33227465.0,
      "step": 15175
    },
    {
      "entropy": 5.630669403076172,
      "epoch": 1.5065502183406112,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.1951,
      "mean_token_accuracy": 0.16781466603279113,
      "num_tokens": 33237510.0,
      "step": 15180
    },
    {
      "entropy": 5.630733013153076,
      "epoch": 1.507046447002779,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.2349,
      "mean_token_accuracy": 0.1692189872264862,
      "num_tokens": 33247349.0,
      "step": 15185
    },
    {
      "entropy": 5.542925024032593,
      "epoch": 1.5075426756649464,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17682314068078994,
      "num_tokens": 33258357.0,
      "step": 15190
    },
    {
      "entropy": 5.5718847751617435,
      "epoch": 1.508038904327114,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.2153,
      "mean_token_accuracy": 0.16720317751169206,
      "num_tokens": 33269355.0,
      "step": 15195
    },
    {
      "entropy": 5.555209922790527,
      "epoch": 1.5085351329892815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047787453983779826,
      "loss": 5.0556,
      "mean_token_accuracy": 0.17974577844142914,
      "num_tokens": 33279129.0,
      "step": 15200
    },
    {
      "entropy": 5.55638484954834,
      "epoch": 1.509031361651449,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.0934,
      "mean_token_accuracy": 0.16814342737197877,
      "num_tokens": 33290681.0,
      "step": 15205
    },
    {
      "entropy": 5.552210569381714,
      "epoch": 1.5095275903136165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.1342,
      "mean_token_accuracy": 0.17126743793487548,
      "num_tokens": 33300738.0,
      "step": 15210
    },
    {
      "entropy": 5.578936338424683,
      "epoch": 1.510023818975784,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.2033,
      "mean_token_accuracy": 0.16902662217617034,
      "num_tokens": 33310829.0,
      "step": 15215
    },
    {
      "entropy": 5.561600255966186,
      "epoch": 1.5105200476379514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.1702,
      "mean_token_accuracy": 0.17709518671035768,
      "num_tokens": 33320814.0,
      "step": 15220
    },
    {
      "entropy": 5.515883588790894,
      "epoch": 1.5110162763001191,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.0677,
      "mean_token_accuracy": 0.17599680721759797,
      "num_tokens": 33330105.0,
      "step": 15225
    },
    {
      "entropy": 5.487015724182129,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.08,
      "mean_token_accuracy": 0.17613619714975357,
      "num_tokens": 33340986.0,
      "step": 15230
    },
    {
      "entropy": 5.525114297866821,
      "epoch": 1.512008733624454,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.193,
      "mean_token_accuracy": 0.1639440029859543,
      "num_tokens": 33352096.0,
      "step": 15235
    },
    {
      "entropy": 5.603683567047119,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1652282178401947,
      "num_tokens": 33362581.0,
      "step": 15240
    },
    {
      "entropy": 5.589505863189697,
      "epoch": 1.5130011909487893,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.1987,
      "mean_token_accuracy": 0.17284175753593445,
      "num_tokens": 33374020.0,
      "step": 15245
    },
    {
      "entropy": 5.52776083946228,
      "epoch": 1.5134974196109567,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.1555,
      "mean_token_accuracy": 0.1705294132232666,
      "num_tokens": 33384755.0,
      "step": 15250
    },
    {
      "entropy": 5.5843220233917235,
      "epoch": 1.5139936482731242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.1361,
      "mean_token_accuracy": 0.179190556704998,
      "num_tokens": 33395837.0,
      "step": 15255
    },
    {
      "entropy": 5.552399969100952,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477690315464711,
      "loss": 5.1877,
      "mean_token_accuracy": 0.16576566696166992,
      "num_tokens": 33408083.0,
      "step": 15260
    },
    {
      "entropy": 5.55109395980835,
      "epoch": 1.5149861055974592,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.2284,
      "mean_token_accuracy": 0.17430276572704315,
      "num_tokens": 33419433.0,
      "step": 15265
    },
    {
      "entropy": 5.592087078094482,
      "epoch": 1.5154823342596269,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000477659541016784,
      "loss": 5.2118,
      "mean_token_accuracy": 0.16935246139764787,
      "num_tokens": 33431920.0,
      "step": 15270
    },
    {
      "entropy": 5.598186492919922,
      "epoch": 1.5159785629217943,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.1704,
      "mean_token_accuracy": 0.1651158258318901,
      "num_tokens": 33443428.0,
      "step": 15275
    },
    {
      "entropy": 5.514695501327514,
      "epoch": 1.516474791583962,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.1075,
      "mean_token_accuracy": 0.16407959908246994,
      "num_tokens": 33455324.0,
      "step": 15280
    },
    {
      "entropy": 5.556764936447143,
      "epoch": 1.5169710202461295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.1667,
      "mean_token_accuracy": 0.16894119530916213,
      "num_tokens": 33468640.0,
      "step": 15285
    },
    {
      "entropy": 5.539021682739258,
      "epoch": 1.517467248908297,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.1617,
      "mean_token_accuracy": 0.16957267820835115,
      "num_tokens": 33479681.0,
      "step": 15290
    },
    {
      "entropy": 5.628204393386841,
      "epoch": 1.5179634775704645,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.1945,
      "mean_token_accuracy": 0.1694916695356369,
      "num_tokens": 33490797.0,
      "step": 15295
    },
    {
      "entropy": 5.56722936630249,
      "epoch": 1.518459706232632,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.2308,
      "mean_token_accuracy": 0.1679074451327324,
      "num_tokens": 33501649.0,
      "step": 15300
    },
    {
      "entropy": 5.498570919036865,
      "epoch": 1.5189559348947994,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.0683,
      "mean_token_accuracy": 0.17583872377872467,
      "num_tokens": 33512169.0,
      "step": 15305
    },
    {
      "entropy": 5.539388990402221,
      "epoch": 1.519452163556967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.1588,
      "mean_token_accuracy": 0.16298311054706574,
      "num_tokens": 33522866.0,
      "step": 15310
    },
    {
      "entropy": 5.473582124710083,
      "epoch": 1.5199483922191346,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1742237016558647,
      "num_tokens": 33533671.0,
      "step": 15315
    },
    {
      "entropy": 5.530866289138794,
      "epoch": 1.520444620881302,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.1277,
      "mean_token_accuracy": 0.1744606152176857,
      "num_tokens": 33544475.0,
      "step": 15320
    },
    {
      "entropy": 5.582953882217407,
      "epoch": 1.5209408495434698,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.2252,
      "mean_token_accuracy": 0.1712804526090622,
      "num_tokens": 33555691.0,
      "step": 15325
    },
    {
      "entropy": 5.5883420467376705,
      "epoch": 1.5214370782056372,
      "grad_norm": 1.125,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.164,
      "mean_token_accuracy": 0.16761189550161362,
      "num_tokens": 33566980.0,
      "step": 15330
    },
    {
      "entropy": 5.541941976547241,
      "epoch": 1.5219333068678047,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.0987,
      "mean_token_accuracy": 0.16535341292619704,
      "num_tokens": 33577947.0,
      "step": 15335
    },
    {
      "entropy": 5.5574281215667725,
      "epoch": 1.5224295355299722,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.172,
      "mean_token_accuracy": 0.1737917557358742,
      "num_tokens": 33589703.0,
      "step": 15340
    },
    {
      "entropy": 5.507535600662232,
      "epoch": 1.5229257641921397,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.1508,
      "mean_token_accuracy": 0.17001756876707078,
      "num_tokens": 33599418.0,
      "step": 15345
    },
    {
      "entropy": 5.576603794097901,
      "epoch": 1.5234219928543071,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000477412622015337,
      "loss": 5.0971,
      "mean_token_accuracy": 0.17569808214902877,
      "num_tokens": 33609914.0,
      "step": 15350
    },
    {
      "entropy": 5.565807247161866,
      "epoch": 1.5239182215164748,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.1884,
      "mean_token_accuracy": 0.16434087753295898,
      "num_tokens": 33621134.0,
      "step": 15355
    },
    {
      "entropy": 5.593252325057984,
      "epoch": 1.5244144501786423,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16705452054738998,
      "num_tokens": 33631665.0,
      "step": 15360
    },
    {
      "entropy": 5.590158224105835,
      "epoch": 1.52491067884081,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004773661816024335,
      "loss": 5.1614,
      "mean_token_accuracy": 0.16521507352590561,
      "num_tokens": 33643161.0,
      "step": 15365
    },
    {
      "entropy": 5.591151857376099,
      "epoch": 1.5254069075029775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.1555,
      "mean_token_accuracy": 0.1680409938097,
      "num_tokens": 33654893.0,
      "step": 15370
    },
    {
      "entropy": 5.5371345520019535,
      "epoch": 1.525903136165145,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.1594,
      "mean_token_accuracy": 0.1635119155049324,
      "num_tokens": 33664688.0,
      "step": 15375
    },
    {
      "entropy": 5.530398273468018,
      "epoch": 1.5263993648273124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.1494,
      "mean_token_accuracy": 0.17213926315307618,
      "num_tokens": 33676084.0,
      "step": 15380
    },
    {
      "entropy": 5.603587484359741,
      "epoch": 1.52689559348948,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004773041907982561,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16321431696414948,
      "num_tokens": 33686946.0,
      "step": 15385
    },
    {
      "entropy": 5.531073522567749,
      "epoch": 1.5273918221516474,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17650165110826493,
      "num_tokens": 33697994.0,
      "step": 15390
    },
    {
      "entropy": 5.535098028182984,
      "epoch": 1.5278880508138148,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.184,
      "mean_token_accuracy": 0.16685742139816284,
      "num_tokens": 33709618.0,
      "step": 15395
    },
    {
      "entropy": 5.570789337158203,
      "epoch": 1.5283842794759825,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.1723,
      "mean_token_accuracy": 0.17637065201997756,
      "num_tokens": 33720282.0,
      "step": 15400
    },
    {
      "entropy": 5.534037828445435,
      "epoch": 1.52888050813815,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.1548,
      "mean_token_accuracy": 0.16516810208559035,
      "num_tokens": 33730581.0,
      "step": 15405
    },
    {
      "entropy": 5.598129987716675,
      "epoch": 1.5293767368003177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.1981,
      "mean_token_accuracy": 0.16923441737890244,
      "num_tokens": 33741732.0,
      "step": 15410
    },
    {
      "entropy": 5.636341142654419,
      "epoch": 1.5298729654624852,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.2287,
      "mean_token_accuracy": 0.16746146976947784,
      "num_tokens": 33753043.0,
      "step": 15415
    },
    {
      "entropy": 5.531795454025269,
      "epoch": 1.5303691941246527,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17612006962299348,
      "num_tokens": 33763974.0,
      "step": 15420
    },
    {
      "entropy": 5.521078586578369,
      "epoch": 1.5308654227868201,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.0521,
      "mean_token_accuracy": 0.17892154753208162,
      "num_tokens": 33775247.0,
      "step": 15425
    },
    {
      "entropy": 5.5863121509552,
      "epoch": 1.5313616514489876,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.1965,
      "mean_token_accuracy": 0.1777326613664627,
      "num_tokens": 33786843.0,
      "step": 15430
    },
    {
      "entropy": 5.538761758804322,
      "epoch": 1.531857880111155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.1049,
      "mean_token_accuracy": 0.17465351670980453,
      "num_tokens": 33797974.0,
      "step": 15435
    },
    {
      "entropy": 5.563059949874878,
      "epoch": 1.5323541087733228,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.1564,
      "mean_token_accuracy": 0.1694395586848259,
      "num_tokens": 33807965.0,
      "step": 15440
    },
    {
      "entropy": 5.548739385604859,
      "epoch": 1.5328503374354903,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.1482,
      "mean_token_accuracy": 0.17239087224006652,
      "num_tokens": 33819835.0,
      "step": 15445
    },
    {
      "entropy": 5.503639554977417,
      "epoch": 1.533346566097658,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17667288035154344,
      "num_tokens": 33830041.0,
      "step": 15450
    },
    {
      "entropy": 5.584738111495971,
      "epoch": 1.5338427947598254,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.1576,
      "mean_token_accuracy": 0.1644434615969658,
      "num_tokens": 33841681.0,
      "step": 15455
    },
    {
      "entropy": 5.542841291427612,
      "epoch": 1.534339023421993,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.1203,
      "mean_token_accuracy": 0.16787735968828202,
      "num_tokens": 33852411.0,
      "step": 15460
    },
    {
      "entropy": 5.543913793563843,
      "epoch": 1.5348352520841604,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17359764277935028,
      "num_tokens": 33862130.0,
      "step": 15465
    },
    {
      "entropy": 5.5496134757995605,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.1746,
      "mean_token_accuracy": 0.16826476603746415,
      "num_tokens": 33872308.0,
      "step": 15470
    },
    {
      "entropy": 5.614943313598633,
      "epoch": 1.5358277094084953,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.122,
      "mean_token_accuracy": 0.1706447795033455,
      "num_tokens": 33882704.0,
      "step": 15475
    },
    {
      "entropy": 5.558654546737671,
      "epoch": 1.5363239380706628,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.1797,
      "mean_token_accuracy": 0.16936301440000534,
      "num_tokens": 33893796.0,
      "step": 15480
    },
    {
      "entropy": 5.47291316986084,
      "epoch": 1.5368201667328305,
      "grad_norm": 1.015625,
      "learning_rate": 0.000476993033260516,
      "loss": 5.0639,
      "mean_token_accuracy": 0.17623965442180634,
      "num_tokens": 33903723.0,
      "step": 15485
    },
    {
      "entropy": 5.516610431671142,
      "epoch": 1.537316395394998,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.0449,
      "mean_token_accuracy": 0.18071333765983583,
      "num_tokens": 33914295.0,
      "step": 15490
    },
    {
      "entropy": 5.718691873550415,
      "epoch": 1.5378126240571657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1701073929667473,
      "num_tokens": 33925414.0,
      "step": 15495
    },
    {
      "entropy": 5.58011417388916,
      "epoch": 1.5383088527193332,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.1379,
      "mean_token_accuracy": 0.16405633389949797,
      "num_tokens": 33936878.0,
      "step": 15500
    },
    {
      "entropy": 5.510344076156616,
      "epoch": 1.5388050813815006,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.1075,
      "mean_token_accuracy": 0.16689598709344863,
      "num_tokens": 33947684.0,
      "step": 15505
    },
    {
      "entropy": 5.588282108306885,
      "epoch": 1.539301310043668,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.1687,
      "mean_token_accuracy": 0.17027779072523117,
      "num_tokens": 33958102.0,
      "step": 15510
    },
    {
      "entropy": 5.6110011577606205,
      "epoch": 1.5397975387058356,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.1646,
      "mean_token_accuracy": 0.1644342914223671,
      "num_tokens": 33968606.0,
      "step": 15515
    },
    {
      "entropy": 5.555943012237549,
      "epoch": 1.540293767368003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.0792,
      "mean_token_accuracy": 0.18280257880687714,
      "num_tokens": 33978859.0,
      "step": 15520
    },
    {
      "entropy": 5.51783766746521,
      "epoch": 1.5407899960301707,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.1672,
      "mean_token_accuracy": 0.17321260869503022,
      "num_tokens": 33991056.0,
      "step": 15525
    },
    {
      "entropy": 5.530650568008423,
      "epoch": 1.5412862246923382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.1563,
      "mean_token_accuracy": 0.17044303864240645,
      "num_tokens": 34001257.0,
      "step": 15530
    },
    {
      "entropy": 5.552424573898316,
      "epoch": 1.541782453354506,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1705116629600525,
      "num_tokens": 34012881.0,
      "step": 15535
    },
    {
      "entropy": 5.5938009262084964,
      "epoch": 1.5422786820166734,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.1309,
      "mean_token_accuracy": 0.17061381414532661,
      "num_tokens": 34023503.0,
      "step": 15540
    },
    {
      "entropy": 5.630060148239136,
      "epoch": 1.5427749106788409,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.1895,
      "mean_token_accuracy": 0.17193544805049896,
      "num_tokens": 34033137.0,
      "step": 15545
    },
    {
      "entropy": 5.549626731872559,
      "epoch": 1.5432711393410083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17453866750001906,
      "num_tokens": 34044584.0,
      "step": 15550
    },
    {
      "entropy": 5.592932844161988,
      "epoch": 1.5437673680031758,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.1146,
      "mean_token_accuracy": 0.16939475238323212,
      "num_tokens": 34055575.0,
      "step": 15555
    },
    {
      "entropy": 5.5567848682403564,
      "epoch": 1.5442635966653433,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.1862,
      "mean_token_accuracy": 0.17123686969280244,
      "num_tokens": 34067036.0,
      "step": 15560
    },
    {
      "entropy": 5.580525255203247,
      "epoch": 1.5447598253275108,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.1592,
      "mean_token_accuracy": 0.1642957329750061,
      "num_tokens": 34078705.0,
      "step": 15565
    },
    {
      "entropy": 5.599431324005127,
      "epoch": 1.5452560539896785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.2133,
      "mean_token_accuracy": 0.1679215610027313,
      "num_tokens": 34089437.0,
      "step": 15570
    },
    {
      "entropy": 5.574877023696899,
      "epoch": 1.545752282651846,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.167,
      "mean_token_accuracy": 0.17329792976379393,
      "num_tokens": 34101714.0,
      "step": 15575
    },
    {
      "entropy": 5.560558795928955,
      "epoch": 1.5462485113140136,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000476695577721796,
      "loss": 5.1785,
      "mean_token_accuracy": 0.16942770183086395,
      "num_tokens": 34112596.0,
      "step": 15580
    },
    {
      "entropy": 5.5641889572143555,
      "epoch": 1.5467447399761811,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.0614,
      "mean_token_accuracy": 0.1779106765985489,
      "num_tokens": 34122777.0,
      "step": 15585
    },
    {
      "entropy": 5.566963148117066,
      "epoch": 1.5472409686383486,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.1874,
      "mean_token_accuracy": 0.17185969799757003,
      "num_tokens": 34133887.0,
      "step": 15590
    },
    {
      "entropy": 5.564784145355224,
      "epoch": 1.547737197300516,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.1936,
      "mean_token_accuracy": 0.16507910639047624,
      "num_tokens": 34144494.0,
      "step": 15595
    },
    {
      "entropy": 5.600232315063477,
      "epoch": 1.5482334259626835,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.193,
      "mean_token_accuracy": 0.1723092809319496,
      "num_tokens": 34156604.0,
      "step": 15600
    },
    {
      "entropy": 5.544495248794556,
      "epoch": 1.548729654624851,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1690809518098831,
      "num_tokens": 34167342.0,
      "step": 15605
    },
    {
      "entropy": 5.590365028381347,
      "epoch": 1.5492258832870187,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17457469552755356,
      "num_tokens": 34178611.0,
      "step": 15610
    },
    {
      "entropy": 5.546537590026856,
      "epoch": 1.5497221119491862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.0959,
      "mean_token_accuracy": 0.1693306878209114,
      "num_tokens": 34189658.0,
      "step": 15615
    },
    {
      "entropy": 5.497526741027832,
      "epoch": 1.5502183406113537,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1753316342830658,
      "num_tokens": 34200080.0,
      "step": 15620
    },
    {
      "entropy": 5.5555400371551515,
      "epoch": 1.5507145692735214,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.153,
      "mean_token_accuracy": 0.17906749993562698,
      "num_tokens": 34210184.0,
      "step": 15625
    },
    {
      "entropy": 5.600201511383057,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.1576,
      "mean_token_accuracy": 0.1742034822702408,
      "num_tokens": 34220448.0,
      "step": 15630
    },
    {
      "entropy": 5.52303295135498,
      "epoch": 1.5517070265978563,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.1552,
      "mean_token_accuracy": 0.17226925641298294,
      "num_tokens": 34231595.0,
      "step": 15635
    },
    {
      "entropy": 5.563747978210449,
      "epoch": 1.5522032552600238,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.1592,
      "mean_token_accuracy": 0.16784761548042298,
      "num_tokens": 34244066.0,
      "step": 15640
    },
    {
      "entropy": 5.523954391479492,
      "epoch": 1.5526994839221913,
      "grad_norm": 1.0,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.162,
      "mean_token_accuracy": 0.1666826218366623,
      "num_tokens": 34255505.0,
      "step": 15645
    },
    {
      "entropy": 5.5645464897155765,
      "epoch": 1.5531957125843587,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.0583,
      "mean_token_accuracy": 0.17606594562530517,
      "num_tokens": 34265386.0,
      "step": 15650
    },
    {
      "entropy": 5.547397327423096,
      "epoch": 1.5536919412465264,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.1155,
      "mean_token_accuracy": 0.1745871916413307,
      "num_tokens": 34278317.0,
      "step": 15655
    },
    {
      "entropy": 5.601050996780396,
      "epoch": 1.554188169908694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.2121,
      "mean_token_accuracy": 0.16379949152469636,
      "num_tokens": 34290814.0,
      "step": 15660
    },
    {
      "entropy": 5.5828345775604244,
      "epoch": 1.5546843985708616,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.1899,
      "mean_token_accuracy": 0.17162858098745346,
      "num_tokens": 34302819.0,
      "step": 15665
    },
    {
      "entropy": 5.512557554244995,
      "epoch": 1.555180627233029,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.1281,
      "mean_token_accuracy": 0.17675762623548508,
      "num_tokens": 34314055.0,
      "step": 15670
    },
    {
      "entropy": 5.551833534240723,
      "epoch": 1.5556768558951966,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.077,
      "mean_token_accuracy": 0.18261767774820328,
      "num_tokens": 34324826.0,
      "step": 15675
    },
    {
      "entropy": 5.589658260345459,
      "epoch": 1.556173084557364,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17956203669309617,
      "num_tokens": 34334540.0,
      "step": 15680
    },
    {
      "entropy": 5.529135704040527,
      "epoch": 1.5566693132195315,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.1191,
      "mean_token_accuracy": 0.1727798566222191,
      "num_tokens": 34344501.0,
      "step": 15685
    },
    {
      "entropy": 5.525330924987793,
      "epoch": 1.557165541881699,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.0969,
      "mean_token_accuracy": 0.17129005044698714,
      "num_tokens": 34355207.0,
      "step": 15690
    },
    {
      "entropy": 5.573146867752075,
      "epoch": 1.5576617705438665,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.1606,
      "mean_token_accuracy": 0.16888036727905273,
      "num_tokens": 34366839.0,
      "step": 15695
    },
    {
      "entropy": 5.642684602737427,
      "epoch": 1.5581579992060341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.281,
      "mean_token_accuracy": 0.17238802313804627,
      "num_tokens": 34377863.0,
      "step": 15700
    },
    {
      "entropy": 5.688420629501342,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.0,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.2661,
      "mean_token_accuracy": 0.15733095705509187,
      "num_tokens": 34390490.0,
      "step": 15705
    },
    {
      "entropy": 5.537191152572632,
      "epoch": 1.5591504565303693,
      "grad_norm": 0.875,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.11,
      "mean_token_accuracy": 0.17391955852508545,
      "num_tokens": 34401979.0,
      "step": 15710
    },
    {
      "entropy": 5.550140428543091,
      "epoch": 1.5596466851925368,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.178,
      "mean_token_accuracy": 0.16616002321243287,
      "num_tokens": 34413470.0,
      "step": 15715
    },
    {
      "entropy": 5.66965913772583,
      "epoch": 1.5601429138547043,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16978975534439086,
      "num_tokens": 34424805.0,
      "step": 15720
    },
    {
      "entropy": 5.5640708923339846,
      "epoch": 1.5606391425168717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17576806694269181,
      "num_tokens": 34433994.0,
      "step": 15725
    },
    {
      "entropy": 5.560881662368774,
      "epoch": 1.5611353711790392,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.1458,
      "mean_token_accuracy": 0.16550110578536986,
      "num_tokens": 34444230.0,
      "step": 15730
    },
    {
      "entropy": 5.5187279224395756,
      "epoch": 1.5616315998412067,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.0507,
      "mean_token_accuracy": 0.17704780995845795,
      "num_tokens": 34453424.0,
      "step": 15735
    },
    {
      "entropy": 5.58669114112854,
      "epoch": 1.5621278285033744,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.1471,
      "mean_token_accuracy": 0.16934033036231994,
      "num_tokens": 34464283.0,
      "step": 15740
    },
    {
      "entropy": 5.529719495773316,
      "epoch": 1.5626240571655419,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.0353,
      "mean_token_accuracy": 0.17911873906850814,
      "num_tokens": 34475397.0,
      "step": 15745
    },
    {
      "entropy": 5.54973669052124,
      "epoch": 1.5631202858277096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.1382,
      "mean_token_accuracy": 0.1755244702100754,
      "num_tokens": 34486534.0,
      "step": 15750
    },
    {
      "entropy": 5.537407398223877,
      "epoch": 1.563616514489877,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000476142908753089,
      "loss": 5.2045,
      "mean_token_accuracy": 0.16967225819826126,
      "num_tokens": 34498245.0,
      "step": 15755
    },
    {
      "entropy": 5.6115011215209964,
      "epoch": 1.5641127431520445,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.2166,
      "mean_token_accuracy": 0.16684400141239167,
      "num_tokens": 34509502.0,
      "step": 15760
    },
    {
      "entropy": 5.588599395751953,
      "epoch": 1.564608971814212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18621499836444855,
      "num_tokens": 34519754.0,
      "step": 15765
    },
    {
      "entropy": 5.574887275695801,
      "epoch": 1.5651052004763795,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.122,
      "mean_token_accuracy": 0.17285457551479338,
      "num_tokens": 34530952.0,
      "step": 15770
    },
    {
      "entropy": 5.4881195545196535,
      "epoch": 1.565601429138547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.0891,
      "mean_token_accuracy": 0.16801812797784804,
      "num_tokens": 34540780.0,
      "step": 15775
    },
    {
      "entropy": 5.536787414550782,
      "epoch": 1.5660976578007144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.1746,
      "mean_token_accuracy": 0.16900371760129929,
      "num_tokens": 34550828.0,
      "step": 15780
    },
    {
      "entropy": 5.558574056625366,
      "epoch": 1.5665938864628821,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.0495,
      "mean_token_accuracy": 0.17537857145071029,
      "num_tokens": 34561842.0,
      "step": 15785
    },
    {
      "entropy": 5.5459557056427,
      "epoch": 1.5670901151250496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17557242661714553,
      "num_tokens": 34573894.0,
      "step": 15790
    },
    {
      "entropy": 5.494121646881103,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.141,
      "mean_token_accuracy": 0.1677237033843994,
      "num_tokens": 34583003.0,
      "step": 15795
    },
    {
      "entropy": 5.480137538909912,
      "epoch": 1.5680825724493848,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047599980559546757,
      "loss": 5.0159,
      "mean_token_accuracy": 0.17718454599380493,
      "num_tokens": 34594912.0,
      "step": 15800
    },
    {
      "entropy": 5.572283506393433,
      "epoch": 1.5685788011115522,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.162,
      "mean_token_accuracy": 0.17084705382585524,
      "num_tokens": 34606638.0,
      "step": 15805
    },
    {
      "entropy": 5.514507722854614,
      "epoch": 1.5690750297737197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17732035666704177,
      "num_tokens": 34617438.0,
      "step": 15810
    },
    {
      "entropy": 5.560735559463501,
      "epoch": 1.5695712584358872,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.1752,
      "mean_token_accuracy": 0.1651997908949852,
      "num_tokens": 34627857.0,
      "step": 15815
    },
    {
      "entropy": 5.512501859664917,
      "epoch": 1.5700674870980547,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.0608,
      "mean_token_accuracy": 0.17504971474409103,
      "num_tokens": 34638243.0,
      "step": 15820
    },
    {
      "entropy": 5.563844203948975,
      "epoch": 1.5705637157602224,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.0991,
      "mean_token_accuracy": 0.169374880194664,
      "num_tokens": 34648355.0,
      "step": 15825
    },
    {
      "entropy": 5.504838514328003,
      "epoch": 1.5710599444223898,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047590417917635423,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1774387314915657,
      "num_tokens": 34658644.0,
      "step": 15830
    },
    {
      "entropy": 5.4788525104522705,
      "epoch": 1.5715561730845575,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1709710404276848,
      "num_tokens": 34669892.0,
      "step": 15835
    },
    {
      "entropy": 5.632106018066406,
      "epoch": 1.572052401746725,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.162,
      "mean_token_accuracy": 0.17143699228763581,
      "num_tokens": 34681197.0,
      "step": 15840
    },
    {
      "entropy": 5.58644380569458,
      "epoch": 1.5725486304088925,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.1588,
      "mean_token_accuracy": 0.16603389233350754,
      "num_tokens": 34692070.0,
      "step": 15845
    },
    {
      "entropy": 5.554009246826172,
      "epoch": 1.57304485907106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.1993,
      "mean_token_accuracy": 0.16704439818859101,
      "num_tokens": 34702667.0,
      "step": 15850
    },
    {
      "entropy": 5.512634325027466,
      "epoch": 1.5735410877332274,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17666802555322647,
      "num_tokens": 34713936.0,
      "step": 15855
    },
    {
      "entropy": 5.623472404479981,
      "epoch": 1.574037316395395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.1538,
      "mean_token_accuracy": 0.17577076703310013,
      "num_tokens": 34724195.0,
      "step": 15860
    },
    {
      "entropy": 5.602540588378906,
      "epoch": 1.5745335450575624,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.1441,
      "mean_token_accuracy": 0.1660414829850197,
      "num_tokens": 34735932.0,
      "step": 15865
    },
    {
      "entropy": 5.635334110260009,
      "epoch": 1.57502977371973,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.2442,
      "mean_token_accuracy": 0.16334135979413986,
      "num_tokens": 34747451.0,
      "step": 15870
    },
    {
      "entropy": 5.608651065826416,
      "epoch": 1.5755260023818976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17487634271383284,
      "num_tokens": 34757591.0,
      "step": 15875
    },
    {
      "entropy": 5.554347133636474,
      "epoch": 1.5760222310440652,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000475744403294165,
      "loss": 5.1383,
      "mean_token_accuracy": 0.17464955598115922,
      "num_tokens": 34768420.0,
      "step": 15880
    },
    {
      "entropy": 5.566205310821533,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.199,
      "mean_token_accuracy": 0.1693756699562073,
      "num_tokens": 34778984.0,
      "step": 15885
    },
    {
      "entropy": 5.565372896194458,
      "epoch": 1.5770146883684002,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1632111817598343,
      "num_tokens": 34790913.0,
      "step": 15890
    },
    {
      "entropy": 5.530621814727783,
      "epoch": 1.5775109170305677,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.1198,
      "mean_token_accuracy": 0.1711557239294052,
      "num_tokens": 34801873.0,
      "step": 15895
    },
    {
      "entropy": 5.684178638458252,
      "epoch": 1.5780071456927351,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.3451,
      "mean_token_accuracy": 0.16494930386543274,
      "num_tokens": 34812871.0,
      "step": 15900
    },
    {
      "entropy": 5.581216049194336,
      "epoch": 1.5785033743549026,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.17,
      "mean_token_accuracy": 0.1770874708890915,
      "num_tokens": 34824405.0,
      "step": 15905
    },
    {
      "entropy": 5.568099594116211,
      "epoch": 1.5789996030170703,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.1954,
      "mean_token_accuracy": 0.16722864806652069,
      "num_tokens": 34834048.0,
      "step": 15910
    },
    {
      "entropy": 5.58236665725708,
      "epoch": 1.5794958316792378,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.156,
      "mean_token_accuracy": 0.16762006133794785,
      "num_tokens": 34844915.0,
      "step": 15915
    },
    {
      "entropy": 5.543608951568603,
      "epoch": 1.5799920603414053,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.1007,
      "mean_token_accuracy": 0.16643791198730468,
      "num_tokens": 34855282.0,
      "step": 15920
    },
    {
      "entropy": 5.566107177734375,
      "epoch": 1.580488289003573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1735369011759758,
      "num_tokens": 34865541.0,
      "step": 15925
    },
    {
      "entropy": 5.566701364517212,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.1543,
      "mean_token_accuracy": 0.16991167962551118,
      "num_tokens": 34876326.0,
      "step": 15930
    },
    {
      "entropy": 5.589658594131469,
      "epoch": 1.581480746327908,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.1734,
      "mean_token_accuracy": 0.17049886882305146,
      "num_tokens": 34887319.0,
      "step": 15935
    },
    {
      "entropy": 5.62866621017456,
      "epoch": 1.5819769749900754,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.2306,
      "mean_token_accuracy": 0.1720321372151375,
      "num_tokens": 34900511.0,
      "step": 15940
    },
    {
      "entropy": 5.577267456054687,
      "epoch": 1.5824732036522429,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.102,
      "mean_token_accuracy": 0.16747928708791732,
      "num_tokens": 34910037.0,
      "step": 15945
    },
    {
      "entropy": 5.495161676406861,
      "epoch": 1.5829694323144103,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.089,
      "mean_token_accuracy": 0.17572314590215682,
      "num_tokens": 34921618.0,
      "step": 15950
    },
    {
      "entropy": 5.56949896812439,
      "epoch": 1.583465660976578,
      "grad_norm": 0.90625,
      "learning_rate": 0.000475503806214856,
      "loss": 5.1991,
      "mean_token_accuracy": 0.16458560079336165,
      "num_tokens": 34932010.0,
      "step": 15955
    },
    {
      "entropy": 5.534655380249023,
      "epoch": 1.5839618896387455,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17335041016340255,
      "num_tokens": 34942252.0,
      "step": 15960
    },
    {
      "entropy": 5.607513856887818,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475471642049887,
      "loss": 5.1638,
      "mean_token_accuracy": 0.1718481093645096,
      "num_tokens": 34953343.0,
      "step": 15965
    },
    {
      "entropy": 5.619128465652466,
      "epoch": 1.5849543469630807,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.1109,
      "mean_token_accuracy": 0.17458790838718413,
      "num_tokens": 34964537.0,
      "step": 15970
    },
    {
      "entropy": 5.60925030708313,
      "epoch": 1.5854505756252482,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000475439457999859,
      "loss": 5.1919,
      "mean_token_accuracy": 0.16853196620941163,
      "num_tokens": 34973655.0,
      "step": 15975
    },
    {
      "entropy": 5.566146373748779,
      "epoch": 1.5859468042874156,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.1889,
      "mean_token_accuracy": 0.17314601838588714,
      "num_tokens": 34984826.0,
      "step": 15980
    },
    {
      "entropy": 5.5629720211029055,
      "epoch": 1.586443032949583,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17199254184961318,
      "num_tokens": 34995492.0,
      "step": 15985
    },
    {
      "entropy": 5.527230596542358,
      "epoch": 1.5869392616117506,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17511005997657775,
      "num_tokens": 35007617.0,
      "step": 15990
    },
    {
      "entropy": 5.6680638790130615,
      "epoch": 1.587435490273918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000475375030257398,
      "loss": 5.2473,
      "mean_token_accuracy": 0.16354547813534737,
      "num_tokens": 35018022.0,
      "step": 15995
    },
    {
      "entropy": 5.6065216064453125,
      "epoch": 1.5879317189360858,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.1858,
      "mean_token_accuracy": 0.17051806449890136,
      "num_tokens": 35028725.0,
      "step": 16000
    },
    {
      "entropy": 5.648677968978882,
      "epoch": 1.5884279475982532,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.1882,
      "mean_token_accuracy": 0.17366128265857697,
      "num_tokens": 35039067.0,
      "step": 16005
    },
    {
      "entropy": 5.589757251739502,
      "epoch": 1.588924176260421,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.1193,
      "mean_token_accuracy": 0.1717248111963272,
      "num_tokens": 35051292.0,
      "step": 16010
    },
    {
      "entropy": 5.5943012714385985,
      "epoch": 1.5894204049225884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.139,
      "mean_token_accuracy": 0.16765307784080505,
      "num_tokens": 35062046.0,
      "step": 16015
    },
    {
      "entropy": 5.607323884963989,
      "epoch": 1.5899166335847559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.1988,
      "mean_token_accuracy": 0.16944422274827958,
      "num_tokens": 35072275.0,
      "step": 16020
    },
    {
      "entropy": 5.537097406387329,
      "epoch": 1.5904128622469234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1693381518125534,
      "num_tokens": 35083173.0,
      "step": 16025
    },
    {
      "entropy": 5.554601430892944,
      "epoch": 1.5909090909090908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.107,
      "mean_token_accuracy": 0.17372595220804216,
      "num_tokens": 35094323.0,
      "step": 16030
    },
    {
      "entropy": 5.5238124370574955,
      "epoch": 1.5914053195712583,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.1071,
      "mean_token_accuracy": 0.17000172436237335,
      "num_tokens": 35105216.0,
      "step": 16035
    },
    {
      "entropy": 5.575453853607177,
      "epoch": 1.591901548233426,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.1443,
      "mean_token_accuracy": 0.17014622390270234,
      "num_tokens": 35116283.0,
      "step": 16040
    },
    {
      "entropy": 5.631569766998291,
      "epoch": 1.5923977768955935,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.1398,
      "mean_token_accuracy": 0.16856106370687485,
      "num_tokens": 35126821.0,
      "step": 16045
    },
    {
      "entropy": 5.584452295303345,
      "epoch": 1.5928940055577612,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.1318,
      "mean_token_accuracy": 0.16946475505828856,
      "num_tokens": 35137860.0,
      "step": 16050
    },
    {
      "entropy": 5.550068664550781,
      "epoch": 1.5933902342199286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.1188,
      "mean_token_accuracy": 0.1760944679379463,
      "num_tokens": 35149103.0,
      "step": 16055
    },
    {
      "entropy": 5.6125232696533205,
      "epoch": 1.5938864628820961,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.1739,
      "mean_token_accuracy": 0.16938043385744095,
      "num_tokens": 35158981.0,
      "step": 16060
    },
    {
      "entropy": 5.641307830810547,
      "epoch": 1.5943826915442636,
      "grad_norm": 0.875,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.1575,
      "mean_token_accuracy": 0.17334930896759032,
      "num_tokens": 35171189.0,
      "step": 16065
    },
    {
      "entropy": 5.608633375167846,
      "epoch": 1.594878920206431,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1637456715106964,
      "num_tokens": 35185034.0,
      "step": 16070
    },
    {
      "entropy": 5.623081684112549,
      "epoch": 1.5953751488685985,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.2388,
      "mean_token_accuracy": 0.1639787569642067,
      "num_tokens": 35196106.0,
      "step": 16075
    },
    {
      "entropy": 5.536793565750122,
      "epoch": 1.595871377530766,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.1074,
      "mean_token_accuracy": 0.16945667713880538,
      "num_tokens": 35206736.0,
      "step": 16080
    },
    {
      "entropy": 5.4802933692932125,
      "epoch": 1.5963676061929337,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.0849,
      "mean_token_accuracy": 0.17439293414354323,
      "num_tokens": 35217279.0,
      "step": 16085
    },
    {
      "entropy": 5.527122592926025,
      "epoch": 1.5968638348551012,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.0665,
      "mean_token_accuracy": 0.16856844276189803,
      "num_tokens": 35227443.0,
      "step": 16090
    },
    {
      "entropy": 5.549749135971069,
      "epoch": 1.597360063517269,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.1568,
      "mean_token_accuracy": 0.1634203851222992,
      "num_tokens": 35238107.0,
      "step": 16095
    },
    {
      "entropy": 5.565900373458862,
      "epoch": 1.5978562921794364,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.1563,
      "mean_token_accuracy": 0.16927742660045625,
      "num_tokens": 35247960.0,
      "step": 16100
    },
    {
      "entropy": 5.635844278335571,
      "epoch": 1.5983525208416038,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.243,
      "mean_token_accuracy": 0.16321902126073837,
      "num_tokens": 35259560.0,
      "step": 16105
    },
    {
      "entropy": 5.594776010513305,
      "epoch": 1.5988487495037713,
      "grad_norm": 1.0,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17410812377929688,
      "num_tokens": 35270363.0,
      "step": 16110
    },
    {
      "entropy": 5.6355362892150875,
      "epoch": 1.5993449781659388,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1727300316095352,
      "num_tokens": 35281282.0,
      "step": 16115
    },
    {
      "entropy": 5.609845161437988,
      "epoch": 1.5998412068281063,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.1687,
      "mean_token_accuracy": 0.17779066562652587,
      "num_tokens": 35291916.0,
      "step": 16120
    },
    {
      "entropy": 5.546169805526733,
      "epoch": 1.600337435490274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.17,
      "mean_token_accuracy": 0.16773585677146913,
      "num_tokens": 35302493.0,
      "step": 16125
    },
    {
      "entropy": 5.557813692092895,
      "epoch": 1.6008336641524414,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000474938064809171,
      "loss": 5.1167,
      "mean_token_accuracy": 0.17052248120307922,
      "num_tokens": 35313364.0,
      "step": 16130
    },
    {
      "entropy": 5.553959083557129,
      "epoch": 1.6013298928146091,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17559956908226013,
      "num_tokens": 35324072.0,
      "step": 16135
    },
    {
      "entropy": 5.572145891189575,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.1758,
      "mean_token_accuracy": 0.16758000552654267,
      "num_tokens": 35335290.0,
      "step": 16140
    },
    {
      "entropy": 5.546107530593872,
      "epoch": 1.602322350138944,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.1555,
      "mean_token_accuracy": 0.17290777564048768,
      "num_tokens": 35347189.0,
      "step": 16145
    },
    {
      "entropy": 5.556348419189453,
      "epoch": 1.6028185788011116,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047487302159800545,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1727117195725441,
      "num_tokens": 35360095.0,
      "step": 16150
    },
    {
      "entropy": 5.5642599105834964,
      "epoch": 1.603314807463279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000474856748403155,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17190011590719223,
      "num_tokens": 35370075.0,
      "step": 16155
    },
    {
      "entropy": 5.542391157150268,
      "epoch": 1.6038110361254465,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17999490797519685,
      "num_tokens": 35381647.0,
      "step": 16160
    },
    {
      "entropy": 5.558360624313354,
      "epoch": 1.604307264787614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18228810876607895,
      "num_tokens": 35391629.0,
      "step": 16165
    },
    {
      "entropy": 5.529161643981934,
      "epoch": 1.6048034934497817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.1283,
      "mean_token_accuracy": 0.17277828007936477,
      "num_tokens": 35401385.0,
      "step": 16170
    },
    {
      "entropy": 5.5468226909637455,
      "epoch": 1.6052997221119492,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.0984,
      "mean_token_accuracy": 0.1746475249528885,
      "num_tokens": 35414289.0,
      "step": 16175
    },
    {
      "entropy": 5.5038971424102785,
      "epoch": 1.6057959507741169,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17246668338775634,
      "num_tokens": 35424397.0,
      "step": 16180
    },
    {
      "entropy": 5.61233811378479,
      "epoch": 1.6062921794362843,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.1913,
      "mean_token_accuracy": 0.17515736669301987,
      "num_tokens": 35435897.0,
      "step": 16185
    },
    {
      "entropy": 5.570881319046021,
      "epoch": 1.6067884080984518,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.1368,
      "mean_token_accuracy": 0.17160869091749192,
      "num_tokens": 35446794.0,
      "step": 16190
    },
    {
      "entropy": 5.5047258853912355,
      "epoch": 1.6072846367606193,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.0596,
      "mean_token_accuracy": 0.17391697764396669,
      "num_tokens": 35456854.0,
      "step": 16195
    },
    {
      "entropy": 5.54476957321167,
      "epoch": 1.6077808654227868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.1331,
      "mean_token_accuracy": 0.17030350863933563,
      "num_tokens": 35467924.0,
      "step": 16200
    },
    {
      "entropy": 5.512268543243408,
      "epoch": 1.6082770940849542,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000474693743940377,
      "loss": 5.0403,
      "mean_token_accuracy": 0.17718614488840104,
      "num_tokens": 35479290.0,
      "step": 16205
    },
    {
      "entropy": 5.56484899520874,
      "epoch": 1.608773322747122,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.1272,
      "mean_token_accuracy": 0.17352051436901092,
      "num_tokens": 35489477.0,
      "step": 16210
    },
    {
      "entropy": 5.652667617797851,
      "epoch": 1.6092695514092894,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.2804,
      "mean_token_accuracy": 0.16263461112976074,
      "num_tokens": 35501531.0,
      "step": 16215
    },
    {
      "entropy": 5.682536697387695,
      "epoch": 1.6097657800714569,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.1947,
      "mean_token_accuracy": 0.16438548266887665,
      "num_tokens": 35513325.0,
      "step": 16220
    },
    {
      "entropy": 5.578372478485107,
      "epoch": 1.6102620087336246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.1104,
      "mean_token_accuracy": 0.1732337072491646,
      "num_tokens": 35524683.0,
      "step": 16225
    },
    {
      "entropy": 5.479498815536499,
      "epoch": 1.610758237395792,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.0409,
      "mean_token_accuracy": 0.17952865362167358,
      "num_tokens": 35534179.0,
      "step": 16230
    },
    {
      "entropy": 5.514772939682007,
      "epoch": 1.6112544660579595,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17262389063835143,
      "num_tokens": 35544894.0,
      "step": 16235
    },
    {
      "entropy": 5.526379632949829,
      "epoch": 1.611750694720127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.1061,
      "mean_token_accuracy": 0.17255004793405532,
      "num_tokens": 35556915.0,
      "step": 16240
    },
    {
      "entropy": 5.539622354507446,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.1372,
      "mean_token_accuracy": 0.18155924528837203,
      "num_tokens": 35568037.0,
      "step": 16245
    },
    {
      "entropy": 5.609771394729615,
      "epoch": 1.612743152044462,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.167,
      "mean_token_accuracy": 0.1666193574666977,
      "num_tokens": 35578969.0,
      "step": 16250
    },
    {
      "entropy": 5.563062477111816,
      "epoch": 1.6132393807066296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.077,
      "mean_token_accuracy": 0.17654875814914703,
      "num_tokens": 35588637.0,
      "step": 16255
    },
    {
      "entropy": 5.539796543121338,
      "epoch": 1.6137356093687971,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.1756,
      "mean_token_accuracy": 0.1781233474612236,
      "num_tokens": 35600865.0,
      "step": 16260
    },
    {
      "entropy": 5.584414863586426,
      "epoch": 1.6142318380309648,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.1791,
      "mean_token_accuracy": 0.16325150579214096,
      "num_tokens": 35612440.0,
      "step": 16265
    },
    {
      "entropy": 5.611085653305054,
      "epoch": 1.6147280666931323,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.155,
      "mean_token_accuracy": 0.16747966408729553,
      "num_tokens": 35623143.0,
      "step": 16270
    },
    {
      "entropy": 5.603085708618164,
      "epoch": 1.6152242953552998,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.1173,
      "mean_token_accuracy": 0.17767415046691895,
      "num_tokens": 35636245.0,
      "step": 16275
    },
    {
      "entropy": 5.589819812774659,
      "epoch": 1.6157205240174672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.1788,
      "mean_token_accuracy": 0.17004350274801255,
      "num_tokens": 35646954.0,
      "step": 16280
    },
    {
      "entropy": 5.533909702301026,
      "epoch": 1.6162167526796347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17897745370864868,
      "num_tokens": 35657822.0,
      "step": 16285
    },
    {
      "entropy": 5.5564830780029295,
      "epoch": 1.6167129813418022,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.1556,
      "mean_token_accuracy": 0.1701216831803322,
      "num_tokens": 35669306.0,
      "step": 16290
    },
    {
      "entropy": 5.546204423904419,
      "epoch": 1.6172092100039697,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.1596,
      "mean_token_accuracy": 0.1729285940527916,
      "num_tokens": 35681701.0,
      "step": 16295
    },
    {
      "entropy": 5.6116949081420895,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1739622712135315,
      "num_tokens": 35692561.0,
      "step": 16300
    },
    {
      "entropy": 5.62092809677124,
      "epoch": 1.6182016673283048,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.2034,
      "mean_token_accuracy": 0.1663534864783287,
      "num_tokens": 35703097.0,
      "step": 16305
    },
    {
      "entropy": 5.5829089164733885,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17447851300239564,
      "num_tokens": 35713010.0,
      "step": 16310
    },
    {
      "entropy": 5.618849039077759,
      "epoch": 1.61919412465264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.2373,
      "mean_token_accuracy": 0.16886207461357117,
      "num_tokens": 35724240.0,
      "step": 16315
    },
    {
      "entropy": 5.601583862304688,
      "epoch": 1.6196903533148075,
      "grad_norm": 0.875,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.2039,
      "mean_token_accuracy": 0.16837774962186813,
      "num_tokens": 35735022.0,
      "step": 16320
    },
    {
      "entropy": 5.628663444519043,
      "epoch": 1.620186581976975,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.2668,
      "mean_token_accuracy": 0.1646926760673523,
      "num_tokens": 35745994.0,
      "step": 16325
    },
    {
      "entropy": 5.62616982460022,
      "epoch": 1.6206828106391424,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.1946,
      "mean_token_accuracy": 0.16699250936508178,
      "num_tokens": 35756611.0,
      "step": 16330
    },
    {
      "entropy": 5.561268281936646,
      "epoch": 1.62117903930131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474267615825987,
      "loss": 5.0998,
      "mean_token_accuracy": 0.17533102929592131,
      "num_tokens": 35767155.0,
      "step": 16335
    },
    {
      "entropy": 5.579569339752197,
      "epoch": 1.6216752679634776,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.1091,
      "mean_token_accuracy": 0.1804988831281662,
      "num_tokens": 35777763.0,
      "step": 16340
    },
    {
      "entropy": 5.479135370254516,
      "epoch": 1.622171496625645,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000474234698289638,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17669909447431564,
      "num_tokens": 35787730.0,
      "step": 16345
    },
    {
      "entropy": 5.5520142078399655,
      "epoch": 1.6226677252878128,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.1345,
      "mean_token_accuracy": 0.16960862874984742,
      "num_tokens": 35798959.0,
      "step": 16350
    },
    {
      "entropy": 5.558948183059693,
      "epoch": 1.6231639539499803,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004742017609909242,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1772479832172394,
      "num_tokens": 35808426.0,
      "step": 16355
    },
    {
      "entropy": 5.481992101669311,
      "epoch": 1.6236601826121477,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.0313,
      "mean_token_accuracy": 0.17902833968400955,
      "num_tokens": 35818798.0,
      "step": 16360
    },
    {
      "entropy": 5.596526718139648,
      "epoch": 1.6241564112743152,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.208,
      "mean_token_accuracy": 0.16899891197681427,
      "num_tokens": 35830429.0,
      "step": 16365
    },
    {
      "entropy": 5.610493230819702,
      "epoch": 1.6246526399364827,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.1234,
      "mean_token_accuracy": 0.17017478346824647,
      "num_tokens": 35840834.0,
      "step": 16370
    },
    {
      "entropy": 5.5416477680206295,
      "epoch": 1.6251488685986502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17444256097078323,
      "num_tokens": 35850889.0,
      "step": 16375
    },
    {
      "entropy": 5.4726176261901855,
      "epoch": 1.6256450972608176,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17872782200574874,
      "num_tokens": 35862491.0,
      "step": 16380
    },
    {
      "entropy": 5.581227159500122,
      "epoch": 1.6261413259229853,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16669266074895858,
      "num_tokens": 35873567.0,
      "step": 16385
    },
    {
      "entropy": 5.551181364059448,
      "epoch": 1.6266375545851528,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.0984,
      "mean_token_accuracy": 0.17482926696538925,
      "num_tokens": 35884970.0,
      "step": 16390
    },
    {
      "entropy": 5.6201540470123295,
      "epoch": 1.6271337832473205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.1572,
      "mean_token_accuracy": 0.17444658279418945,
      "num_tokens": 35895352.0,
      "step": 16395
    },
    {
      "entropy": 5.5894362449646,
      "epoch": 1.627630011909488,
      "grad_norm": 0.875,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.1939,
      "mean_token_accuracy": 0.16944705545902253,
      "num_tokens": 35908224.0,
      "step": 16400
    },
    {
      "entropy": 5.513410949707032,
      "epoch": 1.6281262405716554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047403677817630226,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17693785279989244,
      "num_tokens": 35918484.0,
      "step": 16405
    },
    {
      "entropy": 5.522650718688965,
      "epoch": 1.628622469233823,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.0646,
      "mean_token_accuracy": 0.17389994859695435,
      "num_tokens": 35929682.0,
      "step": 16410
    },
    {
      "entropy": 5.545888948440552,
      "epoch": 1.6291186978959904,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000474003722372078,
      "loss": 5.1338,
      "mean_token_accuracy": 0.1724226728081703,
      "num_tokens": 35939692.0,
      "step": 16415
    },
    {
      "entropy": 5.599477672576905,
      "epoch": 1.6296149265581579,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.1598,
      "mean_token_accuracy": 0.1716681867837906,
      "num_tokens": 35950871.0,
      "step": 16420
    },
    {
      "entropy": 5.592494010925293,
      "epoch": 1.6301111552203256,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.1926,
      "mean_token_accuracy": 0.16629267632961273,
      "num_tokens": 35962292.0,
      "step": 16425
    },
    {
      "entropy": 5.55762505531311,
      "epoch": 1.630607383882493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17774157971143723,
      "num_tokens": 35973838.0,
      "step": 16430
    },
    {
      "entropy": 5.545197772979736,
      "epoch": 1.6311036125446607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17368272244930266,
      "num_tokens": 35984309.0,
      "step": 16435
    },
    {
      "entropy": 5.5290357112884525,
      "epoch": 1.6315998412068282,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.1277,
      "mean_token_accuracy": 0.17245234549045563,
      "num_tokens": 35994200.0,
      "step": 16440
    },
    {
      "entropy": 5.636565637588501,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000473904436535807,
      "loss": 5.237,
      "mean_token_accuracy": 0.16564082950353623,
      "num_tokens": 36004839.0,
      "step": 16445
    },
    {
      "entropy": 5.4973756790161135,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.1116,
      "mean_token_accuracy": 0.16888900101184845,
      "num_tokens": 36015422.0,
      "step": 16450
    },
    {
      "entropy": 5.633340167999267,
      "epoch": 1.6330885271933306,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.2152,
      "mean_token_accuracy": 0.16588245034217836,
      "num_tokens": 36027453.0,
      "step": 16455
    },
    {
      "entropy": 5.611720180511474,
      "epoch": 1.6335847558554981,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.2309,
      "mean_token_accuracy": 0.17329298406839372,
      "num_tokens": 36039671.0,
      "step": 16460
    },
    {
      "entropy": 5.565054082870484,
      "epoch": 1.6340809845176656,
      "grad_norm": 0.921875,
      "learning_rate": 0.000473838147324794,
      "loss": 5.1728,
      "mean_token_accuracy": 0.16812328547239302,
      "num_tokens": 36053006.0,
      "step": 16465
    },
    {
      "entropy": 5.6080029010772705,
      "epoch": 1.6345772131798333,
      "grad_norm": 1.015625,
      "learning_rate": 0.000473821562694119,
      "loss": 5.1473,
      "mean_token_accuracy": 0.17244131863117218,
      "num_tokens": 36063764.0,
      "step": 16470
    },
    {
      "entropy": 5.526020193099976,
      "epoch": 1.6350734418420008,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.0949,
      "mean_token_accuracy": 0.17380722165107726,
      "num_tokens": 36073449.0,
      "step": 16475
    },
    {
      "entropy": 5.552924299240113,
      "epoch": 1.6355696705041685,
      "grad_norm": 1.0,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.1656,
      "mean_token_accuracy": 0.17469855695962905,
      "num_tokens": 36084912.0,
      "step": 16480
    },
    {
      "entropy": 5.473834705352783,
      "epoch": 1.636065899166336,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17293040007352828,
      "num_tokens": 36095031.0,
      "step": 16485
    },
    {
      "entropy": 5.541517496109009,
      "epoch": 1.6365621278285034,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17501949816942214,
      "num_tokens": 36106689.0,
      "step": 16490
    },
    {
      "entropy": 5.540836763381958,
      "epoch": 1.6370583564906709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17585390508174897,
      "num_tokens": 36118043.0,
      "step": 16495
    },
    {
      "entropy": 5.557962131500244,
      "epoch": 1.6375545851528384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.142,
      "mean_token_accuracy": 0.1731821745634079,
      "num_tokens": 36128155.0,
      "step": 16500
    },
    {
      "entropy": 5.584091520309448,
      "epoch": 1.6380508138150058,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.136,
      "mean_token_accuracy": 0.17413569390773773,
      "num_tokens": 36139377.0,
      "step": 16505
    },
    {
      "entropy": 5.491154193878174,
      "epoch": 1.6385470424771735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17960765063762665,
      "num_tokens": 36149575.0,
      "step": 16510
    },
    {
      "entropy": 5.5118492603302,
      "epoch": 1.639043271139341,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.1155,
      "mean_token_accuracy": 0.16733430474996566,
      "num_tokens": 36161707.0,
      "step": 16515
    },
    {
      "entropy": 5.526417207717896,
      "epoch": 1.6395394998015085,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.1245,
      "mean_token_accuracy": 0.1712793678045273,
      "num_tokens": 36173095.0,
      "step": 16520
    },
    {
      "entropy": 5.616968250274658,
      "epoch": 1.6400357284636762,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.1195,
      "mean_token_accuracy": 0.17503081560134887,
      "num_tokens": 36184895.0,
      "step": 16525
    },
    {
      "entropy": 5.503020095825195,
      "epoch": 1.6405319571258437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.0967,
      "mean_token_accuracy": 0.17783630937337874,
      "num_tokens": 36195277.0,
      "step": 16530
    },
    {
      "entropy": 5.522178792953492,
      "epoch": 1.6410281857880111,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.1411,
      "mean_token_accuracy": 0.17162981033325195,
      "num_tokens": 36207555.0,
      "step": 16535
    },
    {
      "entropy": 5.511473703384399,
      "epoch": 1.6415244144501786,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1846883624792099,
      "num_tokens": 36217632.0,
      "step": 16540
    },
    {
      "entropy": 5.592891550064087,
      "epoch": 1.642020643112346,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.1273,
      "mean_token_accuracy": 0.17273232489824294,
      "num_tokens": 36227368.0,
      "step": 16545
    },
    {
      "entropy": 5.572576999664307,
      "epoch": 1.6425168717745136,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.192,
      "mean_token_accuracy": 0.16771386712789535,
      "num_tokens": 36237932.0,
      "step": 16550
    },
    {
      "entropy": 5.5240638256073,
      "epoch": 1.6430131004366813,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.0841,
      "mean_token_accuracy": 0.17427465468645095,
      "num_tokens": 36249365.0,
      "step": 16555
    },
    {
      "entropy": 5.631076431274414,
      "epoch": 1.6435093290988487,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.1527,
      "mean_token_accuracy": 0.16402993500232696,
      "num_tokens": 36259254.0,
      "step": 16560
    },
    {
      "entropy": 5.546822738647461,
      "epoch": 1.6440055577610164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.0785,
      "mean_token_accuracy": 0.17386584877967834,
      "num_tokens": 36269249.0,
      "step": 16565
    },
    {
      "entropy": 5.57869062423706,
      "epoch": 1.644501786423184,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.178,
      "mean_token_accuracy": 0.1728519767522812,
      "num_tokens": 36280275.0,
      "step": 16570
    },
    {
      "entropy": 5.643891334533691,
      "epoch": 1.6449980150853514,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.2114,
      "mean_token_accuracy": 0.17100379467010499,
      "num_tokens": 36292750.0,
      "step": 16575
    },
    {
      "entropy": 5.568701076507568,
      "epoch": 1.6454942437475188,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.1487,
      "mean_token_accuracy": 0.1725194126367569,
      "num_tokens": 36303148.0,
      "step": 16580
    },
    {
      "entropy": 5.5605346202850345,
      "epoch": 1.6459904724096863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.1384,
      "mean_token_accuracy": 0.1729944348335266,
      "num_tokens": 36314939.0,
      "step": 16585
    },
    {
      "entropy": 5.565080213546753,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.2111,
      "mean_token_accuracy": 0.17128254175186158,
      "num_tokens": 36327036.0,
      "step": 16590
    },
    {
      "entropy": 5.498984098434448,
      "epoch": 1.6469829297340213,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1765209600329399,
      "num_tokens": 36336844.0,
      "step": 16595
    },
    {
      "entropy": 5.556779861450195,
      "epoch": 1.647479158396189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17248484045267104,
      "num_tokens": 36347538.0,
      "step": 16600
    },
    {
      "entropy": 5.504328012466431,
      "epoch": 1.6479753870583564,
      "grad_norm": 1.0,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.0583,
      "mean_token_accuracy": 0.17865814119577408,
      "num_tokens": 36358221.0,
      "step": 16605
    },
    {
      "entropy": 5.535958957672119,
      "epoch": 1.6484716157205241,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.0994,
      "mean_token_accuracy": 0.1757394164800644,
      "num_tokens": 36368756.0,
      "step": 16610
    },
    {
      "entropy": 5.515193796157837,
      "epoch": 1.6489678443826916,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.135,
      "mean_token_accuracy": 0.17524597495794297,
      "num_tokens": 36379201.0,
      "step": 16615
    },
    {
      "entropy": 5.569980621337891,
      "epoch": 1.649464073044859,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.1046,
      "mean_token_accuracy": 0.17163171917200087,
      "num_tokens": 36389375.0,
      "step": 16620
    },
    {
      "entropy": 5.585546779632568,
      "epoch": 1.6499603017070266,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.1723,
      "mean_token_accuracy": 0.16765961647033692,
      "num_tokens": 36400064.0,
      "step": 16625
    },
    {
      "entropy": 5.571654033660889,
      "epoch": 1.650456530369194,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.1217,
      "mean_token_accuracy": 0.17913615703582764,
      "num_tokens": 36410165.0,
      "step": 16630
    },
    {
      "entropy": 5.550129318237305,
      "epoch": 1.6509527590313615,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.1352,
      "mean_token_accuracy": 0.16735111624002458,
      "num_tokens": 36421600.0,
      "step": 16635
    },
    {
      "entropy": 5.592658233642578,
      "epoch": 1.6514489876935292,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17113130986690522,
      "num_tokens": 36433439.0,
      "step": 16640
    },
    {
      "entropy": 5.530459451675415,
      "epoch": 1.6519452163556967,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1730123743414879,
      "num_tokens": 36444001.0,
      "step": 16645
    },
    {
      "entropy": 5.575653696060181,
      "epoch": 1.6524414450178644,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16327238082885742,
      "num_tokens": 36454911.0,
      "step": 16650
    },
    {
      "entropy": 5.553426122665405,
      "epoch": 1.6529376736800319,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.1289,
      "mean_token_accuracy": 0.17162616848945617,
      "num_tokens": 36465537.0,
      "step": 16655
    },
    {
      "entropy": 5.561120796203613,
      "epoch": 1.6534339023421993,
      "grad_norm": 1.0,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.1536,
      "mean_token_accuracy": 0.17259007841348648,
      "num_tokens": 36476487.0,
      "step": 16660
    },
    {
      "entropy": 5.560362339019775,
      "epoch": 1.6539301310043668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.1653,
      "mean_token_accuracy": 0.17410876899957656,
      "num_tokens": 36486447.0,
      "step": 16665
    },
    {
      "entropy": 5.560991621017456,
      "epoch": 1.6544263596665343,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.1745,
      "mean_token_accuracy": 0.17099374234676362,
      "num_tokens": 36497704.0,
      "step": 16670
    },
    {
      "entropy": 5.658302259445191,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.1744,
      "mean_token_accuracy": 0.16709312945604324,
      "num_tokens": 36508369.0,
      "step": 16675
    },
    {
      "entropy": 5.628046321868896,
      "epoch": 1.6554188169908692,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.1967,
      "mean_token_accuracy": 0.1657721295952797,
      "num_tokens": 36519766.0,
      "step": 16680
    },
    {
      "entropy": 5.528705215454101,
      "epoch": 1.655915045653037,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047310376493243754,
      "loss": 5.1191,
      "mean_token_accuracy": 0.1771553337574005,
      "num_tokens": 36531022.0,
      "step": 16685
    },
    {
      "entropy": 5.570150852203369,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17424995452165604,
      "num_tokens": 36542325.0,
      "step": 16690
    },
    {
      "entropy": 5.620160007476807,
      "epoch": 1.656907502977372,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.1614,
      "mean_token_accuracy": 0.17124625891447068,
      "num_tokens": 36553671.0,
      "step": 16695
    },
    {
      "entropy": 5.600968933105468,
      "epoch": 1.6574037316395396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.1022,
      "mean_token_accuracy": 0.1686378076672554,
      "num_tokens": 36564835.0,
      "step": 16700
    },
    {
      "entropy": 5.547097015380859,
      "epoch": 1.657899960301707,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17679499834775925,
      "num_tokens": 36575793.0,
      "step": 16705
    },
    {
      "entropy": 5.5952507019042965,
      "epoch": 1.6583961889638745,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.1777,
      "mean_token_accuracy": 0.17638331651687622,
      "num_tokens": 36587097.0,
      "step": 16710
    },
    {
      "entropy": 5.5118207931518555,
      "epoch": 1.658892417626042,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.0883,
      "mean_token_accuracy": 0.16899824291467666,
      "num_tokens": 36598609.0,
      "step": 16715
    },
    {
      "entropy": 5.575468826293945,
      "epoch": 1.6593886462882095,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.0915,
      "mean_token_accuracy": 0.17723340541124344,
      "num_tokens": 36609476.0,
      "step": 16720
    },
    {
      "entropy": 5.596626806259155,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.1386,
      "mean_token_accuracy": 0.17457910180091857,
      "num_tokens": 36620160.0,
      "step": 16725
    },
    {
      "entropy": 5.578025913238525,
      "epoch": 1.6603811036125447,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17346347272396087,
      "num_tokens": 36630742.0,
      "step": 16730
    },
    {
      "entropy": 5.547439098358154,
      "epoch": 1.6608773322747123,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.0613,
      "mean_token_accuracy": 0.18049250692129135,
      "num_tokens": 36640980.0,
      "step": 16735
    },
    {
      "entropy": 5.5570323944091795,
      "epoch": 1.6613735609368798,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.1268,
      "mean_token_accuracy": 0.17028154879808427,
      "num_tokens": 36651703.0,
      "step": 16740
    },
    {
      "entropy": 5.520350694656372,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.076,
      "mean_token_accuracy": 0.18021786957979202,
      "num_tokens": 36662362.0,
      "step": 16745
    },
    {
      "entropy": 5.581615781784057,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17710793763399124,
      "num_tokens": 36671775.0,
      "step": 16750
    },
    {
      "entropy": 5.564860105514526,
      "epoch": 1.6628622469233822,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.1027,
      "mean_token_accuracy": 0.17490515261888503,
      "num_tokens": 36683183.0,
      "step": 16755
    },
    {
      "entropy": 5.591960382461548,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17104759067296982,
      "num_tokens": 36694106.0,
      "step": 16760
    },
    {
      "entropy": 5.536469554901123,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17609127461910248,
      "num_tokens": 36704607.0,
      "step": 16765
    },
    {
      "entropy": 5.5718268871307375,
      "epoch": 1.664350932909885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.0907,
      "mean_token_accuracy": 0.1717880129814148,
      "num_tokens": 36715330.0,
      "step": 16770
    },
    {
      "entropy": 5.585845184326172,
      "epoch": 1.6648471615720524,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.1501,
      "mean_token_accuracy": 0.16589295417070388,
      "num_tokens": 36727238.0,
      "step": 16775
    },
    {
      "entropy": 5.558612251281739,
      "epoch": 1.66534339023422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.0873,
      "mean_token_accuracy": 0.17090577632188797,
      "num_tokens": 36737355.0,
      "step": 16780
    },
    {
      "entropy": 5.495054435729981,
      "epoch": 1.6658396188963875,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1795773684978485,
      "num_tokens": 36748331.0,
      "step": 16785
    },
    {
      "entropy": 5.549136781692505,
      "epoch": 1.666335847558555,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17451054900884627,
      "num_tokens": 36758999.0,
      "step": 16790
    },
    {
      "entropy": 5.546716403961182,
      "epoch": 1.6668320762207225,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.0414,
      "mean_token_accuracy": 0.17957437932491302,
      "num_tokens": 36771646.0,
      "step": 16795
    },
    {
      "entropy": 5.501601982116699,
      "epoch": 1.66732830488289,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17465578615665436,
      "num_tokens": 36781670.0,
      "step": 16800
    },
    {
      "entropy": 5.543744945526123,
      "epoch": 1.6678245335450574,
      "grad_norm": 1.0,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.1467,
      "mean_token_accuracy": 0.16759931445121765,
      "num_tokens": 36792664.0,
      "step": 16805
    },
    {
      "entropy": 5.494076299667358,
      "epoch": 1.6683207622072251,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.0297,
      "mean_token_accuracy": 0.17740671336650848,
      "num_tokens": 36803551.0,
      "step": 16810
    },
    {
      "entropy": 5.510613918304443,
      "epoch": 1.6688169908693926,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17325358390808104,
      "num_tokens": 36818202.0,
      "step": 16815
    },
    {
      "entropy": 5.601619386672974,
      "epoch": 1.66931321953156,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.1488,
      "mean_token_accuracy": 0.17675337493419646,
      "num_tokens": 36829861.0,
      "step": 16820
    },
    {
      "entropy": 5.594704103469849,
      "epoch": 1.6698094481937278,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004726314766116223,
      "loss": 5.1337,
      "mean_token_accuracy": 0.17359722256660462,
      "num_tokens": 36841115.0,
      "step": 16825
    },
    {
      "entropy": 5.546628141403199,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472614538054793,
      "loss": 5.0985,
      "mean_token_accuracy": 0.17283360809087753,
      "num_tokens": 36851176.0,
      "step": 16830
    },
    {
      "entropy": 5.53088526725769,
      "epoch": 1.6708019055180627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.0895,
      "mean_token_accuracy": 0.17304913997650145,
      "num_tokens": 36861408.0,
      "step": 16835
    },
    {
      "entropy": 5.512455749511719,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18245433568954467,
      "num_tokens": 36871244.0,
      "step": 16840
    },
    {
      "entropy": 5.585089540481567,
      "epoch": 1.6717943628423977,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.1406,
      "mean_token_accuracy": 0.16606016755104064,
      "num_tokens": 36882955.0,
      "step": 16845
    },
    {
      "entropy": 5.574730253219604,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.0536,
      "mean_token_accuracy": 0.17618249654769896,
      "num_tokens": 36893118.0,
      "step": 16850
    },
    {
      "entropy": 5.629086303710937,
      "epoch": 1.6727868201667329,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.2536,
      "mean_token_accuracy": 0.16652028262615204,
      "num_tokens": 36903484.0,
      "step": 16855
    },
    {
      "entropy": 5.516356134414673,
      "epoch": 1.6732830488289003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.0746,
      "mean_token_accuracy": 0.17645945996046067,
      "num_tokens": 36913439.0,
      "step": 16860
    },
    {
      "entropy": 5.5640356063842775,
      "epoch": 1.673779277491068,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.1662,
      "mean_token_accuracy": 0.17025782018899918,
      "num_tokens": 36925977.0,
      "step": 16865
    },
    {
      "entropy": 5.588336181640625,
      "epoch": 1.6742755061532355,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17377639412879944,
      "num_tokens": 36937318.0,
      "step": 16870
    },
    {
      "entropy": 5.619341516494751,
      "epoch": 1.674771734815403,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.1326,
      "mean_token_accuracy": 0.16857100576162337,
      "num_tokens": 36948199.0,
      "step": 16875
    },
    {
      "entropy": 5.550583028793335,
      "epoch": 1.6752679634775705,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17058052867650986,
      "num_tokens": 36959813.0,
      "step": 16880
    },
    {
      "entropy": 5.553938961029052,
      "epoch": 1.675764192139738,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.0758,
      "mean_token_accuracy": 0.17186720073223113,
      "num_tokens": 36970548.0,
      "step": 16885
    },
    {
      "entropy": 5.525807046890259,
      "epoch": 1.6762604208019054,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.1886,
      "mean_token_accuracy": 0.16610622107982637,
      "num_tokens": 36983072.0,
      "step": 16890
    },
    {
      "entropy": 5.568398666381836,
      "epoch": 1.6767566494640729,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17136835157871247,
      "num_tokens": 36993684.0,
      "step": 16895
    },
    {
      "entropy": 5.570712518692017,
      "epoch": 1.6772528781262406,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.0396,
      "mean_token_accuracy": 0.17761206030845642,
      "num_tokens": 37004776.0,
      "step": 16900
    },
    {
      "entropy": 5.566556406021118,
      "epoch": 1.677749106788408,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17205696702003478,
      "num_tokens": 37014669.0,
      "step": 16905
    },
    {
      "entropy": 5.557174634933472,
      "epoch": 1.6782453354505757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1795705333352089,
      "num_tokens": 37025645.0,
      "step": 16910
    },
    {
      "entropy": 5.549451637268066,
      "epoch": 1.6787415641127432,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.1575,
      "mean_token_accuracy": 0.16867785602808,
      "num_tokens": 37037065.0,
      "step": 16915
    },
    {
      "entropy": 5.549391555786133,
      "epoch": 1.6792377927749107,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.1247,
      "mean_token_accuracy": 0.16746118664741516,
      "num_tokens": 37047957.0,
      "step": 16920
    },
    {
      "entropy": 5.620020914077759,
      "epoch": 1.6797340214370782,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17400882691144942,
      "num_tokens": 37059798.0,
      "step": 16925
    },
    {
      "entropy": 5.609007835388184,
      "epoch": 1.6802302500992456,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.1846,
      "mean_token_accuracy": 0.16993080526590348,
      "num_tokens": 37071229.0,
      "step": 16930
    },
    {
      "entropy": 5.5604085445404055,
      "epoch": 1.6807264787614131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.1549,
      "mean_token_accuracy": 0.16805250942707062,
      "num_tokens": 37082191.0,
      "step": 16935
    },
    {
      "entropy": 5.495192527770996,
      "epoch": 1.6812227074235808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17819166779518128,
      "num_tokens": 37092718.0,
      "step": 16940
    },
    {
      "entropy": 5.508884620666504,
      "epoch": 1.6817189360857483,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17841824442148208,
      "num_tokens": 37104287.0,
      "step": 16945
    },
    {
      "entropy": 5.552346563339233,
      "epoch": 1.682215164747916,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.1882,
      "mean_token_accuracy": 0.17255532890558242,
      "num_tokens": 37114994.0,
      "step": 16950
    },
    {
      "entropy": 5.583250617980957,
      "epoch": 1.6827113934100835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.0295,
      "mean_token_accuracy": 0.17901719510555267,
      "num_tokens": 37125258.0,
      "step": 16955
    },
    {
      "entropy": 5.591987800598145,
      "epoch": 1.683207622072251,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.1589,
      "mean_token_accuracy": 0.17840666025877,
      "num_tokens": 37136055.0,
      "step": 16960
    },
    {
      "entropy": 5.561094951629639,
      "epoch": 1.6837038507344184,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.09,
      "mean_token_accuracy": 0.17484714537858964,
      "num_tokens": 37147312.0,
      "step": 16965
    },
    {
      "entropy": 5.641918897628784,
      "epoch": 1.684200079396586,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000472138270433348,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1730018064379692,
      "num_tokens": 37157940.0,
      "step": 16970
    },
    {
      "entropy": 5.586187410354614,
      "epoch": 1.6846963080587534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16608615070581437,
      "num_tokens": 37169863.0,
      "step": 16975
    },
    {
      "entropy": 5.580465269088745,
      "epoch": 1.6851925367209208,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.1623,
      "mean_token_accuracy": 0.16686749458312988,
      "num_tokens": 37181840.0,
      "step": 16980
    },
    {
      "entropy": 5.550250005722046,
      "epoch": 1.6856887653830885,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.0772,
      "mean_token_accuracy": 0.17856526672840117,
      "num_tokens": 37191015.0,
      "step": 16985
    },
    {
      "entropy": 5.530923748016358,
      "epoch": 1.686184994045256,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047206991911465,
      "loss": 5.0055,
      "mean_token_accuracy": 0.1746320679783821,
      "num_tokens": 37201829.0,
      "step": 16990
    },
    {
      "entropy": 5.490995597839356,
      "epoch": 1.6866812227074237,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.0835,
      "mean_token_accuracy": 0.16554984599351882,
      "num_tokens": 37213870.0,
      "step": 16995
    },
    {
      "entropy": 5.563897228240966,
      "epoch": 1.6871774513695912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17030187994241713,
      "num_tokens": 37225232.0,
      "step": 17000
    },
    {
      "entropy": 5.6181731700897215,
      "epoch": 1.6876736800317587,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.2292,
      "mean_token_accuracy": 0.16745668649673462,
      "num_tokens": 37236329.0,
      "step": 17005
    },
    {
      "entropy": 5.6509161472320555,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.209,
      "mean_token_accuracy": 0.1676446184515953,
      "num_tokens": 37247402.0,
      "step": 17010
    },
    {
      "entropy": 5.598942375183105,
      "epoch": 1.6886661373560936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.1358,
      "mean_token_accuracy": 0.16231524497270583,
      "num_tokens": 37258629.0,
      "step": 17015
    },
    {
      "entropy": 5.549152660369873,
      "epoch": 1.689162366018261,
      "grad_norm": 0.875,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.1116,
      "mean_token_accuracy": 0.17736204862594604,
      "num_tokens": 37270154.0,
      "step": 17020
    },
    {
      "entropy": 5.58444447517395,
      "epoch": 1.6896585946804288,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000471950116177256,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1783924803137779,
      "num_tokens": 37280801.0,
      "step": 17025
    },
    {
      "entropy": 5.610301542282104,
      "epoch": 1.6901548233425963,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.1626,
      "mean_token_accuracy": 0.17282964289188385,
      "num_tokens": 37293057.0,
      "step": 17030
    },
    {
      "entropy": 5.577498149871826,
      "epoch": 1.690651052004764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.1625,
      "mean_token_accuracy": 0.17618027329444885,
      "num_tokens": 37305784.0,
      "step": 17035
    },
    {
      "entropy": 5.5358928680419925,
      "epoch": 1.6911472806669314,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.1304,
      "mean_token_accuracy": 0.16793324053287506,
      "num_tokens": 37317067.0,
      "step": 17040
    },
    {
      "entropy": 5.628410863876343,
      "epoch": 1.691643509329099,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.1497,
      "mean_token_accuracy": 0.17207755893468857,
      "num_tokens": 37327568.0,
      "step": 17045
    },
    {
      "entropy": 5.6081705570220945,
      "epoch": 1.6921397379912664,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.1758,
      "mean_token_accuracy": 0.17571752220392228,
      "num_tokens": 37339235.0,
      "step": 17050
    },
    {
      "entropy": 5.540337562561035,
      "epoch": 1.6926359666534339,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.0358,
      "mean_token_accuracy": 0.17690628618001938,
      "num_tokens": 37348943.0,
      "step": 17055
    },
    {
      "entropy": 5.529344654083252,
      "epoch": 1.6931321953156013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.1193,
      "mean_token_accuracy": 0.17414300590753556,
      "num_tokens": 37359792.0,
      "step": 17060
    },
    {
      "entropy": 5.562683963775635,
      "epoch": 1.6936284239777688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.1564,
      "mean_token_accuracy": 0.17573115229606628,
      "num_tokens": 37371905.0,
      "step": 17065
    },
    {
      "entropy": 5.5884064674377445,
      "epoch": 1.6941246526399365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1685376524925232,
      "num_tokens": 37383262.0,
      "step": 17070
    },
    {
      "entropy": 5.589658594131469,
      "epoch": 1.694620881302104,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.1369,
      "mean_token_accuracy": 0.17654308527708054,
      "num_tokens": 37394666.0,
      "step": 17075
    },
    {
      "entropy": 5.642005395889282,
      "epoch": 1.6951171099642717,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1669481173157692,
      "num_tokens": 37405860.0,
      "step": 17080
    },
    {
      "entropy": 5.586106014251709,
      "epoch": 1.6956133386264391,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.0659,
      "mean_token_accuracy": 0.1757259964942932,
      "num_tokens": 37416596.0,
      "step": 17085
    },
    {
      "entropy": 5.5851655960083,
      "epoch": 1.6961095672886066,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.1715,
      "mean_token_accuracy": 0.16451943218708037,
      "num_tokens": 37427293.0,
      "step": 17090
    },
    {
      "entropy": 5.479903697967529,
      "epoch": 1.696605795950774,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.062,
      "mean_token_accuracy": 0.17856266647577285,
      "num_tokens": 37438753.0,
      "step": 17095
    },
    {
      "entropy": 5.549706554412841,
      "epoch": 1.6971020246129416,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.104,
      "mean_token_accuracy": 0.17403829842805862,
      "num_tokens": 37449917.0,
      "step": 17100
    },
    {
      "entropy": 5.554020833969116,
      "epoch": 1.697598253275109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004716753824819654,
      "loss": 5.004,
      "mean_token_accuracy": 0.17841964960098267,
      "num_tokens": 37460248.0,
      "step": 17105
    },
    {
      "entropy": 5.51776442527771,
      "epoch": 1.6980944819372767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.0805,
      "mean_token_accuracy": 0.17064074277877808,
      "num_tokens": 37470404.0,
      "step": 17110
    },
    {
      "entropy": 5.551681327819824,
      "epoch": 1.6985907105994442,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.1686,
      "mean_token_accuracy": 0.16874366253614426,
      "num_tokens": 37481764.0,
      "step": 17115
    },
    {
      "entropy": 5.60221905708313,
      "epoch": 1.699086939261612,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.0856,
      "mean_token_accuracy": 0.17710286974906922,
      "num_tokens": 37492927.0,
      "step": 17120
    },
    {
      "entropy": 5.676514720916748,
      "epoch": 1.6995831679237794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1737758472561836,
      "num_tokens": 37503101.0,
      "step": 17125
    },
    {
      "entropy": 5.61646466255188,
      "epoch": 1.7000793965859469,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.1868,
      "mean_token_accuracy": 0.17101385891437532,
      "num_tokens": 37514265.0,
      "step": 17130
    },
    {
      "entropy": 5.557420492172241,
      "epoch": 1.7005756252481143,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.1391,
      "mean_token_accuracy": 0.16693486273288727,
      "num_tokens": 37524564.0,
      "step": 17135
    },
    {
      "entropy": 5.556945323944092,
      "epoch": 1.7010718539102818,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.1977,
      "mean_token_accuracy": 0.1692441076040268,
      "num_tokens": 37536876.0,
      "step": 17140
    },
    {
      "entropy": 5.572968482971191,
      "epoch": 1.7015680825724493,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17236080765724182,
      "num_tokens": 37547868.0,
      "step": 17145
    },
    {
      "entropy": 5.541991519927978,
      "epoch": 1.7020643112346168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.116,
      "mean_token_accuracy": 0.1758270338177681,
      "num_tokens": 37558487.0,
      "step": 17150
    },
    {
      "entropy": 5.58472843170166,
      "epoch": 1.7025605398967845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004715030397853651,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17693308889865875,
      "num_tokens": 37568226.0,
      "step": 17155
    },
    {
      "entropy": 5.626507377624511,
      "epoch": 1.703056768558952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.2172,
      "mean_token_accuracy": 0.16391189098358155,
      "num_tokens": 37579571.0,
      "step": 17160
    },
    {
      "entropy": 5.617984390258789,
      "epoch": 1.7035529972211196,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.1532,
      "mean_token_accuracy": 0.17032649964094163,
      "num_tokens": 37590532.0,
      "step": 17165
    },
    {
      "entropy": 5.597557592391968,
      "epoch": 1.704049225883287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.1782,
      "mean_token_accuracy": 0.1689898505806923,
      "num_tokens": 37601946.0,
      "step": 17170
    },
    {
      "entropy": 5.458157205581665,
      "epoch": 1.7045454545454546,
      "grad_norm": 0.84375,
      "learning_rate": 0.00047143396624292454,
      "loss": 5.0086,
      "mean_token_accuracy": 0.17932301610708237,
      "num_tokens": 37613944.0,
      "step": 17175
    },
    {
      "entropy": 5.509756946563721,
      "epoch": 1.705041683207622,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.0161,
      "mean_token_accuracy": 0.17974766790866853,
      "num_tokens": 37625308.0,
      "step": 17180
    },
    {
      "entropy": 5.547039651870728,
      "epoch": 1.7055379118697895,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.0882,
      "mean_token_accuracy": 0.16557158380746842,
      "num_tokens": 37635570.0,
      "step": 17185
    },
    {
      "entropy": 5.606826400756836,
      "epoch": 1.706034140531957,
      "grad_norm": 1.0,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.1483,
      "mean_token_accuracy": 0.16864078491926193,
      "num_tokens": 37646279.0,
      "step": 17190
    },
    {
      "entropy": 5.586287546157837,
      "epoch": 1.7065303691941247,
      "grad_norm": 0.84375,
      "learning_rate": 0.000471364814762261,
      "loss": 5.0516,
      "mean_token_accuracy": 0.17428333163261414,
      "num_tokens": 37657182.0,
      "step": 17195
    },
    {
      "entropy": 5.5239636421203615,
      "epoch": 1.7070265978562922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17677990198135377,
      "num_tokens": 37668996.0,
      "step": 17200
    },
    {
      "entropy": 5.5084692478179935,
      "epoch": 1.7075228265184597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.1131,
      "mean_token_accuracy": 0.173566772043705,
      "num_tokens": 37679501.0,
      "step": 17205
    },
    {
      "entropy": 5.549026584625244,
      "epoch": 1.7080190551806274,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16813481897115706,
      "num_tokens": 37691173.0,
      "step": 17210
    },
    {
      "entropy": 5.578010702133179,
      "epoch": 1.7085152838427948,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1733037441968918,
      "num_tokens": 37701875.0,
      "step": 17215
    },
    {
      "entropy": 5.5706971168518065,
      "epoch": 1.7090115125049623,
      "grad_norm": 1.03125,
      "learning_rate": 0.000471278265852607,
      "loss": 5.1165,
      "mean_token_accuracy": 0.17306775003671646,
      "num_tokens": 37711003.0,
      "step": 17220
    },
    {
      "entropy": 5.625764799118042,
      "epoch": 1.7095077411671298,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.1741,
      "mean_token_accuracy": 0.1696472942829132,
      "num_tokens": 37721335.0,
      "step": 17225
    },
    {
      "entropy": 5.561184310913086,
      "epoch": 1.7100039698292973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17938615679740905,
      "num_tokens": 37731664.0,
      "step": 17230
    },
    {
      "entropy": 5.6217193603515625,
      "epoch": 1.7105001984914647,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.2541,
      "mean_token_accuracy": 0.16740294843912124,
      "num_tokens": 37741619.0,
      "step": 17235
    },
    {
      "entropy": 5.610430765151977,
      "epoch": 1.7109964271536324,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.153,
      "mean_token_accuracy": 0.1751936122775078,
      "num_tokens": 37752387.0,
      "step": 17240
    },
    {
      "entropy": 5.517894983291626,
      "epoch": 1.7114926558158,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.0251,
      "mean_token_accuracy": 0.17199427187442778,
      "num_tokens": 37763459.0,
      "step": 17245
    },
    {
      "entropy": 5.5598796844482425,
      "epoch": 1.7119888844779676,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.0878,
      "mean_token_accuracy": 0.1737575799226761,
      "num_tokens": 37773848.0,
      "step": 17250
    },
    {
      "entropy": 5.557477760314941,
      "epoch": 1.712485113140135,
      "grad_norm": 1.0,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.1692,
      "mean_token_accuracy": 0.17377965450286864,
      "num_tokens": 37784074.0,
      "step": 17255
    },
    {
      "entropy": 5.5807554721832275,
      "epoch": 1.7129813418023025,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.1669,
      "mean_token_accuracy": 0.17161007076501847,
      "num_tokens": 37794926.0,
      "step": 17260
    },
    {
      "entropy": 5.5839362144470215,
      "epoch": 1.71347757046447,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.091,
      "mean_token_accuracy": 0.17425415068864822,
      "num_tokens": 37804926.0,
      "step": 17265
    },
    {
      "entropy": 5.6021209239959715,
      "epoch": 1.7139737991266375,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16806328892707825,
      "num_tokens": 37815241.0,
      "step": 17270
    },
    {
      "entropy": 5.568073749542236,
      "epoch": 1.714470027788805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.1,
      "mean_token_accuracy": 0.17393898367881774,
      "num_tokens": 37825047.0,
      "step": 17275
    },
    {
      "entropy": 5.6083849430084225,
      "epoch": 1.7149662564509724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.1485,
      "mean_token_accuracy": 0.17576317638158798,
      "num_tokens": 37836201.0,
      "step": 17280
    },
    {
      "entropy": 5.577491664886475,
      "epoch": 1.7154624851131401,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.1044,
      "mean_token_accuracy": 0.17419223338365555,
      "num_tokens": 37848242.0,
      "step": 17285
    },
    {
      "entropy": 5.488332223892212,
      "epoch": 1.7159587137753076,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.092,
      "mean_token_accuracy": 0.18224983811378478,
      "num_tokens": 37859956.0,
      "step": 17290
    },
    {
      "entropy": 5.547891902923584,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.2104,
      "mean_token_accuracy": 0.170424248278141,
      "num_tokens": 37870054.0,
      "step": 17295
    },
    {
      "entropy": 5.553718709945679,
      "epoch": 1.7169511710996428,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.0878,
      "mean_token_accuracy": 0.17395429611206054,
      "num_tokens": 37879871.0,
      "step": 17300
    },
    {
      "entropy": 5.5926477909088135,
      "epoch": 1.7174473997618103,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.1896,
      "mean_token_accuracy": 0.16903069615364075,
      "num_tokens": 37892642.0,
      "step": 17305
    },
    {
      "entropy": 5.565938949584961,
      "epoch": 1.7179436284239777,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17335714250802994,
      "num_tokens": 37904371.0,
      "step": 17310
    },
    {
      "entropy": 5.614768218994141,
      "epoch": 1.7184398570861452,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.1231,
      "mean_token_accuracy": 0.16958910822868348,
      "num_tokens": 37915538.0,
      "step": 17315
    },
    {
      "entropy": 5.525193691253662,
      "epoch": 1.7189360857483127,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17317724823951722,
      "num_tokens": 37926254.0,
      "step": 17320
    },
    {
      "entropy": 5.550056886672974,
      "epoch": 1.7194323144104804,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.1351,
      "mean_token_accuracy": 0.17549198418855666,
      "num_tokens": 37939242.0,
      "step": 17325
    },
    {
      "entropy": 5.552377510070801,
      "epoch": 1.7199285430726479,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.078,
      "mean_token_accuracy": 0.18029043525457383,
      "num_tokens": 37948970.0,
      "step": 17330
    },
    {
      "entropy": 5.631630611419678,
      "epoch": 1.7204247717348156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.295,
      "mean_token_accuracy": 0.16190911829471588,
      "num_tokens": 37960290.0,
      "step": 17335
    },
    {
      "entropy": 5.624152326583863,
      "epoch": 1.720921000396983,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17062582075595856,
      "num_tokens": 37971004.0,
      "step": 17340
    },
    {
      "entropy": 5.67258472442627,
      "epoch": 1.7214172290591505,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.167,
      "mean_token_accuracy": 0.1724618747830391,
      "num_tokens": 37982455.0,
      "step": 17345
    },
    {
      "entropy": 5.553209114074707,
      "epoch": 1.721913457721318,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.0717,
      "mean_token_accuracy": 0.1760333925485611,
      "num_tokens": 37993623.0,
      "step": 17350
    },
    {
      "entropy": 5.535041618347168,
      "epoch": 1.7224096863834855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.1053,
      "mean_token_accuracy": 0.17346712052822114,
      "num_tokens": 38002636.0,
      "step": 17355
    },
    {
      "entropy": 5.562751007080078,
      "epoch": 1.722905915045653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16984858810901643,
      "num_tokens": 38014469.0,
      "step": 17360
    },
    {
      "entropy": 5.561086130142212,
      "epoch": 1.7234021437078204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1762968733906746,
      "num_tokens": 38024558.0,
      "step": 17365
    },
    {
      "entropy": 5.528510808944702,
      "epoch": 1.723898372369988,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17843884229660034,
      "num_tokens": 38034872.0,
      "step": 17370
    },
    {
      "entropy": 5.577870082855225,
      "epoch": 1.7243946010321556,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.1154,
      "mean_token_accuracy": 0.16986772865056993,
      "num_tokens": 38044947.0,
      "step": 17375
    },
    {
      "entropy": 5.644710683822632,
      "epoch": 1.7248908296943233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.2684,
      "mean_token_accuracy": 0.16696690768003464,
      "num_tokens": 38055349.0,
      "step": 17380
    },
    {
      "entropy": 5.601987409591675,
      "epoch": 1.7253870583564908,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.1805,
      "mean_token_accuracy": 0.16832386702299118,
      "num_tokens": 38066337.0,
      "step": 17385
    },
    {
      "entropy": 5.55857400894165,
      "epoch": 1.7258832870186582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.1852,
      "mean_token_accuracy": 0.17110035866498946,
      "num_tokens": 38078402.0,
      "step": 17390
    },
    {
      "entropy": 5.556513738632202,
      "epoch": 1.7263795156808257,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.033,
      "mean_token_accuracy": 0.1807978019118309,
      "num_tokens": 38089781.0,
      "step": 17395
    },
    {
      "entropy": 5.574331569671631,
      "epoch": 1.7268757443429932,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17329989224672318,
      "num_tokens": 38100496.0,
      "step": 17400
    },
    {
      "entropy": 5.596127128601074,
      "epoch": 1.7273719730051607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.1148,
      "mean_token_accuracy": 0.17237476110458375,
      "num_tokens": 38110471.0,
      "step": 17405
    },
    {
      "entropy": 5.601821088790894,
      "epoch": 1.7278682016673284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.1149,
      "mean_token_accuracy": 0.17447086721658706,
      "num_tokens": 38120511.0,
      "step": 17410
    },
    {
      "entropy": 5.582054471969604,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.099,
      "mean_token_accuracy": 0.1697555288672447,
      "num_tokens": 38130501.0,
      "step": 17415
    },
    {
      "entropy": 5.4959958553314205,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17214010506868363,
      "num_tokens": 38139634.0,
      "step": 17420
    },
    {
      "entropy": 5.6000657081604,
      "epoch": 1.729356887653831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.1478,
      "mean_token_accuracy": 0.16409500390291215,
      "num_tokens": 38151216.0,
      "step": 17425
    },
    {
      "entropy": 5.5658196926116945,
      "epoch": 1.7298531163159985,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1730649247765541,
      "num_tokens": 38162014.0,
      "step": 17430
    },
    {
      "entropy": 5.560622978210449,
      "epoch": 1.730349344978166,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.1753,
      "mean_token_accuracy": 0.16950931400060654,
      "num_tokens": 38174928.0,
      "step": 17435
    },
    {
      "entropy": 5.598149299621582,
      "epoch": 1.7308455736403334,
      "grad_norm": 1.0,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.1619,
      "mean_token_accuracy": 0.16418444067239762,
      "num_tokens": 38185818.0,
      "step": 17440
    },
    {
      "entropy": 5.623664855957031,
      "epoch": 1.731341802302501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.1706,
      "mean_token_accuracy": 0.16976717859506607,
      "num_tokens": 38197595.0,
      "step": 17445
    },
    {
      "entropy": 5.608973217010498,
      "epoch": 1.7318380309646684,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.144,
      "mean_token_accuracy": 0.17007811963558198,
      "num_tokens": 38209874.0,
      "step": 17450
    },
    {
      "entropy": 5.597997665405273,
      "epoch": 1.732334259626836,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.1388,
      "mean_token_accuracy": 0.16732779890298843,
      "num_tokens": 38221054.0,
      "step": 17455
    },
    {
      "entropy": 5.652759552001953,
      "epoch": 1.7328304882890035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1692784994840622,
      "num_tokens": 38230766.0,
      "step": 17460
    },
    {
      "entropy": 5.533802032470703,
      "epoch": 1.7333267169511712,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1708204597234726,
      "num_tokens": 38242925.0,
      "step": 17465
    },
    {
      "entropy": 5.522342920303345,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17968921214342118,
      "num_tokens": 38254228.0,
      "step": 17470
    },
    {
      "entropy": 5.546210765838623,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.0928,
      "mean_token_accuracy": 0.17397087514400483,
      "num_tokens": 38265515.0,
      "step": 17475
    },
    {
      "entropy": 5.569547653198242,
      "epoch": 1.7348154029376737,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.1274,
      "mean_token_accuracy": 0.1714722380042076,
      "num_tokens": 38277232.0,
      "step": 17480
    },
    {
      "entropy": 5.524055624008179,
      "epoch": 1.7353116315998411,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.1217,
      "mean_token_accuracy": 0.16819774657487868,
      "num_tokens": 38288038.0,
      "step": 17485
    },
    {
      "entropy": 5.609125995635987,
      "epoch": 1.7358078602620086,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.1579,
      "mean_token_accuracy": 0.1689736694097519,
      "num_tokens": 38299053.0,
      "step": 17490
    },
    {
      "entropy": 5.668013334274292,
      "epoch": 1.7363040889241763,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1730890616774559,
      "num_tokens": 38311029.0,
      "step": 17495
    },
    {
      "entropy": 5.7066315650939945,
      "epoch": 1.7368003175863438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.2637,
      "mean_token_accuracy": 0.1738849014043808,
      "num_tokens": 38321525.0,
      "step": 17500
    },
    {
      "entropy": 5.5450835704803465,
      "epoch": 1.7372965462485113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.0698,
      "mean_token_accuracy": 0.1748120218515396,
      "num_tokens": 38331035.0,
      "step": 17505
    },
    {
      "entropy": 5.567495107650757,
      "epoch": 1.737792774910679,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.1701,
      "mean_token_accuracy": 0.17348637878894807,
      "num_tokens": 38342046.0,
      "step": 17510
    },
    {
      "entropy": 5.55100154876709,
      "epoch": 1.7382890035728464,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000470247813956351,
      "loss": 5.1041,
      "mean_token_accuracy": 0.17561482787132263,
      "num_tokens": 38352614.0,
      "step": 17515
    },
    {
      "entropy": 5.586223125457764,
      "epoch": 1.738785232235014,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.1864,
      "mean_token_accuracy": 0.16958013474941253,
      "num_tokens": 38364379.0,
      "step": 17520
    },
    {
      "entropy": 5.6004129409790036,
      "epoch": 1.7392814608971814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.201,
      "mean_token_accuracy": 0.16780415326356887,
      "num_tokens": 38373823.0,
      "step": 17525
    },
    {
      "entropy": 5.608495855331421,
      "epoch": 1.7397776895593489,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.1342,
      "mean_token_accuracy": 0.17163187116384507,
      "num_tokens": 38384238.0,
      "step": 17530
    },
    {
      "entropy": 5.556384181976318,
      "epoch": 1.7402739182215163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1790621593594551,
      "num_tokens": 38394980.0,
      "step": 17535
    },
    {
      "entropy": 5.499869537353516,
      "epoch": 1.740770146883684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.1748,
      "mean_token_accuracy": 0.16942480057477952,
      "num_tokens": 38406722.0,
      "step": 17540
    },
    {
      "entropy": 5.557612895965576,
      "epoch": 1.7412663755458515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000470142076603618,
      "loss": 5.0724,
      "mean_token_accuracy": 0.17250989228487015,
      "num_tokens": 38416204.0,
      "step": 17545
    },
    {
      "entropy": 5.551488924026489,
      "epoch": 1.7417626042080192,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.1237,
      "mean_token_accuracy": 0.1758826658129692,
      "num_tokens": 38426166.0,
      "step": 17550
    },
    {
      "entropy": 5.546440744400025,
      "epoch": 1.7422588328701867,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.1276,
      "mean_token_accuracy": 0.17287513017654418,
      "num_tokens": 38439195.0,
      "step": 17555
    },
    {
      "entropy": 5.556669807434082,
      "epoch": 1.7427550615323542,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17618490755558014,
      "num_tokens": 38449276.0,
      "step": 17560
    },
    {
      "entropy": 5.63215217590332,
      "epoch": 1.7432512901945216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.1791,
      "mean_token_accuracy": 0.16668294966220856,
      "num_tokens": 38460961.0,
      "step": 17565
    },
    {
      "entropy": 5.599747037887573,
      "epoch": 1.743747518856689,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17766036540269853,
      "num_tokens": 38470961.0,
      "step": 17570
    },
    {
      "entropy": 5.566015577316284,
      "epoch": 1.7442437475188566,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17804176807403566,
      "num_tokens": 38481233.0,
      "step": 17575
    },
    {
      "entropy": 5.544550371170044,
      "epoch": 1.744739976181024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.0376,
      "mean_token_accuracy": 0.17650674134492875,
      "num_tokens": 38492260.0,
      "step": 17580
    },
    {
      "entropy": 5.597955751419067,
      "epoch": 1.7452362048431918,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.1268,
      "mean_token_accuracy": 0.17658177614212037,
      "num_tokens": 38503492.0,
      "step": 17585
    },
    {
      "entropy": 5.536380863189697,
      "epoch": 1.7457324335053592,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17240659147500992,
      "num_tokens": 38514680.0,
      "step": 17590
    },
    {
      "entropy": 5.515688943862915,
      "epoch": 1.746228662167527,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004699654606008684,
      "loss": 5.0233,
      "mean_token_accuracy": 0.17473705857992172,
      "num_tokens": 38525948.0,
      "step": 17595
    },
    {
      "entropy": 5.6328285217285154,
      "epoch": 1.7467248908296944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.1516,
      "mean_token_accuracy": 0.16996827572584153,
      "num_tokens": 38535955.0,
      "step": 17600
    },
    {
      "entropy": 5.581509733200074,
      "epoch": 1.7472211194918619,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17425957471132278,
      "num_tokens": 38548420.0,
      "step": 17605
    },
    {
      "entropy": 5.688747406005859,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004699123815029993,
      "loss": 5.2814,
      "mean_token_accuracy": 0.16564230918884276,
      "num_tokens": 38559220.0,
      "step": 17610
    },
    {
      "entropy": 5.616655540466309,
      "epoch": 1.7482135768161968,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.1924,
      "mean_token_accuracy": 0.1625903829932213,
      "num_tokens": 38570045.0,
      "step": 17615
    },
    {
      "entropy": 5.629323577880859,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.1205,
      "mean_token_accuracy": 0.16595842093229293,
      "num_tokens": 38582191.0,
      "step": 17620
    },
    {
      "entropy": 5.6338108539581295,
      "epoch": 1.749206034140532,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17823554575443268,
      "num_tokens": 38592545.0,
      "step": 17625
    },
    {
      "entropy": 5.527970027923584,
      "epoch": 1.7497022628026995,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.1067,
      "mean_token_accuracy": 0.17270385473966599,
      "num_tokens": 38604024.0,
      "step": 17630
    },
    {
      "entropy": 5.558313512802124,
      "epoch": 1.7501984914648672,
      "grad_norm": 0.875,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.098,
      "mean_token_accuracy": 0.17283066660165786,
      "num_tokens": 38614961.0,
      "step": 17635
    },
    {
      "entropy": 5.554338598251343,
      "epoch": 1.7506947201270346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.1102,
      "mean_token_accuracy": 0.1769152835011482,
      "num_tokens": 38624706.0,
      "step": 17640
    },
    {
      "entropy": 5.61465220451355,
      "epoch": 1.7511909487892021,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.0972,
      "mean_token_accuracy": 0.17708284705877303,
      "num_tokens": 38636006.0,
      "step": 17645
    },
    {
      "entropy": 5.569019603729248,
      "epoch": 1.7516871774513696,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17268340289592743,
      "num_tokens": 38647826.0,
      "step": 17650
    },
    {
      "entropy": 5.564844846725464,
      "epoch": 1.752183406113537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.1102,
      "mean_token_accuracy": 0.17221035808324814,
      "num_tokens": 38659463.0,
      "step": 17655
    },
    {
      "entropy": 5.605877876281738,
      "epoch": 1.7526796347757045,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.1987,
      "mean_token_accuracy": 0.16510094702243805,
      "num_tokens": 38670593.0,
      "step": 17660
    },
    {
      "entropy": 5.560161685943603,
      "epoch": 1.753175863437872,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1796002060174942,
      "num_tokens": 38681560.0,
      "step": 17665
    },
    {
      "entropy": 5.565459299087524,
      "epoch": 1.7536720921000397,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.0106,
      "mean_token_accuracy": 0.17378750145435334,
      "num_tokens": 38693624.0,
      "step": 17670
    },
    {
      "entropy": 5.585840797424316,
      "epoch": 1.7541683207622072,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1705695539712906,
      "num_tokens": 38704347.0,
      "step": 17675
    },
    {
      "entropy": 5.585345458984375,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.131,
      "mean_token_accuracy": 0.17121277153491973,
      "num_tokens": 38715578.0,
      "step": 17680
    },
    {
      "entropy": 5.504002285003662,
      "epoch": 1.7551607780865424,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.1398,
      "mean_token_accuracy": 0.16965821385383606,
      "num_tokens": 38729539.0,
      "step": 17685
    },
    {
      "entropy": 5.561998319625855,
      "epoch": 1.7556570067487098,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.16,
      "mean_token_accuracy": 0.17648142278194429,
      "num_tokens": 38742031.0,
      "step": 17690
    },
    {
      "entropy": 5.588738632202149,
      "epoch": 1.7561532354108773,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.0533,
      "mean_token_accuracy": 0.17965624183416368,
      "num_tokens": 38754046.0,
      "step": 17695
    },
    {
      "entropy": 5.513803291320801,
      "epoch": 1.7566494640730448,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.0609,
      "mean_token_accuracy": 0.17319609224796295,
      "num_tokens": 38764678.0,
      "step": 17700
    },
    {
      "entropy": 5.588637924194336,
      "epoch": 1.7571456927352123,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.102,
      "mean_token_accuracy": 0.1737789183855057,
      "num_tokens": 38775954.0,
      "step": 17705
    },
    {
      "entropy": 5.606778526306153,
      "epoch": 1.75764192139738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.0787,
      "mean_token_accuracy": 0.17167526632547378,
      "num_tokens": 38785856.0,
      "step": 17710
    },
    {
      "entropy": 5.550051641464234,
      "epoch": 1.7581381500595474,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17436802685260772,
      "num_tokens": 38796811.0,
      "step": 17715
    },
    {
      "entropy": 5.547880935668945,
      "epoch": 1.7586343787217151,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.1851,
      "mean_token_accuracy": 0.17074999064207078,
      "num_tokens": 38808536.0,
      "step": 17720
    },
    {
      "entropy": 5.5596670627594,
      "epoch": 1.7591306073838826,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.0732,
      "mean_token_accuracy": 0.17389263957738876,
      "num_tokens": 38819828.0,
      "step": 17725
    },
    {
      "entropy": 5.608418941497803,
      "epoch": 1.75962683604605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046948618369330533,
      "loss": 5.1319,
      "mean_token_accuracy": 0.17222689539194108,
      "num_tokens": 38830975.0,
      "step": 17730
    },
    {
      "entropy": 5.552244472503662,
      "epoch": 1.7601230647082176,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.0714,
      "mean_token_accuracy": 0.17467777132987977,
      "num_tokens": 38841201.0,
      "step": 17735
    },
    {
      "entropy": 5.567065906524658,
      "epoch": 1.760619293370385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.102,
      "mean_token_accuracy": 0.1730594664812088,
      "num_tokens": 38851570.0,
      "step": 17740
    },
    {
      "entropy": 5.608945894241333,
      "epoch": 1.7611155220325525,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17228432595729828,
      "num_tokens": 38862869.0,
      "step": 17745
    },
    {
      "entropy": 5.5734617710113525,
      "epoch": 1.76161175069472,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.1148,
      "mean_token_accuracy": 0.17154959589242935,
      "num_tokens": 38874162.0,
      "step": 17750
    },
    {
      "entropy": 5.492464399337768,
      "epoch": 1.7621079793568877,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004693970426246635,
      "loss": 4.9958,
      "mean_token_accuracy": 0.18095029443502425,
      "num_tokens": 38884480.0,
      "step": 17755
    },
    {
      "entropy": 5.611266279220581,
      "epoch": 1.7626042080190552,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1648872673511505,
      "num_tokens": 38895623.0,
      "step": 17760
    },
    {
      "entropy": 5.55257568359375,
      "epoch": 1.7631004366812228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.0707,
      "mean_token_accuracy": 0.1775823712348938,
      "num_tokens": 38907286.0,
      "step": 17765
    },
    {
      "entropy": 5.488652896881104,
      "epoch": 1.7635966653433903,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.0269,
      "mean_token_accuracy": 0.1805002883076668,
      "num_tokens": 38919342.0,
      "step": 17770
    },
    {
      "entropy": 5.499943733215332,
      "epoch": 1.7640928940055578,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.0757,
      "mean_token_accuracy": 0.1736500382423401,
      "num_tokens": 38930289.0,
      "step": 17775
    },
    {
      "entropy": 5.5259946346282955,
      "epoch": 1.7645891226677253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.0497,
      "mean_token_accuracy": 0.1753482311964035,
      "num_tokens": 38941053.0,
      "step": 17780
    },
    {
      "entropy": 5.570055675506592,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.076,
      "mean_token_accuracy": 0.1749258294701576,
      "num_tokens": 38952474.0,
      "step": 17785
    },
    {
      "entropy": 5.5248171329498295,
      "epoch": 1.7655815799920602,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.0602,
      "mean_token_accuracy": 0.18056851029396057,
      "num_tokens": 38963049.0,
      "step": 17790
    },
    {
      "entropy": 5.6903918266296385,
      "epoch": 1.766077808654228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.2872,
      "mean_token_accuracy": 0.16286114305257798,
      "num_tokens": 38975290.0,
      "step": 17795
    },
    {
      "entropy": 5.574919605255127,
      "epoch": 1.7665740373163954,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17251146882772445,
      "num_tokens": 38985825.0,
      "step": 17800
    },
    {
      "entropy": 5.6242790699005125,
      "epoch": 1.7670702659785629,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.1738,
      "mean_token_accuracy": 0.16609287708997728,
      "num_tokens": 38998119.0,
      "step": 17805
    },
    {
      "entropy": 5.628975582122803,
      "epoch": 1.7675664946407306,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.1557,
      "mean_token_accuracy": 0.17918626666069032,
      "num_tokens": 39009204.0,
      "step": 17810
    },
    {
      "entropy": 5.479180002212525,
      "epoch": 1.768062723302898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1690761774778366,
      "num_tokens": 39020138.0,
      "step": 17815
    },
    {
      "entropy": 5.543305683135986,
      "epoch": 1.7685589519650655,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000469164711970924,
      "loss": 5.1273,
      "mean_token_accuracy": 0.16922591477632523,
      "num_tokens": 39031590.0,
      "step": 17820
    },
    {
      "entropy": 5.595354843139648,
      "epoch": 1.769055180627233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.2011,
      "mean_token_accuracy": 0.17158617675304413,
      "num_tokens": 39042361.0,
      "step": 17825
    },
    {
      "entropy": 5.600746917724609,
      "epoch": 1.7695514092894005,
      "grad_norm": 1.0,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.0655,
      "mean_token_accuracy": 0.17768279016017913,
      "num_tokens": 39053560.0,
      "step": 17830
    },
    {
      "entropy": 5.5046909809112545,
      "epoch": 1.770047637951568,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.0405,
      "mean_token_accuracy": 0.17571342140436172,
      "num_tokens": 39064172.0,
      "step": 17835
    },
    {
      "entropy": 5.613195466995239,
      "epoch": 1.7705438666137356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.137,
      "mean_token_accuracy": 0.17072379738092422,
      "num_tokens": 39075860.0,
      "step": 17840
    },
    {
      "entropy": 5.587795114517212,
      "epoch": 1.7710400952759031,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.1401,
      "mean_token_accuracy": 0.16498076170682907,
      "num_tokens": 39087148.0,
      "step": 17845
    },
    {
      "entropy": 5.591925096511841,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.1571,
      "mean_token_accuracy": 0.16948329508304597,
      "num_tokens": 39097860.0,
      "step": 17850
    },
    {
      "entropy": 5.5378812789917,
      "epoch": 1.7720325526002383,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.0274,
      "mean_token_accuracy": 0.17340326756238938,
      "num_tokens": 39108475.0,
      "step": 17855
    },
    {
      "entropy": 5.435478258132934,
      "epoch": 1.7725287812624058,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.0907,
      "mean_token_accuracy": 0.16897138059139252,
      "num_tokens": 39119167.0,
      "step": 17860
    },
    {
      "entropy": 5.63852972984314,
      "epoch": 1.7730250099245732,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.1107,
      "mean_token_accuracy": 0.17212416529655455,
      "num_tokens": 39130387.0,
      "step": 17865
    },
    {
      "entropy": 5.553347444534301,
      "epoch": 1.7735212385867407,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046898544232289126,
      "loss": 4.9923,
      "mean_token_accuracy": 0.17875964790582657,
      "num_tokens": 39140673.0,
      "step": 17870
    },
    {
      "entropy": 5.527171039581299,
      "epoch": 1.7740174672489082,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17716972678899764,
      "num_tokens": 39151808.0,
      "step": 17875
    },
    {
      "entropy": 5.523415040969849,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.1233,
      "mean_token_accuracy": 0.17539796531200408,
      "num_tokens": 39163432.0,
      "step": 17880
    },
    {
      "entropy": 5.558017015457153,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.1379,
      "mean_token_accuracy": 0.1685480996966362,
      "num_tokens": 39173802.0,
      "step": 17885
    },
    {
      "entropy": 5.565626859664917,
      "epoch": 1.7755061532354108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.1321,
      "mean_token_accuracy": 0.17639424055814742,
      "num_tokens": 39184475.0,
      "step": 17890
    },
    {
      "entropy": 5.61225004196167,
      "epoch": 1.7760023818975785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.1178,
      "mean_token_accuracy": 0.17048851549625396,
      "num_tokens": 39195607.0,
      "step": 17895
    },
    {
      "entropy": 5.622340488433838,
      "epoch": 1.776498610559746,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004688776496054117,
      "loss": 5.1595,
      "mean_token_accuracy": 0.16812739372253419,
      "num_tokens": 39205721.0,
      "step": 17900
    },
    {
      "entropy": 5.617865324020386,
      "epoch": 1.7769948392219135,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.1199,
      "mean_token_accuracy": 0.17173972874879836,
      "num_tokens": 39216352.0,
      "step": 17905
    },
    {
      "entropy": 5.528640222549439,
      "epoch": 1.777491067884081,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046884168023284787,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18199142515659333,
      "num_tokens": 39225872.0,
      "step": 17910
    },
    {
      "entropy": 5.479281997680664,
      "epoch": 1.7779872965462484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.0259,
      "mean_token_accuracy": 0.17971389889717101,
      "num_tokens": 39237751.0,
      "step": 17915
    },
    {
      "entropy": 5.5109375,
      "epoch": 1.778483525208416,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1777319312095642,
      "num_tokens": 39247854.0,
      "step": 17920
    },
    {
      "entropy": 5.537512016296387,
      "epoch": 1.7789797538705836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.1467,
      "mean_token_accuracy": 0.17625917196273805,
      "num_tokens": 39258220.0,
      "step": 17925
    },
    {
      "entropy": 5.602388048171997,
      "epoch": 1.779475982532751,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.0753,
      "mean_token_accuracy": 0.17355095148086547,
      "num_tokens": 39269303.0,
      "step": 17930
    },
    {
      "entropy": 5.575857162475586,
      "epoch": 1.7799722111949188,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.07,
      "mean_token_accuracy": 0.17262728810310363,
      "num_tokens": 39280048.0,
      "step": 17935
    },
    {
      "entropy": 5.526679277420044,
      "epoch": 1.7804684398570862,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1845131352543831,
      "num_tokens": 39291174.0,
      "step": 17940
    },
    {
      "entropy": 5.539323043823242,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1739985540509224,
      "num_tokens": 39301332.0,
      "step": 17945
    },
    {
      "entropy": 5.569856119155884,
      "epoch": 1.7814608971814212,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.1135,
      "mean_token_accuracy": 0.18091185241937638,
      "num_tokens": 39311935.0,
      "step": 17950
    },
    {
      "entropy": 5.599349403381348,
      "epoch": 1.7819571258435887,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.1725,
      "mean_token_accuracy": 0.16307041943073272,
      "num_tokens": 39321973.0,
      "step": 17955
    },
    {
      "entropy": 5.565472030639649,
      "epoch": 1.7824533545057561,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.1328,
      "mean_token_accuracy": 0.17671920508146285,
      "num_tokens": 39332384.0,
      "step": 17960
    },
    {
      "entropy": 5.508353137969971,
      "epoch": 1.7829495831679236,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004686435050893056,
      "loss": 5.054,
      "mean_token_accuracy": 0.17592542618513107,
      "num_tokens": 39342844.0,
      "step": 17965
    },
    {
      "entropy": 5.551001453399659,
      "epoch": 1.7834458118300913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.1555,
      "mean_token_accuracy": 0.17825542241334916,
      "num_tokens": 39354045.0,
      "step": 17970
    },
    {
      "entropy": 5.608150815963745,
      "epoch": 1.7839420404922588,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.1356,
      "mean_token_accuracy": 0.17056632936000823,
      "num_tokens": 39364529.0,
      "step": 17975
    },
    {
      "entropy": 5.4764769077301025,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18548448830842973,
      "num_tokens": 39375671.0,
      "step": 17980
    },
    {
      "entropy": 5.514445972442627,
      "epoch": 1.784934497816594,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17915773540735244,
      "num_tokens": 39387546.0,
      "step": 17985
    },
    {
      "entropy": 5.6080756187438965,
      "epoch": 1.7854307264787614,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.0934,
      "mean_token_accuracy": 0.179030878841877,
      "num_tokens": 39397726.0,
      "step": 17990
    },
    {
      "entropy": 5.5969010353088375,
      "epoch": 1.785926955140929,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.103,
      "mean_token_accuracy": 0.16994453519582747,
      "num_tokens": 39409869.0,
      "step": 17995
    },
    {
      "entropy": 5.466196298599243,
      "epoch": 1.7864231838030964,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17888710647821426,
      "num_tokens": 39419843.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.308328244940463,
      "eval_loss": 5.212491512298584,
      "eval_mean_token_accuracy": 0.17443242069642684,
      "eval_num_tokens": 39419843.0,
      "eval_runtime": 20.3171,
      "eval_samples_per_second": 1657.174,
      "eval_steps_per_second": 207.165,
      "step": 18000
    },
    {
      "entropy": 5.51349835395813,
      "epoch": 1.7869194124652639,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1756940022110939,
      "num_tokens": 39430307.0,
      "step": 18005
    },
    {
      "entropy": 5.6458783626556395,
      "epoch": 1.7874156411274316,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.232,
      "mean_token_accuracy": 0.1657880574464798,
      "num_tokens": 39442134.0,
      "step": 18010
    },
    {
      "entropy": 5.593474245071411,
      "epoch": 1.787911869789599,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18383324891328812,
      "num_tokens": 39451945.0,
      "step": 18015
    },
    {
      "entropy": 5.562695121765136,
      "epoch": 1.7884080984517667,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1739074096083641,
      "num_tokens": 39463045.0,
      "step": 18020
    },
    {
      "entropy": 5.540069150924682,
      "epoch": 1.7889043271139342,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.0943,
      "mean_token_accuracy": 0.1840682327747345,
      "num_tokens": 39474523.0,
      "step": 18025
    },
    {
      "entropy": 5.540053510665894,
      "epoch": 1.7894005557761017,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046840854904534046,
      "loss": 4.9961,
      "mean_token_accuracy": 0.17804229706525804,
      "num_tokens": 39486005.0,
      "step": 18030
    },
    {
      "entropy": 5.5544578552246096,
      "epoch": 1.7898967844382692,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1745083063840866,
      "num_tokens": 39497418.0,
      "step": 18035
    },
    {
      "entropy": 5.53790545463562,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.0582,
      "mean_token_accuracy": 0.17389774918556214,
      "num_tokens": 39507965.0,
      "step": 18040
    },
    {
      "entropy": 5.62037672996521,
      "epoch": 1.7908892417626041,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046835421327716,
      "loss": 5.1212,
      "mean_token_accuracy": 0.1769772857427597,
      "num_tokens": 39518537.0,
      "step": 18045
    },
    {
      "entropy": 5.528034019470215,
      "epoch": 1.7913854704247716,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.096,
      "mean_token_accuracy": 0.16997410953044892,
      "num_tokens": 39530364.0,
      "step": 18050
    },
    {
      "entropy": 5.603397607803345,
      "epoch": 1.7918816990869393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.1824,
      "mean_token_accuracy": 0.1699381574988365,
      "num_tokens": 39541336.0,
      "step": 18055
    },
    {
      "entropy": 5.562083053588867,
      "epoch": 1.7923779277491068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.1252,
      "mean_token_accuracy": 0.16992536783218384,
      "num_tokens": 39552556.0,
      "step": 18060
    },
    {
      "entropy": 5.585334825515747,
      "epoch": 1.7928741564112745,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.0579,
      "mean_token_accuracy": 0.17492354065179824,
      "num_tokens": 39565036.0,
      "step": 18065
    },
    {
      "entropy": 5.609974765777588,
      "epoch": 1.793370385073442,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.1309,
      "mean_token_accuracy": 0.1726563975214958,
      "num_tokens": 39575542.0,
      "step": 18070
    },
    {
      "entropy": 5.548506736755371,
      "epoch": 1.7938666137356094,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17162472009658813,
      "num_tokens": 39586099.0,
      "step": 18075
    },
    {
      "entropy": 5.507031869888306,
      "epoch": 1.7943628423977769,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.0755,
      "mean_token_accuracy": 0.17794099599123,
      "num_tokens": 39597142.0,
      "step": 18080
    },
    {
      "entropy": 5.538154363632202,
      "epoch": 1.7948590710599444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000468209106958531,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1704799234867096,
      "num_tokens": 39608906.0,
      "step": 18085
    },
    {
      "entropy": 5.5578169345855715,
      "epoch": 1.7953552997221118,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.1164,
      "mean_token_accuracy": 0.17682892978191375,
      "num_tokens": 39620269.0,
      "step": 18090
    },
    {
      "entropy": 5.667745685577392,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.125,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.243,
      "mean_token_accuracy": 0.16961012035608292,
      "num_tokens": 39630742.0,
      "step": 18095
    },
    {
      "entropy": 5.586498212814331,
      "epoch": 1.796347757046447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.1534,
      "mean_token_accuracy": 0.1683826580643654,
      "num_tokens": 39640867.0,
      "step": 18100
    },
    {
      "entropy": 5.508498620986939,
      "epoch": 1.7968439857086145,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17958172261714936,
      "num_tokens": 39651883.0,
      "step": 18105
    },
    {
      "entropy": 5.51050386428833,
      "epoch": 1.7973402143707822,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17652324140071868,
      "num_tokens": 39662224.0,
      "step": 18110
    },
    {
      "entropy": 5.577437877655029,
      "epoch": 1.7978364430329496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.1824,
      "mean_token_accuracy": 0.16782035678625107,
      "num_tokens": 39673847.0,
      "step": 18115
    },
    {
      "entropy": 5.657806539535523,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.0607,
      "mean_token_accuracy": 0.18220536559820175,
      "num_tokens": 39684721.0,
      "step": 18120
    },
    {
      "entropy": 5.606435823440552,
      "epoch": 1.7988289003572846,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17282146215438843,
      "num_tokens": 39695779.0,
      "step": 18125
    },
    {
      "entropy": 5.518237686157226,
      "epoch": 1.799325129019452,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468045495837141,
      "loss": 5.1207,
      "mean_token_accuracy": 0.16789035499095917,
      "num_tokens": 39706635.0,
      "step": 18130
    },
    {
      "entropy": 5.514397287368775,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18135659098625184,
      "num_tokens": 39716934.0,
      "step": 18135
    },
    {
      "entropy": 5.574674797058106,
      "epoch": 1.8003175863437872,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.0814,
      "mean_token_accuracy": 0.17956087440252305,
      "num_tokens": 39728232.0,
      "step": 18140
    },
    {
      "entropy": 5.58008041381836,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.037,
      "mean_token_accuracy": 0.18012687861919402,
      "num_tokens": 39738906.0,
      "step": 18145
    },
    {
      "entropy": 5.510091114044189,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1766292631626129,
      "num_tokens": 39749349.0,
      "step": 18150
    },
    {
      "entropy": 5.488127660751343,
      "epoch": 1.80180627233029,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18085013329982758,
      "num_tokens": 39760082.0,
      "step": 18155
    },
    {
      "entropy": 5.595105791091919,
      "epoch": 1.8023025009924574,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.1042,
      "mean_token_accuracy": 0.1774160847067833,
      "num_tokens": 39771365.0,
      "step": 18160
    },
    {
      "entropy": 5.583909130096435,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17261693477630616,
      "num_tokens": 39781890.0,
      "step": 18165
    },
    {
      "entropy": 5.618559837341309,
      "epoch": 1.8032949583167923,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.1554,
      "mean_token_accuracy": 0.17174603790044785,
      "num_tokens": 39791934.0,
      "step": 18170
    },
    {
      "entropy": 5.608896541595459,
      "epoch": 1.8037911869789598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.1371,
      "mean_token_accuracy": 0.1692329838871956,
      "num_tokens": 39802797.0,
      "step": 18175
    },
    {
      "entropy": 5.576122760772705,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.117,
      "mean_token_accuracy": 0.17571600377559662,
      "num_tokens": 39813457.0,
      "step": 18180
    },
    {
      "entropy": 5.5675252914428714,
      "epoch": 1.804783644303295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000467845000139629,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16955844163894654,
      "num_tokens": 39825771.0,
      "step": 18185
    },
    {
      "entropy": 5.597385215759277,
      "epoch": 1.8052798729654624,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000467826744556148,
      "loss": 5.0776,
      "mean_token_accuracy": 0.17580542862415313,
      "num_tokens": 39836388.0,
      "step": 18190
    },
    {
      "entropy": 5.575229930877685,
      "epoch": 1.8057761016276301,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.1718,
      "mean_token_accuracy": 0.16843862235546112,
      "num_tokens": 39847757.0,
      "step": 18195
    },
    {
      "entropy": 5.534408760070801,
      "epoch": 1.8062723302897976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.0801,
      "mean_token_accuracy": 0.16710518151521683,
      "num_tokens": 39858181.0,
      "step": 18200
    },
    {
      "entropy": 5.643390417098999,
      "epoch": 1.806768558951965,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.2331,
      "mean_token_accuracy": 0.16405104994773864,
      "num_tokens": 39871104.0,
      "step": 18205
    },
    {
      "entropy": 5.571788167953491,
      "epoch": 1.8072647876141326,
      "grad_norm": 0.875,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17465649396181107,
      "num_tokens": 39883323.0,
      "step": 18210
    },
    {
      "entropy": 5.478519058227539,
      "epoch": 1.8077610162763,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.0846,
      "mean_token_accuracy": 0.1786652311682701,
      "num_tokens": 39894246.0,
      "step": 18215
    },
    {
      "entropy": 5.512026786804199,
      "epoch": 1.8082572449384675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.089,
      "mean_token_accuracy": 0.172767174243927,
      "num_tokens": 39905208.0,
      "step": 18220
    },
    {
      "entropy": 5.56547212600708,
      "epoch": 1.8087534736006352,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.073,
      "mean_token_accuracy": 0.17250906974077224,
      "num_tokens": 39915531.0,
      "step": 18225
    },
    {
      "entropy": 5.54920711517334,
      "epoch": 1.8092497022628027,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.1788,
      "mean_token_accuracy": 0.17227038294076918,
      "num_tokens": 39926241.0,
      "step": 18230
    },
    {
      "entropy": 5.548025894165039,
      "epoch": 1.8097459309249704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.0772,
      "mean_token_accuracy": 0.18207720667123795,
      "num_tokens": 39936500.0,
      "step": 18235
    },
    {
      "entropy": 5.529899215698242,
      "epoch": 1.8102421595871379,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17328398078680038,
      "num_tokens": 39948865.0,
      "step": 18240
    },
    {
      "entropy": 5.564827823638916,
      "epoch": 1.8107383882493053,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.1629,
      "mean_token_accuracy": 0.16937393844127654,
      "num_tokens": 39960626.0,
      "step": 18245
    },
    {
      "entropy": 5.591396999359131,
      "epoch": 1.8112346169114728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000467607304712321,
      "loss": 5.1282,
      "mean_token_accuracy": 0.16578252166509627,
      "num_tokens": 39971100.0,
      "step": 18250
    },
    {
      "entropy": 5.602129936218262,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.0856,
      "mean_token_accuracy": 0.17259714603424073,
      "num_tokens": 39981332.0,
      "step": 18255
    },
    {
      "entropy": 5.563432931900024,
      "epoch": 1.8122270742358078,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17455360144376755,
      "num_tokens": 39992739.0,
      "step": 18260
    },
    {
      "entropy": 5.5184930801391605,
      "epoch": 1.8127233028979752,
      "grad_norm": 1.09375,
      "learning_rate": 0.000467552337255195,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18556706309318544,
      "num_tokens": 40002841.0,
      "step": 18265
    },
    {
      "entropy": 5.663113975524903,
      "epoch": 1.813219531560143,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.1777,
      "mean_token_accuracy": 0.17586639523506165,
      "num_tokens": 40014176.0,
      "step": 18270
    },
    {
      "entropy": 5.582763957977295,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.0759,
      "mean_token_accuracy": 0.17411395013332367,
      "num_tokens": 40024842.0,
      "step": 18275
    },
    {
      "entropy": 5.547998905181885,
      "epoch": 1.814211988884478,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.0753,
      "mean_token_accuracy": 0.177666375041008,
      "num_tokens": 40034838.0,
      "step": 18280
    },
    {
      "entropy": 5.470091724395752,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.0,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17780368775129318,
      "num_tokens": 40045019.0,
      "step": 18285
    },
    {
      "entropy": 5.590621471405029,
      "epoch": 1.815204446208813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.254,
      "mean_token_accuracy": 0.15972820967435836,
      "num_tokens": 40057283.0,
      "step": 18290
    },
    {
      "entropy": 5.632390117645263,
      "epoch": 1.8157006748709805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.1447,
      "mean_token_accuracy": 0.16668666154146194,
      "num_tokens": 40067542.0,
      "step": 18295
    },
    {
      "entropy": 5.608241128921509,
      "epoch": 1.816196903533148,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.0927,
      "mean_token_accuracy": 0.17656283974647521,
      "num_tokens": 40077538.0,
      "step": 18300
    },
    {
      "entropy": 5.542082595825195,
      "epoch": 1.8166931321953155,
      "grad_norm": 1.109375,
      "learning_rate": 0.000467405547308025,
      "loss": 5.0501,
      "mean_token_accuracy": 0.17996515482664108,
      "num_tokens": 40088206.0,
      "step": 18305
    },
    {
      "entropy": 5.493988704681397,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.0,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16719171702861785,
      "num_tokens": 40100510.0,
      "step": 18310
    },
    {
      "entropy": 5.581872367858887,
      "epoch": 1.8176855895196506,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.103,
      "mean_token_accuracy": 0.1766183689236641,
      "num_tokens": 40110759.0,
      "step": 18315
    },
    {
      "entropy": 5.55856146812439,
      "epoch": 1.8181818181818183,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.1606,
      "mean_token_accuracy": 0.17302494794130324,
      "num_tokens": 40122140.0,
      "step": 18320
    },
    {
      "entropy": 5.624567747116089,
      "epoch": 1.8186780468439858,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.1729,
      "mean_token_accuracy": 0.16987254917621614,
      "num_tokens": 40133479.0,
      "step": 18325
    },
    {
      "entropy": 5.5644862174987795,
      "epoch": 1.8191742755061533,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.0998,
      "mean_token_accuracy": 0.16784086376428603,
      "num_tokens": 40144776.0,
      "step": 18330
    },
    {
      "entropy": 5.632618618011475,
      "epoch": 1.8196705041683208,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.143,
      "mean_token_accuracy": 0.17694283425807952,
      "num_tokens": 40155767.0,
      "step": 18335
    },
    {
      "entropy": 5.6285138607025145,
      "epoch": 1.8201667328304882,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.1926,
      "mean_token_accuracy": 0.16762510538101197,
      "num_tokens": 40167040.0,
      "step": 18340
    },
    {
      "entropy": 5.574040079116822,
      "epoch": 1.8206629614926557,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.0848,
      "mean_token_accuracy": 0.1799247995018959,
      "num_tokens": 40177983.0,
      "step": 18345
    },
    {
      "entropy": 5.613712215423584,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.1125,
      "mean_token_accuracy": 0.17110036164522172,
      "num_tokens": 40187781.0,
      "step": 18350
    },
    {
      "entropy": 5.545492935180664,
      "epoch": 1.821655418816991,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.1063,
      "mean_token_accuracy": 0.17257779836654663,
      "num_tokens": 40198186.0,
      "step": 18355
    },
    {
      "entropy": 5.500664472579956,
      "epoch": 1.8221516474791584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17269857227802277,
      "num_tokens": 40207656.0,
      "step": 18360
    },
    {
      "entropy": 5.543660306930542,
      "epoch": 1.822647876141326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17730205357074738,
      "num_tokens": 40219482.0,
      "step": 18365
    },
    {
      "entropy": 5.6155126094818115,
      "epoch": 1.8231441048034935,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.1188,
      "mean_token_accuracy": 0.17337943017482757,
      "num_tokens": 40230623.0,
      "step": 18370
    },
    {
      "entropy": 5.538521003723145,
      "epoch": 1.823640333465661,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.0429,
      "mean_token_accuracy": 0.17731239199638366,
      "num_tokens": 40240682.0,
      "step": 18375
    },
    {
      "entropy": 5.520022296905518,
      "epoch": 1.8241365621278285,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.1141,
      "mean_token_accuracy": 0.17411809414625168,
      "num_tokens": 40253085.0,
      "step": 18380
    },
    {
      "entropy": 5.589762783050537,
      "epoch": 1.824632790789996,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.0646,
      "mean_token_accuracy": 0.1756399840116501,
      "num_tokens": 40264056.0,
      "step": 18385
    },
    {
      "entropy": 5.6239055633544925,
      "epoch": 1.8251290194521634,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17289505749940873,
      "num_tokens": 40273974.0,
      "step": 18390
    },
    {
      "entropy": 5.502289628982544,
      "epoch": 1.8256252481143311,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.0705,
      "mean_token_accuracy": 0.17695866972208024,
      "num_tokens": 40284455.0,
      "step": 18395
    },
    {
      "entropy": 5.542024087905884,
      "epoch": 1.8261214767764986,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000467055697788742,
      "loss": 5.0837,
      "mean_token_accuracy": 0.17438217550516127,
      "num_tokens": 40294598.0,
      "step": 18400
    },
    {
      "entropy": 5.630302762985229,
      "epoch": 1.826617705438666,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.1408,
      "mean_token_accuracy": 0.1708901271224022,
      "num_tokens": 40306053.0,
      "step": 18405
    },
    {
      "entropy": 5.592396879196167,
      "epoch": 1.8271139341008338,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.1474,
      "mean_token_accuracy": 0.17188413739204406,
      "num_tokens": 40316714.0,
      "step": 18410
    },
    {
      "entropy": 5.594588947296143,
      "epoch": 1.8276101627630013,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.1382,
      "mean_token_accuracy": 0.16857328712940217,
      "num_tokens": 40327696.0,
      "step": 18415
    },
    {
      "entropy": 5.615906620025635,
      "epoch": 1.8281063914251687,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.1254,
      "mean_token_accuracy": 0.17151508182287217,
      "num_tokens": 40338958.0,
      "step": 18420
    },
    {
      "entropy": 5.627157926559448,
      "epoch": 1.8286026200873362,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.2051,
      "mean_token_accuracy": 0.16518543809652328,
      "num_tokens": 40350869.0,
      "step": 18425
    },
    {
      "entropy": 5.535977602005005,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046694486165206573,
      "loss": 5.114,
      "mean_token_accuracy": 0.17391082197427749,
      "num_tokens": 40360932.0,
      "step": 18430
    },
    {
      "entropy": 5.573984527587891,
      "epoch": 1.8295950774116712,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17711306512355804,
      "num_tokens": 40372159.0,
      "step": 18435
    },
    {
      "entropy": 5.556112289428711,
      "epoch": 1.8300913060738389,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.1347,
      "mean_token_accuracy": 0.17568773180246353,
      "num_tokens": 40385268.0,
      "step": 18440
    },
    {
      "entropy": 5.592760372161865,
      "epoch": 1.8305875347360063,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17313764840364457,
      "num_tokens": 40396585.0,
      "step": 18445
    },
    {
      "entropy": 5.573780727386475,
      "epoch": 1.831083763398174,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1772050991654396,
      "num_tokens": 40408348.0,
      "step": 18450
    },
    {
      "entropy": 5.651212406158447,
      "epoch": 1.8315799920603415,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17320408970117568,
      "num_tokens": 40419900.0,
      "step": 18455
    },
    {
      "entropy": 5.587758207321167,
      "epoch": 1.832076220722509,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.1145,
      "mean_token_accuracy": 0.1736462637782097,
      "num_tokens": 40431110.0,
      "step": 18460
    },
    {
      "entropy": 5.505968618392944,
      "epoch": 1.8325724493846764,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18027226775884628,
      "num_tokens": 40442741.0,
      "step": 18465
    },
    {
      "entropy": 5.488784122467041,
      "epoch": 1.833068678046844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.0403,
      "mean_token_accuracy": 0.17639591544866562,
      "num_tokens": 40453720.0,
      "step": 18470
    },
    {
      "entropy": 5.568374443054199,
      "epoch": 1.8335649067090114,
      "grad_norm": 0.875,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.1533,
      "mean_token_accuracy": 0.16489337980747223,
      "num_tokens": 40466097.0,
      "step": 18475
    },
    {
      "entropy": 5.57383713722229,
      "epoch": 1.8340611353711789,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.1058,
      "mean_token_accuracy": 0.17383286505937576,
      "num_tokens": 40475713.0,
      "step": 18480
    },
    {
      "entropy": 5.525574779510498,
      "epoch": 1.8345573640333466,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046674121709660855,
      "loss": 5.0038,
      "mean_token_accuracy": 0.1800436332821846,
      "num_tokens": 40486603.0,
      "step": 18485
    },
    {
      "entropy": 5.6012046337127686,
      "epoch": 1.835053592695514,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000466722675417401,
      "loss": 5.1965,
      "mean_token_accuracy": 0.16458699107170105,
      "num_tokens": 40498057.0,
      "step": 18490
    },
    {
      "entropy": 5.604301071166992,
      "epoch": 1.8355498213576817,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17347632497549056,
      "num_tokens": 40508798.0,
      "step": 18495
    },
    {
      "entropy": 5.553925275802612,
      "epoch": 1.8360460500198492,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.0845,
      "mean_token_accuracy": 0.17613804787397386,
      "num_tokens": 40519736.0,
      "step": 18500
    },
    {
      "entropy": 5.581607532501221,
      "epoch": 1.8365422786820167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.1122,
      "mean_token_accuracy": 0.17420735359191894,
      "num_tokens": 40531582.0,
      "step": 18505
    },
    {
      "entropy": 5.528589153289795,
      "epoch": 1.8370385073441842,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.0425,
      "mean_token_accuracy": 0.17240664660930632,
      "num_tokens": 40542427.0,
      "step": 18510
    },
    {
      "entropy": 5.645935153961181,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.2517,
      "mean_token_accuracy": 0.17054003328084946,
      "num_tokens": 40553432.0,
      "step": 18515
    },
    {
      "entropy": 5.627157497406006,
      "epoch": 1.8380309646685191,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.123,
      "mean_token_accuracy": 0.16991243809461593,
      "num_tokens": 40564995.0,
      "step": 18520
    },
    {
      "entropy": 5.573887729644776,
      "epoch": 1.8385271933306868,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.0967,
      "mean_token_accuracy": 0.17760400772094725,
      "num_tokens": 40574856.0,
      "step": 18525
    },
    {
      "entropy": 5.567026996612549,
      "epoch": 1.8390234219928543,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.1041,
      "mean_token_accuracy": 0.17854958027601242,
      "num_tokens": 40585439.0,
      "step": 18530
    },
    {
      "entropy": 5.582536077499389,
      "epoch": 1.839519650655022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.1155,
      "mean_token_accuracy": 0.17151283025741576,
      "num_tokens": 40596275.0,
      "step": 18535
    },
    {
      "entropy": 5.519898223876953,
      "epoch": 1.8400158793171895,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1741112619638443,
      "num_tokens": 40607586.0,
      "step": 18540
    },
    {
      "entropy": 5.555807161331177,
      "epoch": 1.840512107979357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17188745141029357,
      "num_tokens": 40617339.0,
      "step": 18545
    },
    {
      "entropy": 5.609269285202027,
      "epoch": 1.8410083366415244,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.0737,
      "mean_token_accuracy": 0.18910609930753708,
      "num_tokens": 40630426.0,
      "step": 18550
    },
    {
      "entropy": 5.5866772651672365,
      "epoch": 1.8415045653036919,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.1197,
      "mean_token_accuracy": 0.17397536635398864,
      "num_tokens": 40640694.0,
      "step": 18555
    },
    {
      "entropy": 5.608185577392578,
      "epoch": 1.8420007939658594,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.1619,
      "mean_token_accuracy": 0.17065961807966232,
      "num_tokens": 40651131.0,
      "step": 18560
    },
    {
      "entropy": 5.633810997009277,
      "epoch": 1.8424970226280268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.2101,
      "mean_token_accuracy": 0.1699725344777107,
      "num_tokens": 40662228.0,
      "step": 18565
    },
    {
      "entropy": 5.59955620765686,
      "epoch": 1.8429932512901945,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.0774,
      "mean_token_accuracy": 0.17381449043750763,
      "num_tokens": 40672928.0,
      "step": 18570
    },
    {
      "entropy": 5.593259382247925,
      "epoch": 1.843489479952362,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.1662,
      "mean_token_accuracy": 0.16910812109708787,
      "num_tokens": 40684578.0,
      "step": 18575
    },
    {
      "entropy": 5.561435270309448,
      "epoch": 1.8439857086145297,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17207329124212264,
      "num_tokens": 40694844.0,
      "step": 18580
    },
    {
      "entropy": 5.537709093093872,
      "epoch": 1.8444819372766972,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17422432005405425,
      "num_tokens": 40706494.0,
      "step": 18585
    },
    {
      "entropy": 5.612421703338623,
      "epoch": 1.8449781659388647,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17035920023918152,
      "num_tokens": 40717092.0,
      "step": 18590
    },
    {
      "entropy": 5.53544135093689,
      "epoch": 1.8454743946010321,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18356480896472932,
      "num_tokens": 40729131.0,
      "step": 18595
    },
    {
      "entropy": 5.519265270233154,
      "epoch": 1.8459706232631996,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046631355645014875,
      "loss": 5.0445,
      "mean_token_accuracy": 0.17369847744703293,
      "num_tokens": 40739918.0,
      "step": 18600
    },
    {
      "entropy": 5.593945550918579,
      "epoch": 1.846466851925367,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.1558,
      "mean_token_accuracy": 0.17199598401784896,
      "num_tokens": 40752136.0,
      "step": 18605
    },
    {
      "entropy": 5.584518527984619,
      "epoch": 1.8469630805875348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.0484,
      "mean_token_accuracy": 0.18304478377103806,
      "num_tokens": 40761613.0,
      "step": 18610
    },
    {
      "entropy": 5.546965599060059,
      "epoch": 1.8474593092497023,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.1203,
      "mean_token_accuracy": 0.17571787387132645,
      "num_tokens": 40773405.0,
      "step": 18615
    },
    {
      "entropy": 5.623188495635986,
      "epoch": 1.84795553791187,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.1917,
      "mean_token_accuracy": 0.16865300685167312,
      "num_tokens": 40784795.0,
      "step": 18620
    },
    {
      "entropy": 5.596748352050781,
      "epoch": 1.8484517665740374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1704618290066719,
      "num_tokens": 40795946.0,
      "step": 18625
    },
    {
      "entropy": 5.571191120147705,
      "epoch": 1.848947995236205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.077,
      "mean_token_accuracy": 0.17797884345054626,
      "num_tokens": 40807854.0,
      "step": 18630
    },
    {
      "entropy": 5.549201583862304,
      "epoch": 1.8494442238983724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17689341455698013,
      "num_tokens": 40818677.0,
      "step": 18635
    },
    {
      "entropy": 5.579242324829101,
      "epoch": 1.8499404525605398,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.1086,
      "mean_token_accuracy": 0.16957387775182725,
      "num_tokens": 40829927.0,
      "step": 18640
    },
    {
      "entropy": 5.516990375518799,
      "epoch": 1.8504366812227073,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17352605164051055,
      "num_tokens": 40840661.0,
      "step": 18645
    },
    {
      "entropy": 5.638862133026123,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16971901059150696,
      "num_tokens": 40851985.0,
      "step": 18650
    },
    {
      "entropy": 5.622923040390015,
      "epoch": 1.8514291385470425,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.1221,
      "mean_token_accuracy": 0.17752483040094375,
      "num_tokens": 40863546.0,
      "step": 18655
    },
    {
      "entropy": 5.634589433670044,
      "epoch": 1.85192536720921,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000466089432526647,
      "loss": 5.0938,
      "mean_token_accuracy": 0.16963127106428147,
      "num_tokens": 40873555.0,
      "step": 18660
    },
    {
      "entropy": 5.652401113510132,
      "epoch": 1.8524215958713777,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.2312,
      "mean_token_accuracy": 0.16575287133455277,
      "num_tokens": 40887483.0,
      "step": 18665
    },
    {
      "entropy": 5.617118787765503,
      "epoch": 1.8529178245335451,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.2188,
      "mean_token_accuracy": 0.16653105318546296,
      "num_tokens": 40900412.0,
      "step": 18670
    },
    {
      "entropy": 5.599985551834107,
      "epoch": 1.8534140531957126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17577607929706573,
      "num_tokens": 40910069.0,
      "step": 18675
    },
    {
      "entropy": 5.597972202301025,
      "epoch": 1.85391028185788,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.0995,
      "mean_token_accuracy": 0.1789856657385826,
      "num_tokens": 40921321.0,
      "step": 18680
    },
    {
      "entropy": 5.586864328384399,
      "epoch": 1.8544065105200476,
      "grad_norm": 1.25,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.0815,
      "mean_token_accuracy": 0.17530633211135865,
      "num_tokens": 40931564.0,
      "step": 18685
    },
    {
      "entropy": 5.546919250488282,
      "epoch": 1.854902739182215,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.076,
      "mean_token_accuracy": 0.1729219824075699,
      "num_tokens": 40942537.0,
      "step": 18690
    },
    {
      "entropy": 5.602327966690064,
      "epoch": 1.8553989678443827,
      "grad_norm": 1.0,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.1227,
      "mean_token_accuracy": 0.17576497048139572,
      "num_tokens": 40953583.0,
      "step": 18695
    },
    {
      "entropy": 5.588066148757934,
      "epoch": 1.8558951965065502,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046593963743976163,
      "loss": 5.0835,
      "mean_token_accuracy": 0.1738061174750328,
      "num_tokens": 40964438.0,
      "step": 18700
    },
    {
      "entropy": 5.57550015449524,
      "epoch": 1.8563914251687177,
      "grad_norm": 1.046875,
      "learning_rate": 0.000465920891742011,
      "loss": 5.0438,
      "mean_token_accuracy": 0.17603249698877335,
      "num_tokens": 40975154.0,
      "step": 18705
    },
    {
      "entropy": 5.53535099029541,
      "epoch": 1.8568876538308854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17403220683336257,
      "num_tokens": 40986498.0,
      "step": 18710
    },
    {
      "entropy": 5.5731181621551515,
      "epoch": 1.8573838824930529,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.097,
      "mean_token_accuracy": 0.17261353135108948,
      "num_tokens": 40996095.0,
      "step": 18715
    },
    {
      "entropy": 5.647522592544556,
      "epoch": 1.8578801111552203,
      "grad_norm": 1.0,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17163667529821397,
      "num_tokens": 41007583.0,
      "step": 18720
    },
    {
      "entropy": 5.626029300689697,
      "epoch": 1.8583763398173878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.1291,
      "mean_token_accuracy": 0.16911496669054032,
      "num_tokens": 41018251.0,
      "step": 18725
    },
    {
      "entropy": 5.598599004745483,
      "epoch": 1.8588725684795553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.1454,
      "mean_token_accuracy": 0.1751018673181534,
      "num_tokens": 41029144.0,
      "step": 18730
    },
    {
      "entropy": 5.621309614181518,
      "epoch": 1.8593687971417228,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17074748277664184,
      "num_tokens": 41039385.0,
      "step": 18735
    },
    {
      "entropy": 5.531555318832398,
      "epoch": 1.8598650258038905,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18166848123073578,
      "num_tokens": 41049175.0,
      "step": 18740
    },
    {
      "entropy": 5.5256187438964846,
      "epoch": 1.860361254466058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17808591276407243,
      "num_tokens": 41060517.0,
      "step": 18745
    },
    {
      "entropy": 5.5424909591674805,
      "epoch": 1.8608574831282256,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.1304,
      "mean_token_accuracy": 0.1742803931236267,
      "num_tokens": 41070957.0,
      "step": 18750
    },
    {
      "entropy": 5.606602144241333,
      "epoch": 1.861353711790393,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.1802,
      "mean_token_accuracy": 0.166935595870018,
      "num_tokens": 41083079.0,
      "step": 18755
    },
    {
      "entropy": 5.665310907363891,
      "epoch": 1.8618499404525606,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.2182,
      "mean_token_accuracy": 0.16876950711011887,
      "num_tokens": 41095379.0,
      "step": 18760
    },
    {
      "entropy": 5.544263696670532,
      "epoch": 1.862346169114728,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17967138737440108,
      "num_tokens": 41106412.0,
      "step": 18765
    },
    {
      "entropy": 5.583419990539551,
      "epoch": 1.8628423977768955,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.1649,
      "mean_token_accuracy": 0.17827685475349425,
      "num_tokens": 41116820.0,
      "step": 18770
    },
    {
      "entropy": 5.557677316665649,
      "epoch": 1.863338626439063,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.0986,
      "mean_token_accuracy": 0.1738136127591133,
      "num_tokens": 41128657.0,
      "step": 18775
    },
    {
      "entropy": 5.634247827529907,
      "epoch": 1.8638348551012305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000465639138405143,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17476013153791428,
      "num_tokens": 41139700.0,
      "step": 18780
    },
    {
      "entropy": 5.604368019104004,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.125,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17009503543376922,
      "num_tokens": 41150018.0,
      "step": 18785
    },
    {
      "entropy": 5.592474269866943,
      "epoch": 1.8648273124255657,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.1752,
      "mean_token_accuracy": 0.16917942315340043,
      "num_tokens": 41160683.0,
      "step": 18790
    },
    {
      "entropy": 5.595214605331421,
      "epoch": 1.8653235410877333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004655826600465427,
      "loss": 5.1162,
      "mean_token_accuracy": 0.17839142680168152,
      "num_tokens": 41171602.0,
      "step": 18795
    },
    {
      "entropy": 5.615164852142334,
      "epoch": 1.8658197697499008,
      "grad_norm": 1.0,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.0973,
      "mean_token_accuracy": 0.17292211353778839,
      "num_tokens": 41182087.0,
      "step": 18800
    },
    {
      "entropy": 5.527322912216187,
      "epoch": 1.8663159984120683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18410271406173706,
      "num_tokens": 41194472.0,
      "step": 18805
    },
    {
      "entropy": 5.5331642627716064,
      "epoch": 1.8668122270742358,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.0984,
      "mean_token_accuracy": 0.17751652449369432,
      "num_tokens": 41205082.0,
      "step": 18810
    },
    {
      "entropy": 5.584650468826294,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.0985,
      "mean_token_accuracy": 0.18302377313375473,
      "num_tokens": 41214362.0,
      "step": 18815
    },
    {
      "entropy": 5.617903470993042,
      "epoch": 1.8678046843985707,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.1026,
      "mean_token_accuracy": 0.1806028500199318,
      "num_tokens": 41225108.0,
      "step": 18820
    },
    {
      "entropy": 5.590192985534668,
      "epoch": 1.8683009130607384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.0773,
      "mean_token_accuracy": 0.1799865871667862,
      "num_tokens": 41235771.0,
      "step": 18825
    },
    {
      "entropy": 5.519165134429931,
      "epoch": 1.868797141722906,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046545071182024,
      "loss": 5.0876,
      "mean_token_accuracy": 0.17361727356910706,
      "num_tokens": 41245752.0,
      "step": 18830
    },
    {
      "entropy": 5.632931900024414,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.0,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.2218,
      "mean_token_accuracy": 0.1666522741317749,
      "num_tokens": 41255704.0,
      "step": 18835
    },
    {
      "entropy": 5.585506343841553,
      "epoch": 1.869789599047241,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000465412969815091,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17775716185569762,
      "num_tokens": 41265742.0,
      "step": 18840
    },
    {
      "entropy": 5.57513108253479,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.1181,
      "mean_token_accuracy": 0.1707280695438385,
      "num_tokens": 41276166.0,
      "step": 18845
    },
    {
      "entropy": 5.585367727279663,
      "epoch": 1.870782056371576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.111,
      "mean_token_accuracy": 0.1768019050359726,
      "num_tokens": 41287047.0,
      "step": 18850
    },
    {
      "entropy": 5.563522005081177,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.1046,
      "mean_token_accuracy": 0.17190309017896652,
      "num_tokens": 41297317.0,
      "step": 18855
    },
    {
      "entropy": 5.596819019317627,
      "epoch": 1.871774513695911,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.1575,
      "mean_token_accuracy": 0.17454561442136765,
      "num_tokens": 41308107.0,
      "step": 18860
    },
    {
      "entropy": 5.592664480209351,
      "epoch": 1.8722707423580784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.0438,
      "mean_token_accuracy": 0.1876957818865776,
      "num_tokens": 41320026.0,
      "step": 18865
    },
    {
      "entropy": 5.652701711654663,
      "epoch": 1.8727669710202461,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.1627,
      "mean_token_accuracy": 0.1727445274591446,
      "num_tokens": 41331492.0,
      "step": 18870
    },
    {
      "entropy": 5.549637079238892,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.0422,
      "mean_token_accuracy": 0.17107570320367813,
      "num_tokens": 41342258.0,
      "step": 18875
    },
    {
      "entropy": 5.610552740097046,
      "epoch": 1.8737594283445813,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.1186,
      "mean_token_accuracy": 0.17838337123394013,
      "num_tokens": 41353041.0,
      "step": 18880
    },
    {
      "entropy": 5.575196981430054,
      "epoch": 1.8742556570067488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.1484,
      "mean_token_accuracy": 0.17448002845048904,
      "num_tokens": 41364517.0,
      "step": 18885
    },
    {
      "entropy": 5.5387022495269775,
      "epoch": 1.8747518856689163,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.1349,
      "mean_token_accuracy": 0.17613407522439956,
      "num_tokens": 41376241.0,
      "step": 18890
    },
    {
      "entropy": 5.57519793510437,
      "epoch": 1.8752481143310837,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17599423378705978,
      "num_tokens": 41387627.0,
      "step": 18895
    },
    {
      "entropy": 5.607448530197144,
      "epoch": 1.8757443429932512,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.0926,
      "mean_token_accuracy": 0.18097908794879913,
      "num_tokens": 41399803.0,
      "step": 18900
    },
    {
      "entropy": 5.541554880142212,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17023009657859803,
      "num_tokens": 41409735.0,
      "step": 18905
    },
    {
      "entropy": 5.531721210479736,
      "epoch": 1.8767368003175864,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.0304,
      "mean_token_accuracy": 0.1846806898713112,
      "num_tokens": 41420928.0,
      "step": 18910
    },
    {
      "entropy": 5.5635871410369875,
      "epoch": 1.8772330289797539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.1183,
      "mean_token_accuracy": 0.17753703147172928,
      "num_tokens": 41432309.0,
      "step": 18915
    },
    {
      "entropy": 5.612442064285278,
      "epoch": 1.8777292576419216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.1063,
      "mean_token_accuracy": 0.177027228474617,
      "num_tokens": 41443064.0,
      "step": 18920
    },
    {
      "entropy": 5.618899202346801,
      "epoch": 1.878225486304089,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17555666863918304,
      "num_tokens": 41452698.0,
      "step": 18925
    },
    {
      "entropy": 5.5655121326446535,
      "epoch": 1.8787217149662565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.0852,
      "mean_token_accuracy": 0.17314388155937194,
      "num_tokens": 41463253.0,
      "step": 18930
    },
    {
      "entropy": 5.562196207046509,
      "epoch": 1.879217943628424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.1011,
      "mean_token_accuracy": 0.173806095123291,
      "num_tokens": 41473635.0,
      "step": 18935
    },
    {
      "entropy": 5.578678369522095,
      "epoch": 1.8797141722905915,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.074,
      "mean_token_accuracy": 0.17772893607616425,
      "num_tokens": 41485705.0,
      "step": 18940
    },
    {
      "entropy": 5.647975635528565,
      "epoch": 1.880210400952759,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.1849,
      "mean_token_accuracy": 0.166853129863739,
      "num_tokens": 41497604.0,
      "step": 18945
    },
    {
      "entropy": 5.4962873458862305,
      "epoch": 1.8807066296149264,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17501992285251616,
      "num_tokens": 41508960.0,
      "step": 18950
    },
    {
      "entropy": 5.535485744476318,
      "epoch": 1.881202858277094,
      "grad_norm": 1.125,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.1167,
      "mean_token_accuracy": 0.17563883811235428,
      "num_tokens": 41519494.0,
      "step": 18955
    },
    {
      "entropy": 5.643306684494019,
      "epoch": 1.8816990869392616,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.1185,
      "mean_token_accuracy": 0.1679257407784462,
      "num_tokens": 41530731.0,
      "step": 18960
    },
    {
      "entropy": 5.553531646728516,
      "epoch": 1.8821953156014293,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17247798442840576,
      "num_tokens": 41541839.0,
      "step": 18965
    },
    {
      "entropy": 5.556876945495605,
      "epoch": 1.8826915442635968,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.0891,
      "mean_token_accuracy": 0.17661537677049638,
      "num_tokens": 41552413.0,
      "step": 18970
    },
    {
      "entropy": 5.658810710906982,
      "epoch": 1.8831877729257642,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.2138,
      "mean_token_accuracy": 0.170829676091671,
      "num_tokens": 41563400.0,
      "step": 18975
    },
    {
      "entropy": 5.529514598846435,
      "epoch": 1.8836840015879317,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17661979645490647,
      "num_tokens": 41575902.0,
      "step": 18980
    },
    {
      "entropy": 5.561712265014648,
      "epoch": 1.8841802302500992,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.127,
      "mean_token_accuracy": 0.17369654029607773,
      "num_tokens": 41587613.0,
      "step": 18985
    },
    {
      "entropy": 5.567292261123657,
      "epoch": 1.8846764589122667,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.0926,
      "mean_token_accuracy": 0.18484259992837906,
      "num_tokens": 41598674.0,
      "step": 18990
    },
    {
      "entropy": 5.57949595451355,
      "epoch": 1.8851726875744343,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.1442,
      "mean_token_accuracy": 0.16751063466072083,
      "num_tokens": 41609891.0,
      "step": 18995
    },
    {
      "entropy": 5.514576482772827,
      "epoch": 1.8856689162366018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.0517,
      "mean_token_accuracy": 0.17840394973754883,
      "num_tokens": 41622066.0,
      "step": 19000
    },
    {
      "entropy": 5.5350254535675045,
      "epoch": 1.8861651448987695,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.0476,
      "mean_token_accuracy": 0.1790388911962509,
      "num_tokens": 41633182.0,
      "step": 19005
    },
    {
      "entropy": 5.527015495300293,
      "epoch": 1.886661373560937,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.0743,
      "mean_token_accuracy": 0.18139501065015792,
      "num_tokens": 41644059.0,
      "step": 19010
    },
    {
      "entropy": 5.510572719573974,
      "epoch": 1.8871576022231045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004647494306803764,
      "loss": 5.0116,
      "mean_token_accuracy": 0.17385751605033875,
      "num_tokens": 41654225.0,
      "step": 19015
    },
    {
      "entropy": 5.589991474151612,
      "epoch": 1.887653830885272,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.1501,
      "mean_token_accuracy": 0.1701624274253845,
      "num_tokens": 41666598.0,
      "step": 19020
    },
    {
      "entropy": 5.626532554626465,
      "epoch": 1.8881500595474394,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.1293,
      "mean_token_accuracy": 0.17510353922843933,
      "num_tokens": 41677408.0,
      "step": 19025
    },
    {
      "entropy": 5.519022464752197,
      "epoch": 1.888646288209607,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.0323,
      "mean_token_accuracy": 0.17371483743190766,
      "num_tokens": 41687350.0,
      "step": 19030
    },
    {
      "entropy": 5.565056085586548,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.1266,
      "mean_token_accuracy": 0.17109303623437883,
      "num_tokens": 41698012.0,
      "step": 19035
    },
    {
      "entropy": 5.507843065261841,
      "epoch": 1.889638745533942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17855239808559417,
      "num_tokens": 41708355.0,
      "step": 19040
    },
    {
      "entropy": 5.614826250076294,
      "epoch": 1.8901349741961095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000464635101792902,
      "loss": 5.1848,
      "mean_token_accuracy": 0.17206190824508666,
      "num_tokens": 41720821.0,
      "step": 19045
    },
    {
      "entropy": 5.673423767089844,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.1747,
      "mean_token_accuracy": 0.16767907589673997,
      "num_tokens": 41731167.0,
      "step": 19050
    },
    {
      "entropy": 5.596800708770752,
      "epoch": 1.8911274315204447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.1287,
      "mean_token_accuracy": 0.16858217120170593,
      "num_tokens": 41742175.0,
      "step": 19055
    },
    {
      "entropy": 5.546991777420044,
      "epoch": 1.8916236601826122,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.065,
      "mean_token_accuracy": 0.17255469411611557,
      "num_tokens": 41752435.0,
      "step": 19060
    },
    {
      "entropy": 5.631256771087647,
      "epoch": 1.8921198888447797,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17698732167482376,
      "num_tokens": 41762793.0,
      "step": 19065
    },
    {
      "entropy": 5.633091926574707,
      "epoch": 1.8926161175069471,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.1468,
      "mean_token_accuracy": 0.17256583869457245,
      "num_tokens": 41774139.0,
      "step": 19070
    },
    {
      "entropy": 5.6470255851745605,
      "epoch": 1.8931123461691146,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16507964134216307,
      "num_tokens": 41784908.0,
      "step": 19075
    },
    {
      "entropy": 5.5222039222717285,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.1318,
      "mean_token_accuracy": 0.1641914278268814,
      "num_tokens": 41794742.0,
      "step": 19080
    },
    {
      "entropy": 5.637649917602539,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.1999,
      "mean_token_accuracy": 0.17672694176435472,
      "num_tokens": 41805188.0,
      "step": 19085
    },
    {
      "entropy": 5.611125183105469,
      "epoch": 1.8946010321556173,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.0875,
      "mean_token_accuracy": 0.18361698538064958,
      "num_tokens": 41814689.0,
      "step": 19090
    },
    {
      "entropy": 5.537570524215698,
      "epoch": 1.895097260817785,
      "grad_norm": 1.015625,
      "learning_rate": 0.000464444177494257,
      "loss": 5.0617,
      "mean_token_accuracy": 0.17905312031507492,
      "num_tokens": 41826623.0,
      "step": 19095
    },
    {
      "entropy": 5.635974359512329,
      "epoch": 1.8955934894799524,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.1306,
      "mean_token_accuracy": 0.17857045829296112,
      "num_tokens": 41838045.0,
      "step": 19100
    },
    {
      "entropy": 5.538073492050171,
      "epoch": 1.89608971814212,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046440593624568,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18305373638868333,
      "num_tokens": 41849905.0,
      "step": 19105
    },
    {
      "entropy": 5.618025445938111,
      "epoch": 1.8965859468042874,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.1611,
      "mean_token_accuracy": 0.16903628408908844,
      "num_tokens": 41861679.0,
      "step": 19110
    },
    {
      "entropy": 5.607078218460083,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.1674,
      "mean_token_accuracy": 0.16593395173549652,
      "num_tokens": 41873111.0,
      "step": 19115
    },
    {
      "entropy": 5.586606454849243,
      "epoch": 1.8975784041286223,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.1387,
      "mean_token_accuracy": 0.16990270614624023,
      "num_tokens": 41884181.0,
      "step": 19120
    },
    {
      "entropy": 5.621668195724487,
      "epoch": 1.89807463279079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17830565869808196,
      "num_tokens": 41895086.0,
      "step": 19125
    },
    {
      "entropy": 5.548466920852661,
      "epoch": 1.8985708614529575,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.0532,
      "mean_token_accuracy": 0.1842472642660141,
      "num_tokens": 41907476.0,
      "step": 19130
    },
    {
      "entropy": 5.620968818664551,
      "epoch": 1.8990670901151252,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.2039,
      "mean_token_accuracy": 0.169579941034317,
      "num_tokens": 41918992.0,
      "step": 19135
    },
    {
      "entropy": 5.695254802703857,
      "epoch": 1.8995633187772927,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1664196878671646,
      "num_tokens": 41929332.0,
      "step": 19140
    },
    {
      "entropy": 5.506523561477661,
      "epoch": 1.9000595474394602,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004642527834149696,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1861979231238365,
      "num_tokens": 41940363.0,
      "step": 19145
    },
    {
      "entropy": 5.553765726089478,
      "epoch": 1.9005557761016276,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.1136,
      "mean_token_accuracy": 0.1759836956858635,
      "num_tokens": 41951491.0,
      "step": 19150
    },
    {
      "entropy": 5.579548358917236,
      "epoch": 1.901052004763795,
      "grad_norm": 1.0,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.0351,
      "mean_token_accuracy": 0.1868537575006485,
      "num_tokens": 41962312.0,
      "step": 19155
    },
    {
      "entropy": 5.597741413116455,
      "epoch": 1.9015482334259626,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.1402,
      "mean_token_accuracy": 0.17225602269172668,
      "num_tokens": 41972455.0,
      "step": 19160
    },
    {
      "entropy": 5.5191113471984865,
      "epoch": 1.90204446208813,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.1738,
      "mean_token_accuracy": 0.17295060753822328,
      "num_tokens": 41984608.0,
      "step": 19165
    },
    {
      "entropy": 5.662699460983276,
      "epoch": 1.9025406907502977,
      "grad_norm": 1.03125,
      "learning_rate": 0.000464156910355857,
      "loss": 5.1766,
      "mean_token_accuracy": 0.17166121155023575,
      "num_tokens": 41996838.0,
      "step": 19170
    },
    {
      "entropy": 5.651823902130127,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17022709995508195,
      "num_tokens": 42007720.0,
      "step": 19175
    },
    {
      "entropy": 5.539491939544678,
      "epoch": 1.903533148074633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17489338368177415,
      "num_tokens": 42018859.0,
      "step": 19180
    },
    {
      "entropy": 5.576513624191284,
      "epoch": 1.9040293767368004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1764916568994522,
      "num_tokens": 42029616.0,
      "step": 19185
    },
    {
      "entropy": 5.63318772315979,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.1216,
      "mean_token_accuracy": 0.16559493392705918,
      "num_tokens": 42039870.0,
      "step": 19190
    },
    {
      "entropy": 5.678617334365844,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.1282,
      "mean_token_accuracy": 0.17640409618616104,
      "num_tokens": 42050418.0,
      "step": 19195
    },
    {
      "entropy": 5.481667947769165,
      "epoch": 1.9055180627233028,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004640417078976003,
      "loss": 4.9041,
      "mean_token_accuracy": 0.1873192921280861,
      "num_tokens": 42062273.0,
      "step": 19200
    },
    {
      "entropy": 5.5551951885223385,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.1627,
      "mean_token_accuracy": 0.1736050382256508,
      "num_tokens": 42073573.0,
      "step": 19205
    },
    {
      "entropy": 5.615964889526367,
      "epoch": 1.906510520047638,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000464003269570485,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1682993933558464,
      "num_tokens": 42085217.0,
      "step": 19210
    },
    {
      "entropy": 5.659660530090332,
      "epoch": 1.9070067487098055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.1158,
      "mean_token_accuracy": 0.17119376361370087,
      "num_tokens": 42095993.0,
      "step": 19215
    },
    {
      "entropy": 5.536816692352295,
      "epoch": 1.9075029773719732,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.0408,
      "mean_token_accuracy": 0.18006814420223236,
      "num_tokens": 42107049.0,
      "step": 19220
    },
    {
      "entropy": 5.549579668045044,
      "epoch": 1.9079992060341406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17125475853681565,
      "num_tokens": 42117101.0,
      "step": 19225
    },
    {
      "entropy": 5.539222955703735,
      "epoch": 1.9084954346963081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.097,
      "mean_token_accuracy": 0.1774652436375618,
      "num_tokens": 42128432.0,
      "step": 19230
    },
    {
      "entropy": 5.57769284248352,
      "epoch": 1.9089916633584756,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17728832066059114,
      "num_tokens": 42140379.0,
      "step": 19235
    },
    {
      "entropy": 5.607166862487793,
      "epoch": 1.909487892020643,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17306795716285706,
      "num_tokens": 42151523.0,
      "step": 19240
    },
    {
      "entropy": 5.648929405212402,
      "epoch": 1.9099841206828105,
      "grad_norm": 1.09375,
      "learning_rate": 0.000463868587814095,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1683853104710579,
      "num_tokens": 42162002.0,
      "step": 19245
    },
    {
      "entropy": 5.598668718338013,
      "epoch": 1.910480349344978,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.0346,
      "mean_token_accuracy": 0.18538910746574402,
      "num_tokens": 42172363.0,
      "step": 19250
    },
    {
      "entropy": 5.5652776718139645,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.1231,
      "mean_token_accuracy": 0.17570437043905257,
      "num_tokens": 42184162.0,
      "step": 19255
    },
    {
      "entropy": 5.633339834213257,
      "epoch": 1.9114728066693132,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.1506,
      "mean_token_accuracy": 0.1750887081027031,
      "num_tokens": 42195024.0,
      "step": 19260
    },
    {
      "entropy": 5.6283022403717045,
      "epoch": 1.9119690353314809,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17085965126752853,
      "num_tokens": 42205463.0,
      "step": 19265
    },
    {
      "entropy": 5.481038284301758,
      "epoch": 1.9124652639936484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046377224604397127,
      "loss": 4.9958,
      "mean_token_accuracy": 0.18316214978694917,
      "num_tokens": 42215133.0,
      "step": 19270
    },
    {
      "entropy": 5.617531633377075,
      "epoch": 1.9129614926558158,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.1656,
      "mean_token_accuracy": 0.17819220572710037,
      "num_tokens": 42226822.0,
      "step": 19275
    },
    {
      "entropy": 5.570143318176269,
      "epoch": 1.9134577213179833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.0934,
      "mean_token_accuracy": 0.1747848555445671,
      "num_tokens": 42237298.0,
      "step": 19280
    },
    {
      "entropy": 5.582445573806763,
      "epoch": 1.9139539499801508,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.0332,
      "mean_token_accuracy": 0.17973725199699403,
      "num_tokens": 42247242.0,
      "step": 19285
    },
    {
      "entropy": 5.544363021850586,
      "epoch": 1.9144501786423183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1855648785829544,
      "num_tokens": 42259388.0,
      "step": 19290
    },
    {
      "entropy": 5.5206711769104,
      "epoch": 1.914946407304486,
      "grad_norm": 1.0,
      "learning_rate": 0.00046367578724523,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18491165339946747,
      "num_tokens": 42270200.0,
      "step": 19295
    },
    {
      "entropy": 5.532496929168701,
      "epoch": 1.9154426359666534,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004636564814473087,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17845760136842728,
      "num_tokens": 42280723.0,
      "step": 19300
    },
    {
      "entropy": 5.585100507736206,
      "epoch": 1.9159388646288211,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.0977,
      "mean_token_accuracy": 0.17222286015748978,
      "num_tokens": 42291455.0,
      "step": 19305
    },
    {
      "entropy": 5.590035057067871,
      "epoch": 1.9164350932909886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17254936695098877,
      "num_tokens": 42302807.0,
      "step": 19310
    },
    {
      "entropy": 5.573854398727417,
      "epoch": 1.916931321953156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000463598535985815,
      "loss": 5.0597,
      "mean_token_accuracy": 0.17574475407600404,
      "num_tokens": 42313521.0,
      "step": 19315
    },
    {
      "entropy": 5.585847187042236,
      "epoch": 1.9174275506153236,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.0845,
      "mean_token_accuracy": 0.17035256922245026,
      "num_tokens": 42324666.0,
      "step": 19320
    },
    {
      "entropy": 5.548663234710693,
      "epoch": 1.917923779277491,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.1335,
      "mean_token_accuracy": 0.1762743040919304,
      "num_tokens": 42334904.0,
      "step": 19325
    },
    {
      "entropy": 5.533434057235718,
      "epoch": 1.9184200079396585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.0533,
      "mean_token_accuracy": 0.18032988011837006,
      "num_tokens": 42347164.0,
      "step": 19330
    },
    {
      "entropy": 5.5806622982025145,
      "epoch": 1.918916236601826,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17075996547937394,
      "num_tokens": 42358668.0,
      "step": 19335
    },
    {
      "entropy": 5.569100046157837,
      "epoch": 1.9194124652639937,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000463501866686305,
      "loss": 5.106,
      "mean_token_accuracy": 0.17989685535430908,
      "num_tokens": 42371229.0,
      "step": 19340
    },
    {
      "entropy": 5.5345189571380615,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17949602752923965,
      "num_tokens": 42382383.0,
      "step": 19345
    },
    {
      "entropy": 5.60338864326477,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.2178,
      "mean_token_accuracy": 0.16705350428819657,
      "num_tokens": 42393634.0,
      "step": 19350
    },
    {
      "entropy": 5.617180299758911,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17240063846111298,
      "num_tokens": 42404699.0,
      "step": 19355
    },
    {
      "entropy": 5.57647180557251,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.0461,
      "mean_token_accuracy": 0.18192058354616164,
      "num_tokens": 42415183.0,
      "step": 19360
    },
    {
      "entropy": 5.524668216705322,
      "epoch": 1.9218936085748313,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000463405080525798,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18087807595729827,
      "num_tokens": 42425162.0,
      "step": 19365
    },
    {
      "entropy": 5.522985124588013,
      "epoch": 1.9223898372369987,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17550012916326524,
      "num_tokens": 42438044.0,
      "step": 19370
    },
    {
      "entropy": 5.5598388671875,
      "epoch": 1.9228860658991662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004633663333539542,
      "loss": 5.0797,
      "mean_token_accuracy": 0.17735034674406053,
      "num_tokens": 42449214.0,
      "step": 19375
    },
    {
      "entropy": 5.6088447093963625,
      "epoch": 1.923382294561334,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.1184,
      "mean_token_accuracy": 0.17696808874607087,
      "num_tokens": 42460645.0,
      "step": 19380
    },
    {
      "entropy": 5.558672904968262,
      "epoch": 1.9238785232235014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.0714,
      "mean_token_accuracy": 0.17782512605190276,
      "num_tokens": 42471449.0,
      "step": 19385
    },
    {
      "entropy": 5.601033687591553,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1787564978003502,
      "num_tokens": 42482002.0,
      "step": 19390
    },
    {
      "entropy": 5.538745498657226,
      "epoch": 1.9248709805478366,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046328878296116196,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18875047713518142,
      "num_tokens": 42492068.0,
      "step": 19395
    },
    {
      "entropy": 5.5719951629638675,
      "epoch": 1.925367209210004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.1273,
      "mean_token_accuracy": 0.1705803856253624,
      "num_tokens": 42502766.0,
      "step": 19400
    },
    {
      "entropy": 5.5374617099761965,
      "epoch": 1.9258634378721715,
      "grad_norm": 1.0,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.079,
      "mean_token_accuracy": 0.18367877453565598,
      "num_tokens": 42513706.0,
      "step": 19405
    },
    {
      "entropy": 5.590765953063965,
      "epoch": 1.926359666534339,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.0899,
      "mean_token_accuracy": 0.17219866514205934,
      "num_tokens": 42524164.0,
      "step": 19410
    },
    {
      "entropy": 5.661651372909546,
      "epoch": 1.9268558951965065,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.1363,
      "mean_token_accuracy": 0.17114102393388747,
      "num_tokens": 42535445.0,
      "step": 19415
    },
    {
      "entropy": 5.61998724937439,
      "epoch": 1.927352123858674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.0448,
      "mean_token_accuracy": 0.16988709419965745,
      "num_tokens": 42545655.0,
      "step": 19420
    },
    {
      "entropy": 5.557079696655274,
      "epoch": 1.9278483525208416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.0452,
      "mean_token_accuracy": 0.17941776812076568,
      "num_tokens": 42556424.0,
      "step": 19425
    },
    {
      "entropy": 5.532361078262329,
      "epoch": 1.928344581183009,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.0437,
      "mean_token_accuracy": 0.1758138880133629,
      "num_tokens": 42566877.0,
      "step": 19430
    },
    {
      "entropy": 5.622207927703857,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.0735,
      "mean_token_accuracy": 0.1789550006389618,
      "num_tokens": 42577202.0,
      "step": 19435
    },
    {
      "entropy": 5.627384519577026,
      "epoch": 1.9293370385073443,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17475926280021667,
      "num_tokens": 42588605.0,
      "step": 19440
    },
    {
      "entropy": 5.563139915466309,
      "epoch": 1.9298332671695118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.0808,
      "mean_token_accuracy": 0.17682783156633378,
      "num_tokens": 42599618.0,
      "step": 19445
    },
    {
      "entropy": 5.571786260604858,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17380231320858003,
      "num_tokens": 42610999.0,
      "step": 19450
    },
    {
      "entropy": 5.537028980255127,
      "epoch": 1.9308257244938467,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17581950426101683,
      "num_tokens": 42621167.0,
      "step": 19455
    },
    {
      "entropy": 5.561048221588135,
      "epoch": 1.9313219531560142,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17391788959503174,
      "num_tokens": 42631787.0,
      "step": 19460
    },
    {
      "entropy": 5.654388809204102,
      "epoch": 1.9318181818181817,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.285,
      "mean_token_accuracy": 0.15982785671949387,
      "num_tokens": 42644002.0,
      "step": 19465
    },
    {
      "entropy": 5.582088565826416,
      "epoch": 1.9323144104803494,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.0923,
      "mean_token_accuracy": 0.16947009563446044,
      "num_tokens": 42655043.0,
      "step": 19470
    },
    {
      "entropy": 5.600674057006836,
      "epoch": 1.9328106391425168,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17386596500873566,
      "num_tokens": 42668041.0,
      "step": 19475
    },
    {
      "entropy": 5.548069858551026,
      "epoch": 1.9333068678046845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.0912,
      "mean_token_accuracy": 0.176916266977787,
      "num_tokens": 42678521.0,
      "step": 19480
    },
    {
      "entropy": 5.540034770965576,
      "epoch": 1.933803096466852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1721939891576767,
      "num_tokens": 42690465.0,
      "step": 19485
    },
    {
      "entropy": 5.5867351531982425,
      "epoch": 1.9342993251290195,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.1139,
      "mean_token_accuracy": 0.17374444007873535,
      "num_tokens": 42700712.0,
      "step": 19490
    },
    {
      "entropy": 5.6404437065124515,
      "epoch": 1.934795553791187,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.1532,
      "mean_token_accuracy": 0.1722123846411705,
      "num_tokens": 42712189.0,
      "step": 19495
    },
    {
      "entropy": 5.52061448097229,
      "epoch": 1.9352917824533544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18252720385789872,
      "num_tokens": 42723366.0,
      "step": 19500
    },
    {
      "entropy": 5.563991069793701,
      "epoch": 1.935788011115522,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.112,
      "mean_token_accuracy": 0.17479920834302903,
      "num_tokens": 42733766.0,
      "step": 19505
    },
    {
      "entropy": 5.639567279815674,
      "epoch": 1.9362842397776896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17436151802539826,
      "num_tokens": 42744642.0,
      "step": 19510
    },
    {
      "entropy": 5.54709758758545,
      "epoch": 1.936780468439857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.0463,
      "mean_token_accuracy": 0.18175441026687622,
      "num_tokens": 42755832.0,
      "step": 19515
    },
    {
      "entropy": 5.612672758102417,
      "epoch": 1.9372766971020248,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1732189252972603,
      "num_tokens": 42767606.0,
      "step": 19520
    },
    {
      "entropy": 5.664746809005737,
      "epoch": 1.9377729257641922,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.1311,
      "mean_token_accuracy": 0.174151511490345,
      "num_tokens": 42778845.0,
      "step": 19525
    },
    {
      "entropy": 5.578980159759522,
      "epoch": 1.9382691544263597,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.1009,
      "mean_token_accuracy": 0.17886477410793306,
      "num_tokens": 42789108.0,
      "step": 19530
    },
    {
      "entropy": 5.552608966827393,
      "epoch": 1.9387653830885272,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.1334,
      "mean_token_accuracy": 0.17334591150283812,
      "num_tokens": 42800483.0,
      "step": 19535
    },
    {
      "entropy": 5.567262363433838,
      "epoch": 1.9392616117506947,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046272431034363997,
      "loss": 5.0724,
      "mean_token_accuracy": 0.17770441472530366,
      "num_tokens": 42812669.0,
      "step": 19540
    },
    {
      "entropy": 5.620146894454956,
      "epoch": 1.9397578404128621,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17528989315032958,
      "num_tokens": 42822981.0,
      "step": 19545
    },
    {
      "entropy": 5.589276599884033,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.0976,
      "mean_token_accuracy": 0.1753765881061554,
      "num_tokens": 42833807.0,
      "step": 19550
    },
    {
      "entropy": 5.583070850372314,
      "epoch": 1.9407502977371973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.1358,
      "mean_token_accuracy": 0.17047125548124314,
      "num_tokens": 42844757.0,
      "step": 19555
    },
    {
      "entropy": 5.608278799057007,
      "epoch": 1.9412465263993648,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.1444,
      "mean_token_accuracy": 0.17430474609136581,
      "num_tokens": 42856346.0,
      "step": 19560
    },
    {
      "entropy": 5.6145164489746096,
      "epoch": 1.9417427550615325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17391880601644516,
      "num_tokens": 42866665.0,
      "step": 19565
    },
    {
      "entropy": 5.642249870300293,
      "epoch": 1.9422389837237,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.2289,
      "mean_token_accuracy": 0.16349378526210784,
      "num_tokens": 42879819.0,
      "step": 19570
    },
    {
      "entropy": 5.574222373962402,
      "epoch": 1.9427352123858674,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.1014,
      "mean_token_accuracy": 0.16777559369802475,
      "num_tokens": 42891924.0,
      "step": 19575
    },
    {
      "entropy": 5.589681100845337,
      "epoch": 1.943231441048035,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.1454,
      "mean_token_accuracy": 0.16906750947237015,
      "num_tokens": 42902369.0,
      "step": 19580
    },
    {
      "entropy": 5.699354124069214,
      "epoch": 1.9437276697102024,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.1738,
      "mean_token_accuracy": 0.16499748975038528,
      "num_tokens": 42913466.0,
      "step": 19585
    },
    {
      "entropy": 5.589007139205933,
      "epoch": 1.9442238983723699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.0944,
      "mean_token_accuracy": 0.176913021504879,
      "num_tokens": 42924033.0,
      "step": 19590
    },
    {
      "entropy": 5.577726364135742,
      "epoch": 1.9447201270345376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.1476,
      "mean_token_accuracy": 0.17038049846887587,
      "num_tokens": 42935746.0,
      "step": 19595
    },
    {
      "entropy": 5.560262680053711,
      "epoch": 1.945216355696705,
      "grad_norm": 0.921875,
      "learning_rate": 0.000462489589624776,
      "loss": 4.9682,
      "mean_token_accuracy": 0.1877903714776039,
      "num_tokens": 42945932.0,
      "step": 19600
    },
    {
      "entropy": 5.551028203964234,
      "epoch": 1.9457125843588727,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.0522,
      "mean_token_accuracy": 0.17614099830389024,
      "num_tokens": 42957326.0,
      "step": 19605
    },
    {
      "entropy": 5.576466178894043,
      "epoch": 1.9462088130210402,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17321777641773223,
      "num_tokens": 42968548.0,
      "step": 19610
    },
    {
      "entropy": 5.588423824310302,
      "epoch": 1.9467050416832077,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1753279983997345,
      "num_tokens": 42980627.0,
      "step": 19615
    },
    {
      "entropy": 5.554680299758911,
      "epoch": 1.9472012703453752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.0703,
      "mean_token_accuracy": 0.17789951115846633,
      "num_tokens": 42992212.0,
      "step": 19620
    },
    {
      "entropy": 5.563474988937378,
      "epoch": 1.9476974990075426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.0329,
      "mean_token_accuracy": 0.17912891656160354,
      "num_tokens": 43002833.0,
      "step": 19625
    },
    {
      "entropy": 5.555936574935913,
      "epoch": 1.94819372766971,
      "grad_norm": 1.0,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17625528424978257,
      "num_tokens": 43013861.0,
      "step": 19630
    },
    {
      "entropy": 5.551617765426636,
      "epoch": 1.9486899563318776,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.0722,
      "mean_token_accuracy": 0.17088161557912826,
      "num_tokens": 43025295.0,
      "step": 19635
    },
    {
      "entropy": 5.563519144058228,
      "epoch": 1.9491861849940453,
      "grad_norm": 1.015625,
      "learning_rate": 0.000462332737146781,
      "loss": 5.1236,
      "mean_token_accuracy": 0.17473479956388474,
      "num_tokens": 43035956.0,
      "step": 19640
    },
    {
      "entropy": 5.576566696166992,
      "epoch": 1.9496824136562128,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.1265,
      "mean_token_accuracy": 0.17735406309366225,
      "num_tokens": 43047017.0,
      "step": 19645
    },
    {
      "entropy": 5.604193925857544,
      "epoch": 1.9501786423183805,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1709732949733734,
      "num_tokens": 43057600.0,
      "step": 19650
    },
    {
      "entropy": 5.571951913833618,
      "epoch": 1.950674870980548,
      "grad_norm": 1.25,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.114,
      "mean_token_accuracy": 0.16986027210950852,
      "num_tokens": 43068689.0,
      "step": 19655
    },
    {
      "entropy": 5.5587685108184814,
      "epoch": 1.9511710996427154,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.075,
      "mean_token_accuracy": 0.17564521431922914,
      "num_tokens": 43079576.0,
      "step": 19660
    },
    {
      "entropy": 5.592776489257813,
      "epoch": 1.9516673283048829,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.101,
      "mean_token_accuracy": 0.1711607113480568,
      "num_tokens": 43090522.0,
      "step": 19665
    },
    {
      "entropy": 5.473599863052368,
      "epoch": 1.9521635569670504,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17810677736997604,
      "num_tokens": 43100043.0,
      "step": 19670
    },
    {
      "entropy": 5.565501403808594,
      "epoch": 1.9526597856292178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16827340722084044,
      "num_tokens": 43111023.0,
      "step": 19675
    },
    {
      "entropy": 5.668829822540284,
      "epoch": 1.9531560142913855,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1677851289510727,
      "num_tokens": 43123615.0,
      "step": 19680
    },
    {
      "entropy": 5.586868715286255,
      "epoch": 1.953652242953553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.1416,
      "mean_token_accuracy": 0.16694627404212953,
      "num_tokens": 43134299.0,
      "step": 19685
    },
    {
      "entropy": 5.556848621368408,
      "epoch": 1.9541484716157205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004621362534882192,
      "loss": 4.9927,
      "mean_token_accuracy": 0.1808648869395256,
      "num_tokens": 43143857.0,
      "step": 19690
    },
    {
      "entropy": 5.559931898117066,
      "epoch": 1.9546447002778882,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.1165,
      "mean_token_accuracy": 0.1693742170929909,
      "num_tokens": 43154571.0,
      "step": 19695
    },
    {
      "entropy": 5.596747207641601,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046209690105399623,
      "loss": 5.208,
      "mean_token_accuracy": 0.17275901436805724,
      "num_tokens": 43165812.0,
      "step": 19700
    },
    {
      "entropy": 5.627009010314941,
      "epoch": 1.9556371576022231,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.1337,
      "mean_token_accuracy": 0.17351840138435365,
      "num_tokens": 43178774.0,
      "step": 19705
    },
    {
      "entropy": 5.567388677597046,
      "epoch": 1.9561333862643906,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046205753006137347,
      "loss": 5.0025,
      "mean_token_accuracy": 0.17912761121988297,
      "num_tokens": 43189795.0,
      "step": 19710
    },
    {
      "entropy": 5.530686712265014,
      "epoch": 1.956629614926558,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1744740843772888,
      "num_tokens": 43201093.0,
      "step": 19715
    },
    {
      "entropy": 5.6198700904846195,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.1244,
      "mean_token_accuracy": 0.17316441833972931,
      "num_tokens": 43211370.0,
      "step": 19720
    },
    {
      "entropy": 5.485548830032348,
      "epoch": 1.9576220722508932,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004619984387839829,
      "loss": 5.0002,
      "mean_token_accuracy": 0.17983275204896926,
      "num_tokens": 43222066.0,
      "step": 19725
    },
    {
      "entropy": 5.571327066421508,
      "epoch": 1.9581183009130607,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17553064674139024,
      "num_tokens": 43233852.0,
      "step": 19730
    },
    {
      "entropy": 5.612454652786255,
      "epoch": 1.9586145295752284,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.0804,
      "mean_token_accuracy": 0.180421145260334,
      "num_tokens": 43244090.0,
      "step": 19735
    },
    {
      "entropy": 5.667940330505371,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17344639897346498,
      "num_tokens": 43255079.0,
      "step": 19740
    },
    {
      "entropy": 5.4970039367675785,
      "epoch": 1.9596069868995634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004619195854961939,
      "loss": 4.9672,
      "mean_token_accuracy": 0.1836644098162651,
      "num_tokens": 43266446.0,
      "step": 19745
    },
    {
      "entropy": 5.524586296081543,
      "epoch": 1.9601032155617308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046189986058501687,
      "loss": 5.003,
      "mean_token_accuracy": 0.17671498656272888,
      "num_tokens": 43276854.0,
      "step": 19750
    },
    {
      "entropy": 5.659414434432984,
      "epoch": 1.9605994442238983,
      "grad_norm": 0.984375,
      "learning_rate": 0.000461880131039126,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1738089516758919,
      "num_tokens": 43288515.0,
      "step": 19755
    },
    {
      "entropy": 5.557103013992309,
      "epoch": 1.9610956728860658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17217955738306046,
      "num_tokens": 43299433.0,
      "step": 19760
    },
    {
      "entropy": 5.521183109283447,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1805848315358162,
      "num_tokens": 43309799.0,
      "step": 19765
    },
    {
      "entropy": 5.585933256149292,
      "epoch": 1.962088130210401,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.1281,
      "mean_token_accuracy": 0.17101261764764786,
      "num_tokens": 43321925.0,
      "step": 19770
    },
    {
      "entropy": 5.534637022018432,
      "epoch": 1.9625843588725684,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.0599,
      "mean_token_accuracy": 0.178400519490242,
      "num_tokens": 43333336.0,
      "step": 19775
    },
    {
      "entropy": 5.574430513381958,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004617814138060926,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17958050221204758,
      "num_tokens": 43344343.0,
      "step": 19780
    },
    {
      "entropy": 5.622323894500733,
      "epoch": 1.9635768161969036,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17647516131401061,
      "num_tokens": 43355585.0,
      "step": 19785
    },
    {
      "entropy": 5.527980852127075,
      "epoch": 1.964073044859071,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17853750586509703,
      "num_tokens": 43365938.0,
      "step": 19790
    },
    {
      "entropy": 5.542626476287841,
      "epoch": 1.9645692735212386,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.046,
      "mean_token_accuracy": 0.1783442571759224,
      "num_tokens": 43377114.0,
      "step": 19795
    },
    {
      "entropy": 5.584360074996948,
      "epoch": 1.965065502183406,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.0198,
      "mean_token_accuracy": 0.1799922913312912,
      "num_tokens": 43389869.0,
      "step": 19800
    },
    {
      "entropy": 5.596322345733642,
      "epoch": 1.9655617308455735,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17245468944311143,
      "num_tokens": 43400989.0,
      "step": 19805
    },
    {
      "entropy": 5.6092535018920895,
      "epoch": 1.9660579595077412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.1828,
      "mean_token_accuracy": 0.17201780676841735,
      "num_tokens": 43411688.0,
      "step": 19810
    },
    {
      "entropy": 5.681933879852295,
      "epoch": 1.9665541881699087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.1574,
      "mean_token_accuracy": 0.1777813106775284,
      "num_tokens": 43423362.0,
      "step": 19815
    },
    {
      "entropy": 5.614991807937622,
      "epoch": 1.9670504168320764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004616232254063544,
      "loss": 5.0995,
      "mean_token_accuracy": 0.17673288583755492,
      "num_tokens": 43433651.0,
      "step": 19820
    },
    {
      "entropy": 5.498886823654175,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046160343102589384,
      "loss": 4.92,
      "mean_token_accuracy": 0.1972435086965561,
      "num_tokens": 43444048.0,
      "step": 19825
    },
    {
      "entropy": 5.58025164604187,
      "epoch": 1.9680428741564113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.1065,
      "mean_token_accuracy": 0.17573139667510987,
      "num_tokens": 43453913.0,
      "step": 19830
    },
    {
      "entropy": 5.596156167984009,
      "epoch": 1.9685391028185788,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.1599,
      "mean_token_accuracy": 0.17220928221940995,
      "num_tokens": 43466800.0,
      "step": 19835
    },
    {
      "entropy": 5.608273315429687,
      "epoch": 1.9690353314807463,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.1155,
      "mean_token_accuracy": 0.17922585010528563,
      "num_tokens": 43477435.0,
      "step": 19840
    },
    {
      "entropy": 5.582707071304322,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.083,
      "mean_token_accuracy": 0.17774913609027862,
      "num_tokens": 43486772.0,
      "step": 19845
    },
    {
      "entropy": 5.586400556564331,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.0425,
      "mean_token_accuracy": 0.17676098346710206,
      "num_tokens": 43497240.0,
      "step": 19850
    },
    {
      "entropy": 5.595136976242065,
      "epoch": 1.970524017467249,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17064619958400726,
      "num_tokens": 43507282.0,
      "step": 19855
    },
    {
      "entropy": 5.657425594329834,
      "epoch": 1.9710202461294164,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.1213,
      "mean_token_accuracy": 0.1727917343378067,
      "num_tokens": 43517782.0,
      "step": 19860
    },
    {
      "entropy": 5.578749084472657,
      "epoch": 1.971516474791584,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.1711,
      "mean_token_accuracy": 0.17165857851505278,
      "num_tokens": 43528139.0,
      "step": 19865
    },
    {
      "entropy": 5.549851131439209,
      "epoch": 1.9720127034537516,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17865224182605743,
      "num_tokens": 43540395.0,
      "step": 19870
    },
    {
      "entropy": 5.521418905258178,
      "epoch": 1.972508932115919,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.0206,
      "mean_token_accuracy": 0.1763121709227562,
      "num_tokens": 43551521.0,
      "step": 19875
    },
    {
      "entropy": 5.542658042907715,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.0504,
      "mean_token_accuracy": 0.1832459509372711,
      "num_tokens": 43562468.0,
      "step": 19880
    },
    {
      "entropy": 5.577911043167115,
      "epoch": 1.973501389440254,
      "grad_norm": 1.0,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17632562220096587,
      "num_tokens": 43573144.0,
      "step": 19885
    },
    {
      "entropy": 5.6109308242797855,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17485972344875336,
      "num_tokens": 43584482.0,
      "step": 19890
    },
    {
      "entropy": 5.615381050109863,
      "epoch": 1.9744938467645892,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.1562,
      "mean_token_accuracy": 0.17287420332431794,
      "num_tokens": 43596000.0,
      "step": 19895
    },
    {
      "entropy": 5.61561803817749,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.2492,
      "mean_token_accuracy": 0.1649932458996773,
      "num_tokens": 43606493.0,
      "step": 19900
    },
    {
      "entropy": 5.593475484848023,
      "epoch": 1.9754863040889243,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.1081,
      "mean_token_accuracy": 0.17200236320495604,
      "num_tokens": 43618178.0,
      "step": 19905
    },
    {
      "entropy": 5.612264394760132,
      "epoch": 1.9759825327510918,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.0728,
      "mean_token_accuracy": 0.178529292345047,
      "num_tokens": 43628978.0,
      "step": 19910
    },
    {
      "entropy": 5.616673326492309,
      "epoch": 1.9764787614132593,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000461246341374728,
      "loss": 5.1216,
      "mean_token_accuracy": 0.1746921256184578,
      "num_tokens": 43639283.0,
      "step": 19915
    },
    {
      "entropy": 5.5910721778869625,
      "epoch": 1.9769749900754268,
      "grad_norm": 0.890625,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.1392,
      "mean_token_accuracy": 0.1697324812412262,
      "num_tokens": 43650226.0,
      "step": 19920
    },
    {
      "entropy": 5.555079460144043,
      "epoch": 1.9774712187375942,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.0434,
      "mean_token_accuracy": 0.1696537435054779,
      "num_tokens": 43660289.0,
      "step": 19925
    },
    {
      "entropy": 5.596866941452026,
      "epoch": 1.9779674473997617,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.0858,
      "mean_token_accuracy": 0.17385668754577638,
      "num_tokens": 43671716.0,
      "step": 19930
    },
    {
      "entropy": 5.5378342628479,
      "epoch": 1.9784636760619292,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.0324,
      "mean_token_accuracy": 0.17519512325525283,
      "num_tokens": 43681906.0,
      "step": 19935
    },
    {
      "entropy": 5.531629228591919,
      "epoch": 1.9789599047240969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1721494033932686,
      "num_tokens": 43693765.0,
      "step": 19940
    },
    {
      "entropy": 5.5857789516448975,
      "epoch": 1.9794561333862644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.0492,
      "mean_token_accuracy": 0.1814531609416008,
      "num_tokens": 43703986.0,
      "step": 19945
    },
    {
      "entropy": 5.540805435180664,
      "epoch": 1.979952362048432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.0642,
      "mean_token_accuracy": 0.17312854528427124,
      "num_tokens": 43715361.0,
      "step": 19950
    },
    {
      "entropy": 5.633718776702881,
      "epoch": 1.9804485907105995,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.1081,
      "mean_token_accuracy": 0.17834949493408203,
      "num_tokens": 43725674.0,
      "step": 19955
    },
    {
      "entropy": 5.611129808425903,
      "epoch": 1.980944819372767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.08,
      "mean_token_accuracy": 0.17450591474771499,
      "num_tokens": 43736004.0,
      "step": 19960
    },
    {
      "entropy": 5.591703987121582,
      "epoch": 1.9814410480349345,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.108,
      "mean_token_accuracy": 0.1737539529800415,
      "num_tokens": 43747216.0,
      "step": 19965
    },
    {
      "entropy": 5.558017778396606,
      "epoch": 1.981937276697102,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.08,
      "mean_token_accuracy": 0.17503258138895034,
      "num_tokens": 43758865.0,
      "step": 19970
    },
    {
      "entropy": 5.544087696075439,
      "epoch": 1.9824335053592694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18277803361415862,
      "num_tokens": 43769428.0,
      "step": 19975
    },
    {
      "entropy": 5.581091070175171,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.1115,
      "mean_token_accuracy": 0.17605590373277663,
      "num_tokens": 43779334.0,
      "step": 19980
    },
    {
      "entropy": 5.628533554077149,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.1477,
      "mean_token_accuracy": 0.17384569942951203,
      "num_tokens": 43791308.0,
      "step": 19985
    },
    {
      "entropy": 5.610135316848755,
      "epoch": 1.983922191345772,
      "grad_norm": 1.0,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.105,
      "mean_token_accuracy": 0.17467275857925416,
      "num_tokens": 43801871.0,
      "step": 19990
    },
    {
      "entropy": 5.521086645126343,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046092767202867156,
      "loss": 4.9414,
      "mean_token_accuracy": 0.18358289897441865,
      "num_tokens": 43812897.0,
      "step": 19995
    },
    {
      "entropy": 5.550377082824707,
      "epoch": 1.9849146486701073,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.1587,
      "mean_token_accuracy": 0.1735536575317383,
      "num_tokens": 43823104.0,
      "step": 20000
    },
    {
      "entropy": 5.573636770248413,
      "epoch": 1.9854108773322747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.0511,
      "mean_token_accuracy": 0.18056807965040206,
      "num_tokens": 43833889.0,
      "step": 20005
    },
    {
      "entropy": 5.667809724807739,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.0,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1692133665084839,
      "num_tokens": 43845407.0,
      "step": 20010
    },
    {
      "entropy": 5.636029434204102,
      "epoch": 1.9864033346566097,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1820644408464432,
      "num_tokens": 43857304.0,
      "step": 20015
    },
    {
      "entropy": 5.610428190231323,
      "epoch": 1.9868995633187772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17815715968608856,
      "num_tokens": 43867993.0,
      "step": 20020
    },
    {
      "entropy": 5.552400255203247,
      "epoch": 1.9873957919809448,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17730575948953628,
      "num_tokens": 43877724.0,
      "step": 20025
    },
    {
      "entropy": 5.544674444198608,
      "epoch": 1.9878920206431123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004607878829166696,
      "loss": 5.014,
      "mean_token_accuracy": 0.18456718176603318,
      "num_tokens": 43887595.0,
      "step": 20030
    },
    {
      "entropy": 5.667248296737671,
      "epoch": 1.98838824930528,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17523867189884185,
      "num_tokens": 43898218.0,
      "step": 20035
    },
    {
      "entropy": 5.613915634155274,
      "epoch": 1.9888844779674475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17720010578632356,
      "num_tokens": 43908154.0,
      "step": 20040
    },
    {
      "entropy": 5.5151442527771,
      "epoch": 1.989380706629615,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17695724964141846,
      "num_tokens": 43918840.0,
      "step": 20045
    },
    {
      "entropy": 5.584105062484741,
      "epoch": 1.9898769352917824,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17090997099876404,
      "num_tokens": 43928836.0,
      "step": 20050
    },
    {
      "entropy": 5.649444532394409,
      "epoch": 1.99037316395395,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.1433,
      "mean_token_accuracy": 0.17418249398469926,
      "num_tokens": 43941213.0,
      "step": 20055
    },
    {
      "entropy": 5.603553581237793,
      "epoch": 1.9908693926161174,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004606678839929563,
      "loss": 5.1455,
      "mean_token_accuracy": 0.17339974343776704,
      "num_tokens": 43953254.0,
      "step": 20060
    },
    {
      "entropy": 5.565495920181275,
      "epoch": 1.9913656212782849,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.0764,
      "mean_token_accuracy": 0.17896127849817275,
      "num_tokens": 43963512.0,
      "step": 20065
    },
    {
      "entropy": 5.624778604507446,
      "epoch": 1.9918618499404526,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16585870683193207,
      "num_tokens": 43973735.0,
      "step": 20070
    },
    {
      "entropy": 5.632108879089356,
      "epoch": 1.99235807860262,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.1354,
      "mean_token_accuracy": 0.17154515534639359,
      "num_tokens": 43985619.0,
      "step": 20075
    },
    {
      "entropy": 5.590168523788452,
      "epoch": 1.9928543072647877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.074,
      "mean_token_accuracy": 0.16990781128406524,
      "num_tokens": 43996524.0,
      "step": 20080
    },
    {
      "entropy": 5.53729019165039,
      "epoch": 1.9933505359269552,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17434209138154982,
      "num_tokens": 44007649.0,
      "step": 20085
    },
    {
      "entropy": 5.597844028472901,
      "epoch": 1.9938467645891227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.0794,
      "mean_token_accuracy": 0.1801242247223854,
      "num_tokens": 44018049.0,
      "step": 20090
    },
    {
      "entropy": 5.619947481155395,
      "epoch": 1.9943429932512902,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.1099,
      "mean_token_accuracy": 0.1667921796441078,
      "num_tokens": 44027600.0,
      "step": 20095
    },
    {
      "entropy": 5.566410636901855,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.0627,
      "mean_token_accuracy": 0.18232190161943435,
      "num_tokens": 44038244.0,
      "step": 20100
    },
    {
      "entropy": 5.521130800247192,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046048757485989345,
      "loss": 5.0191,
      "mean_token_accuracy": 0.1823505714535713,
      "num_tokens": 44049565.0,
      "step": 20105
    },
    {
      "entropy": 5.5975524425506595,
      "epoch": 1.9958316792377928,
      "grad_norm": 1.09375,
      "learning_rate": 0.000460467517506687,
      "loss": 5.1316,
      "mean_token_accuracy": 0.17389269471168517,
      "num_tokens": 44060955.0,
      "step": 20110
    },
    {
      "entropy": 5.514666366577148,
      "epoch": 1.9963279078999603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004604474555542858,
      "loss": 4.9716,
      "mean_token_accuracy": 0.18315999060869217,
      "num_tokens": 44071115.0,
      "step": 20115
    },
    {
      "entropy": 5.644526958465576,
      "epoch": 1.996824136562128,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1743818551301956,
      "num_tokens": 44083114.0,
      "step": 20120
    },
    {
      "entropy": 5.574200582504273,
      "epoch": 1.9973203652242955,
      "grad_norm": 0.9375,
      "learning_rate": 0.000460407317853889,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17849387526512145,
      "num_tokens": 44093790.0,
      "step": 20125
    },
    {
      "entropy": 5.586059331893921,
      "epoch": 1.997816593886463,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17540828585624696,
      "num_tokens": 44104688.0,
      "step": 20130
    },
    {
      "entropy": 5.6238936424255375,
      "epoch": 1.9983128225486304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17475575506687163,
      "num_tokens": 44114487.0,
      "step": 20135
    },
    {
      "entropy": 5.5222023010253904,
      "epoch": 1.9988090512107979,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.0739,
      "mean_token_accuracy": 0.17713189572095872,
      "num_tokens": 44126070.0,
      "step": 20140
    },
    {
      "entropy": 5.536852741241455,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.076,
      "mean_token_accuracy": 0.1767069712281227,
      "num_tokens": 44137538.0,
      "step": 20145
    },
    {
      "entropy": 5.621383237838745,
      "epoch": 1.9998015085351328,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17740961015224457,
      "num_tokens": 44148322.0,
      "step": 20150
    },
    {
      "entropy": 5.688109111785889,
      "epoch": 2.0002977371973003,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.2036,
      "mean_token_accuracy": 0.17174942493438722,
      "num_tokens": 44159085.0,
      "step": 20155
    },
    {
      "entropy": 5.590953540802002,
      "epoch": 2.0007939658594682,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.0589,
      "mean_token_accuracy": 0.17163076102733613,
      "num_tokens": 44170269.0,
      "step": 20160
    },
    {
      "entropy": 5.625280046463013,
      "epoch": 2.0012901945216357,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.0075,
      "mean_token_accuracy": 0.17391431331634521,
      "num_tokens": 44181395.0,
      "step": 20165
    },
    {
      "entropy": 5.618778562545776,
      "epoch": 2.001786423183803,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17811291962862014,
      "num_tokens": 44192655.0,
      "step": 20170
    },
    {
      "entropy": 5.553651428222656,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004602063535694791,
      "loss": 4.9596,
      "mean_token_accuracy": 0.18121460527181626,
      "num_tokens": 44202628.0,
      "step": 20175
    },
    {
      "entropy": 5.632187795639038,
      "epoch": 2.002778880508138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000460186231872859,
      "loss": 5.1421,
      "mean_token_accuracy": 0.1737627148628235,
      "num_tokens": 44213281.0,
      "step": 20180
    },
    {
      "entropy": 5.562607002258301,
      "epoch": 2.0032751091703056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046016610558401936,
      "loss": 4.9898,
      "mean_token_accuracy": 0.17712361812591554,
      "num_tokens": 44223919.0,
      "step": 20185
    },
    {
      "entropy": 5.559627056121826,
      "epoch": 2.003771337832473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18738501220941545,
      "num_tokens": 44233637.0,
      "step": 20190
    },
    {
      "entropy": 5.59528169631958,
      "epoch": 2.0042675664946406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004601258392316781,
      "loss": 5.0358,
      "mean_token_accuracy": 0.182068333029747,
      "num_tokens": 44243119.0,
      "step": 20195
    },
    {
      "entropy": 5.667327356338501,
      "epoch": 2.0047637951568085,
      "grad_norm": 1.0,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.1086,
      "mean_token_accuracy": 0.17885542511940003,
      "num_tokens": 44253827.0,
      "step": 20200
    },
    {
      "entropy": 5.533254146575928,
      "epoch": 2.005260023818976,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004600855545164493,
      "loss": 4.8919,
      "mean_token_accuracy": 0.1946847915649414,
      "num_tokens": 44264721.0,
      "step": 20205
    },
    {
      "entropy": 5.520104265213012,
      "epoch": 2.0057562524811434,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004600654052740008,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18483948856592178,
      "num_tokens": 44275510.0,
      "step": 20210
    },
    {
      "entropy": 5.661779546737671,
      "epoch": 2.006252481143311,
      "grad_norm": 0.875,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.1487,
      "mean_token_accuracy": 0.16656716614961625,
      "num_tokens": 44288391.0,
      "step": 20215
    },
    {
      "entropy": 5.597352075576782,
      "epoch": 2.0067487098054784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.0967,
      "mean_token_accuracy": 0.17793567031621932,
      "num_tokens": 44300062.0,
      "step": 20220
    },
    {
      "entropy": 5.585371255874634,
      "epoch": 2.007244938467646,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046000493001331465,
      "loss": 5.0192,
      "mean_token_accuracy": 0.18095086216926576,
      "num_tokens": 44310341.0,
      "step": 20225
    },
    {
      "entropy": 5.595354270935059,
      "epoch": 2.0077411671298133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045998476241697214,
      "loss": 4.992,
      "mean_token_accuracy": 0.1848520964384079,
      "num_tokens": 44320618.0,
      "step": 20230
    },
    {
      "entropy": 5.561444330215454,
      "epoch": 2.008237395791981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004599645902334061,
      "loss": 5.0189,
      "mean_token_accuracy": 0.18483202606439592,
      "num_tokens": 44331627.0,
      "step": 20235
    },
    {
      "entropy": 5.56657772064209,
      "epoch": 2.0087336244541483,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18253988325595855,
      "num_tokens": 44342880.0,
      "step": 20240
    },
    {
      "entropy": 5.622295570373535,
      "epoch": 2.009229853116316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17752134948968887,
      "num_tokens": 44354150.0,
      "step": 20245
    },
    {
      "entropy": 5.602585697174073,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18237360268831254,
      "num_tokens": 44364849.0,
      "step": 20250
    },
    {
      "entropy": 5.567325735092163,
      "epoch": 2.010222310440651,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004598838556369131,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1714005708694458,
      "num_tokens": 44374954.0,
      "step": 20255
    },
    {
      "entropy": 5.6272056102752686,
      "epoch": 2.0107185391028186,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045986366052473525,
      "loss": 5.1055,
      "mean_token_accuracy": 0.17479193210601807,
      "num_tokens": 44387653.0,
      "step": 20260
    },
    {
      "entropy": 5.671513652801513,
      "epoch": 2.011214767764986,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004598434608283369,
      "loss": 5.0696,
      "mean_token_accuracy": 0.18372184038162231,
      "num_tokens": 44398673.0,
      "step": 20265
    },
    {
      "entropy": 5.579389142990112,
      "epoch": 2.0117109964271536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.0697,
      "mean_token_accuracy": 0.1727372959256172,
      "num_tokens": 44409764.0,
      "step": 20270
    },
    {
      "entropy": 5.626053428649902,
      "epoch": 2.012207225089321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.1123,
      "mean_token_accuracy": 0.17160431146621705,
      "num_tokens": 44421998.0,
      "step": 20275
    },
    {
      "entropy": 5.586229276657105,
      "epoch": 2.0127034537514885,
      "grad_norm": 1.0,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.0214,
      "mean_token_accuracy": 0.18545975983142854,
      "num_tokens": 44434236.0,
      "step": 20280
    },
    {
      "entropy": 5.5848963260650635,
      "epoch": 2.0131996824136564,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045976261621055915,
      "loss": 4.9689,
      "mean_token_accuracy": 0.18336813002824784,
      "num_tokens": 44444171.0,
      "step": 20285
    },
    {
      "entropy": 5.541701459884644,
      "epoch": 2.013695911075824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18495579063892365,
      "num_tokens": 44454924.0,
      "step": 20290
    },
    {
      "entropy": 5.540130376815796,
      "epoch": 2.0141921397379914,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045972216640937676,
      "loss": 5.0726,
      "mean_token_accuracy": 0.17897610664367675,
      "num_tokens": 44467072.0,
      "step": 20295
    },
    {
      "entropy": 5.629767179489136,
      "epoch": 2.014688368400159,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17290206253528595,
      "num_tokens": 44477735.0,
      "step": 20300
    },
    {
      "entropy": 5.562438011169434,
      "epoch": 2.0151845970623263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004596816982853477,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17226303070783616,
      "num_tokens": 44489028.0,
      "step": 20305
    },
    {
      "entropy": 5.623924732208252,
      "epoch": 2.015680825724494,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045966145735352,
      "loss": 5.0755,
      "mean_token_accuracy": 0.17665324062108995,
      "num_tokens": 44499313.0,
      "step": 20310
    },
    {
      "entropy": 5.568206548690796,
      "epoch": 2.0161770543866613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045964121184248604,
      "loss": 4.9597,
      "mean_token_accuracy": 0.1854465276002884,
      "num_tokens": 44510331.0,
      "step": 20315
    },
    {
      "entropy": 5.5444379329681395,
      "epoch": 2.0166732830488288,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17659149914979935,
      "num_tokens": 44521249.0,
      "step": 20320
    },
    {
      "entropy": 5.580968952178955,
      "epoch": 2.0171695117109962,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.1091,
      "mean_token_accuracy": 0.1709739089012146,
      "num_tokens": 44531117.0,
      "step": 20325
    },
    {
      "entropy": 5.614404392242432,
      "epoch": 2.017665740373164,
      "grad_norm": 0.875,
      "learning_rate": 0.00045958044783916795,
      "loss": 4.9879,
      "mean_token_accuracy": 0.1825644537806511,
      "num_tokens": 44543807.0,
      "step": 20330
    },
    {
      "entropy": 5.585230445861816,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1794258400797844,
      "num_tokens": 44553247.0,
      "step": 20335
    },
    {
      "entropy": 5.536841106414795,
      "epoch": 2.018658197697499,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.0203,
      "mean_token_accuracy": 0.17546658515930175,
      "num_tokens": 44564566.0,
      "step": 20340
    },
    {
      "entropy": 5.590663051605224,
      "epoch": 2.0191544263596666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.0589,
      "mean_token_accuracy": 0.18081258833408356,
      "num_tokens": 44575968.0,
      "step": 20345
    },
    {
      "entropy": 5.6873555183410645,
      "epoch": 2.019650655021834,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045949936508966015,
      "loss": 5.1413,
      "mean_token_accuracy": 0.1719934344291687,
      "num_tokens": 44587797.0,
      "step": 20350
    },
    {
      "entropy": 5.581501913070679,
      "epoch": 2.0201468836840015,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004594790829630596,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18321543782949448,
      "num_tokens": 44598741.0,
      "step": 20355
    },
    {
      "entropy": 5.639933919906616,
      "epoch": 2.020643112346169,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004594587962617752,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17579226493835448,
      "num_tokens": 44609813.0,
      "step": 20360
    },
    {
      "entropy": 5.561685514450073,
      "epoch": 2.0211393410083365,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004594385049863102,
      "loss": 4.9783,
      "mean_token_accuracy": 0.1807331547141075,
      "num_tokens": 44619996.0,
      "step": 20365
    },
    {
      "entropy": 5.5164858341217045,
      "epoch": 2.021635569670504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045941820913716766,
      "loss": 4.9912,
      "mean_token_accuracy": 0.17847291082143785,
      "num_tokens": 44629596.0,
      "step": 20370
    },
    {
      "entropy": 5.534903287887573,
      "epoch": 2.022131798332672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18027735203504563,
      "num_tokens": 44639204.0,
      "step": 20375
    },
    {
      "entropy": 5.595822525024414,
      "epoch": 2.0226280269948393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045937760371986334,
      "loss": 5.0676,
      "mean_token_accuracy": 0.17132706046104432,
      "num_tokens": 44650140.0,
      "step": 20380
    },
    {
      "entropy": 5.5493934631347654,
      "epoch": 2.023124255657007,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045935729415270857,
      "loss": 4.9482,
      "mean_token_accuracy": 0.184607894718647,
      "num_tokens": 44659962.0,
      "step": 20385
    },
    {
      "entropy": 5.5472358703613285,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004593369800138901,
      "loss": 4.9357,
      "mean_token_accuracy": 0.18350269943475722,
      "num_tokens": 44669608.0,
      "step": 20390
    },
    {
      "entropy": 5.504001760482788,
      "epoch": 2.0241167129813418,
      "grad_norm": 1.125,
      "learning_rate": 0.0004593166613039118,
      "loss": 4.9774,
      "mean_token_accuracy": 0.1860184520483017,
      "num_tokens": 44679044.0,
      "step": 20395
    },
    {
      "entropy": 5.63683705329895,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.1573,
      "mean_token_accuracy": 0.18134614676237107,
      "num_tokens": 44691186.0,
      "step": 20400
    },
    {
      "entropy": 5.620012807846069,
      "epoch": 2.0251091703056767,
      "grad_norm": 0.953125,
      "learning_rate": 0.000459276010172491,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1764611318707466,
      "num_tokens": 44702827.0,
      "step": 20405
    },
    {
      "entropy": 5.6459979057312015,
      "epoch": 2.025605398967844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.0351,
      "mean_token_accuracy": 0.18146812915802002,
      "num_tokens": 44713903.0,
      "step": 20410
    },
    {
      "entropy": 5.5854957580566404,
      "epoch": 2.026101627630012,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045923534076247846,
      "loss": 5.092,
      "mean_token_accuracy": 0.1759853780269623,
      "num_tokens": 44724527.0,
      "step": 20415
    },
    {
      "entropy": 5.572423934936523,
      "epoch": 2.0265978562921796,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.081,
      "mean_token_accuracy": 0.17383324652910231,
      "num_tokens": 44734092.0,
      "step": 20420
    },
    {
      "entropy": 5.547878932952881,
      "epoch": 2.027094084954347,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045919465307790817,
      "loss": 4.9696,
      "mean_token_accuracy": 0.18861816376447677,
      "num_tokens": 44745046.0,
      "step": 20425
    },
    {
      "entropy": 5.5901655673980715,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045917430238392505,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18732383400201796,
      "num_tokens": 44754603.0,
      "step": 20430
    },
    {
      "entropy": 5.58499059677124,
      "epoch": 2.028086542278682,
      "grad_norm": 1.078125,
      "learning_rate": 0.000459153947122816,
      "loss": 5.0343,
      "mean_token_accuracy": 0.1831406146287918,
      "num_tokens": 44765263.0,
      "step": 20435
    },
    {
      "entropy": 5.647555828094482,
      "epoch": 2.0285827709408495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004591335872950859,
      "loss": 5.1022,
      "mean_token_accuracy": 0.16931809931993486,
      "num_tokens": 44775861.0,
      "step": 20440
    },
    {
      "entropy": 5.664849138259887,
      "epoch": 2.029078999603017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591132229012397,
      "loss": 5.1039,
      "mean_token_accuracy": 0.17896772772073746,
      "num_tokens": 44785132.0,
      "step": 20445
    },
    {
      "entropy": 5.586290311813355,
      "epoch": 2.0295752282651844,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045909285394178217,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17767455726861953,
      "num_tokens": 44795634.0,
      "step": 20450
    },
    {
      "entropy": 5.5745679378509525,
      "epoch": 2.030071456927352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17802928239107133,
      "num_tokens": 44806315.0,
      "step": 20455
    },
    {
      "entropy": 5.5448802471160885,
      "epoch": 2.03056768558952,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004590521023280543,
      "loss": 4.9952,
      "mean_token_accuracy": 0.187605582177639,
      "num_tokens": 44816670.0,
      "step": 20460
    },
    {
      "entropy": 5.612177467346191,
      "epoch": 2.0310639142516873,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045903171967479425,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1752113252878189,
      "num_tokens": 44827292.0,
      "step": 20465
    },
    {
      "entropy": 5.594625759124756,
      "epoch": 2.031560142913855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.0583,
      "mean_token_accuracy": 0.1747766464948654,
      "num_tokens": 44838501.0,
      "step": 20470
    },
    {
      "entropy": 5.608181142807007,
      "epoch": 2.0320563715760223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004589909406780096,
      "loss": 5.0615,
      "mean_token_accuracy": 0.1755726382136345,
      "num_tokens": 44849416.0,
      "step": 20475
    },
    {
      "entropy": 5.596524477005005,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17170740365982057,
      "num_tokens": 44860365.0,
      "step": 20480
    },
    {
      "entropy": 5.54678258895874,
      "epoch": 2.033048828900357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.0415,
      "mean_token_accuracy": 0.1810987889766693,
      "num_tokens": 44870415.0,
      "step": 20485
    },
    {
      "entropy": 5.561835241317749,
      "epoch": 2.0335450575625247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045892973796475594,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18453701287508012,
      "num_tokens": 44881409.0,
      "step": 20490
    },
    {
      "entropy": 5.6445009231567385,
      "epoch": 2.034041286224692,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000458909327937541,
      "loss": 5.1365,
      "mean_token_accuracy": 0.17532226145267488,
      "num_tokens": 44893580.0,
      "step": 20495
    },
    {
      "entropy": 5.576352310180664,
      "epoch": 2.03453751488686,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045888891334977105,
      "loss": 5.0021,
      "mean_token_accuracy": 0.17982170283794402,
      "num_tokens": 44905154.0,
      "step": 20500
    },
    {
      "entropy": 5.601222991943359,
      "epoch": 2.0350337435490276,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004588684942019524,
      "loss": 5.0073,
      "mean_token_accuracy": 0.17881788313388824,
      "num_tokens": 44916796.0,
      "step": 20505
    },
    {
      "entropy": 5.609200572967529,
      "epoch": 2.035529972211195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1748254954814911,
      "num_tokens": 44929065.0,
      "step": 20510
    },
    {
      "entropy": 5.514501333236694,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.015625,
      "learning_rate": 0.000458827642228194,
      "loss": 4.9505,
      "mean_token_accuracy": 0.18427218347787858,
      "num_tokens": 44938948.0,
      "step": 20515
    },
    {
      "entropy": 5.607034730911255,
      "epoch": 2.03652242953553,
      "grad_norm": 0.875,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.064,
      "mean_token_accuracy": 0.1810918778181076,
      "num_tokens": 44949536.0,
      "step": 20520
    },
    {
      "entropy": 5.650234222412109,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.041,
      "mean_token_accuracy": 0.1803206294775009,
      "num_tokens": 44960667.0,
      "step": 20525
    },
    {
      "entropy": 5.595774698257446,
      "epoch": 2.037514886859865,
      "grad_norm": 1.0,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1807308167219162,
      "num_tokens": 44972289.0,
      "step": 20530
    },
    {
      "entropy": 5.643571519851685,
      "epoch": 2.0380111155220324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17962378412485122,
      "num_tokens": 44983028.0,
      "step": 20535
    },
    {
      "entropy": 5.522988605499267,
      "epoch": 2.0385073441842,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045872543252840236,
      "loss": 5.0124,
      "mean_token_accuracy": 0.17763127535581588,
      "num_tokens": 44994935.0,
      "step": 20540
    },
    {
      "entropy": 5.600840616226196,
      "epoch": 2.039003572846368,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.0288,
      "mean_token_accuracy": 0.17805291563272477,
      "num_tokens": 45005403.0,
      "step": 20545
    },
    {
      "entropy": 5.5372401714324955,
      "epoch": 2.0394998015085353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004586845167529727,
      "loss": 4.9869,
      "mean_token_accuracy": 0.1783970221877098,
      "num_tokens": 45016960.0,
      "step": 20550
    },
    {
      "entropy": 5.576850271224975,
      "epoch": 2.0399960301707027,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17911584973335265,
      "num_tokens": 45027820.0,
      "step": 20555
    },
    {
      "entropy": 5.587003993988037,
      "epoch": 2.04049225883287,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1734564408659935,
      "num_tokens": 45039442.0,
      "step": 20560
    },
    {
      "entropy": 5.576256179809571,
      "epoch": 2.0409884874950377,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004586231089287456,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17893983274698258,
      "num_tokens": 45050565.0,
      "step": 20565
    },
    {
      "entropy": 5.542503070831299,
      "epoch": 2.041484716157205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004586026305464115,
      "loss": 4.9656,
      "mean_token_accuracy": 0.18834819197654723,
      "num_tokens": 45060992.0,
      "step": 20570
    },
    {
      "entropy": 5.585427713394165,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.0507,
      "mean_token_accuracy": 0.1793490782380104,
      "num_tokens": 45071595.0,
      "step": 20575
    },
    {
      "entropy": 5.547119092941284,
      "epoch": 2.04247717348154,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.0142,
      "mean_token_accuracy": 0.17481336742639542,
      "num_tokens": 45082021.0,
      "step": 20580
    },
    {
      "entropy": 5.536283874511719,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004585411680837433,
      "loss": 4.9817,
      "mean_token_accuracy": 0.1792452871799469,
      "num_tokens": 45092064.0,
      "step": 20585
    },
    {
      "entropy": 5.563938570022583,
      "epoch": 2.0434696308058755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045852067149265875,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17827968150377274,
      "num_tokens": 45104389.0,
      "step": 20590
    },
    {
      "entropy": 5.63202600479126,
      "epoch": 2.043965859468043,
      "grad_norm": 1.0625,
      "learning_rate": 0.000458500170350657,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17268544733524321,
      "num_tokens": 45113627.0,
      "step": 20595
    },
    {
      "entropy": 5.599042081832886,
      "epoch": 2.0444620881302105,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1875288113951683,
      "num_tokens": 45124835.0,
      "step": 20600
    },
    {
      "entropy": 5.443780374526978,
      "epoch": 2.044958316792378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004584591544159356,
      "loss": 4.8666,
      "mean_token_accuracy": 0.18821334838867188,
      "num_tokens": 45134966.0,
      "step": 20605
    },
    {
      "entropy": 5.5386651992797855,
      "epoch": 2.0454545454545454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.0337,
      "mean_token_accuracy": 0.1775014415383339,
      "num_tokens": 45145626.0,
      "step": 20610
    },
    {
      "entropy": 5.57825288772583,
      "epoch": 2.045950774116713,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18067193478345872,
      "num_tokens": 45157054.0,
      "step": 20615
    },
    {
      "entropy": 5.536688756942749,
      "epoch": 2.0464470027788804,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.0495,
      "mean_token_accuracy": 0.17701431810855867,
      "num_tokens": 45168275.0,
      "step": 20620
    },
    {
      "entropy": 5.6110045433044435,
      "epoch": 2.046943231441048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18511123657226564,
      "num_tokens": 45178346.0,
      "step": 20625
    },
    {
      "entropy": 5.551061582565308,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004583565349736818,
      "loss": 4.9908,
      "mean_token_accuracy": 0.18501772433519365,
      "num_tokens": 45189033.0,
      "step": 20630
    },
    {
      "entropy": 5.468842601776123,
      "epoch": 2.0479356887653832,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004583359974426538,
      "loss": 4.8975,
      "mean_token_accuracy": 0.199628247320652,
      "num_tokens": 45200470.0,
      "step": 20635
    },
    {
      "entropy": 5.552168130874634,
      "epoch": 2.0484319174275507,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045831545536528806,
      "loss": 4.9831,
      "mean_token_accuracy": 0.18131881803274155,
      "num_tokens": 45211550.0,
      "step": 20640
    },
    {
      "entropy": 5.580772066116333,
      "epoch": 2.048928146089718,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045829490874209404,
      "loss": 4.99,
      "mean_token_accuracy": 0.18417651504278182,
      "num_tokens": 45222230.0,
      "step": 20645
    },
    {
      "entropy": 5.536427307128906,
      "epoch": 2.0494243747518857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.0094,
      "mean_token_accuracy": 0.182955402135849,
      "num_tokens": 45233397.0,
      "step": 20650
    },
    {
      "entropy": 5.576805496215821,
      "epoch": 2.049920603414053,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1718502536416054,
      "num_tokens": 45244617.0,
      "step": 20655
    },
    {
      "entropy": 5.590973567962647,
      "epoch": 2.0504168320762206,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004582332416026381,
      "loss": 4.8965,
      "mean_token_accuracy": 0.19161903858184814,
      "num_tokens": 45254377.0,
      "step": 20660
    },
    {
      "entropy": 5.590862274169922,
      "epoch": 2.050913060738388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.0226,
      "mean_token_accuracy": 0.1794032484292984,
      "num_tokens": 45264945.0,
      "step": 20665
    },
    {
      "entropy": 5.540956830978393,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004581921074565368,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18064985871315004,
      "num_tokens": 45275596.0,
      "step": 20670
    },
    {
      "entropy": 5.571215629577637,
      "epoch": 2.0519055180627235,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18208522498607635,
      "num_tokens": 45285775.0,
      "step": 20675
    },
    {
      "entropy": 5.636067628860474,
      "epoch": 2.052401746724891,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045815095513935747,
      "loss": 5.098,
      "mean_token_accuracy": 0.17814040631055833,
      "num_tokens": 45297760.0,
      "step": 20680
    },
    {
      "entropy": 5.670542240142822,
      "epoch": 2.0528979753870584,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004581303721678891,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1728994518518448,
      "num_tokens": 45308903.0,
      "step": 20685
    },
    {
      "entropy": 5.529480886459351,
      "epoch": 2.053394204049226,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004581097846551821,
      "loss": 4.9898,
      "mean_token_accuracy": 0.18218246400356292,
      "num_tokens": 45321004.0,
      "step": 20690
    },
    {
      "entropy": 5.554506158828735,
      "epoch": 2.0538904327113934,
      "grad_norm": 1.015625,
      "learning_rate": 0.000458089192601747,
      "loss": 4.9938,
      "mean_token_accuracy": 0.17906313836574556,
      "num_tokens": 45331353.0,
      "step": 20695
    },
    {
      "entropy": 5.595145797729492,
      "epoch": 2.054386661373561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1802888661623001,
      "num_tokens": 45342086.0,
      "step": 20700
    },
    {
      "entropy": 5.626032781600952,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004580479948747353,
      "loss": 5.042,
      "mean_token_accuracy": 0.177916219830513,
      "num_tokens": 45353018.0,
      "step": 20705
    },
    {
      "entropy": 5.571772861480713,
      "epoch": 2.055379118697896,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18098251670598983,
      "num_tokens": 45364261.0,
      "step": 20710
    },
    {
      "entropy": 5.578911972045899,
      "epoch": 2.0558753473600637,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1821468487381935,
      "num_tokens": 45374888.0,
      "step": 20715
    },
    {
      "entropy": 5.599845695495605,
      "epoch": 2.056371576022231,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045798616424152654,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17842764407396317,
      "num_tokens": 45385780.0,
      "step": 20720
    },
    {
      "entropy": 5.615222454071045,
      "epoch": 2.0568678046843987,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17822665274143218,
      "num_tokens": 45396554.0,
      "step": 20725
    },
    {
      "entropy": 5.5724396228790285,
      "epoch": 2.057364033346566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.0194,
      "mean_token_accuracy": 0.1846965253353119,
      "num_tokens": 45407258.0,
      "step": 20730
    },
    {
      "entropy": 5.595016813278198,
      "epoch": 2.0578602620087336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17222968339920045,
      "num_tokens": 45417836.0,
      "step": 20735
    },
    {
      "entropy": 5.51149845123291,
      "epoch": 2.058356490670901,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004579036598723598,
      "loss": 4.9305,
      "mean_token_accuracy": 0.1812478259205818,
      "num_tokens": 45428826.0,
      "step": 20740
    },
    {
      "entropy": 5.526144218444824,
      "epoch": 2.0588527193330686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17910019159317017,
      "num_tokens": 45441023.0,
      "step": 20745
    },
    {
      "entropy": 5.561954021453857,
      "epoch": 2.059348947995236,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004578623804720302,
      "loss": 5.0569,
      "mean_token_accuracy": 0.17459167838096618,
      "num_tokens": 45451529.0,
      "step": 20750
    },
    {
      "entropy": 5.670066881179809,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.125,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.06,
      "mean_token_accuracy": 0.17548184245824813,
      "num_tokens": 45461127.0,
      "step": 20755
    },
    {
      "entropy": 5.648116493225098,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.125,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.0928,
      "mean_token_accuracy": 0.17824302166700362,
      "num_tokens": 45471649.0,
      "step": 20760
    },
    {
      "entropy": 5.515666580200195,
      "epoch": 2.060837633981739,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004578004273633687,
      "loss": 4.9282,
      "mean_token_accuracy": 0.18106131255626678,
      "num_tokens": 45482267.0,
      "step": 20765
    },
    {
      "entropy": 5.621349048614502,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1694685474038124,
      "num_tokens": 45493935.0,
      "step": 20770
    },
    {
      "entropy": 5.611789035797119,
      "epoch": 2.061830091306074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.0506,
      "mean_token_accuracy": 0.1767025411128998,
      "num_tokens": 45505347.0,
      "step": 20775
    },
    {
      "entropy": 5.509460306167602,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004577384334571976,
      "loss": 4.9807,
      "mean_token_accuracy": 0.17989071309566498,
      "num_tokens": 45515168.0,
      "step": 20780
    },
    {
      "entropy": 5.6016334056854244,
      "epoch": 2.062822548630409,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004577177597580851,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17347152084112166,
      "num_tokens": 45525792.0,
      "step": 20785
    },
    {
      "entropy": 5.537793493270874,
      "epoch": 2.0633187772925763,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045769708152796593,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18147585093975066,
      "num_tokens": 45536704.0,
      "step": 20790
    },
    {
      "entropy": 5.640101671218872,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18547103852033614,
      "num_tokens": 45548400.0,
      "step": 20795
    },
    {
      "entropy": 5.568090391159058,
      "epoch": 2.0643112346169117,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.0431,
      "mean_token_accuracy": 0.17938539236783982,
      "num_tokens": 45559417.0,
      "step": 20800
    },
    {
      "entropy": 5.661285495758056,
      "epoch": 2.064807463279079,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045763501965669627,
      "loss": 5.1283,
      "mean_token_accuracy": 0.16357256472110748,
      "num_tokens": 45571040.0,
      "step": 20805
    },
    {
      "entropy": 5.580446243286133,
      "epoch": 2.0653036919412466,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000457614323307679,
      "loss": 5.0437,
      "mean_token_accuracy": 0.18175933361053467,
      "num_tokens": 45581180.0,
      "step": 20810
    },
    {
      "entropy": 5.5800457954406735,
      "epoch": 2.065799920603414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045759362243022,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18026960343122483,
      "num_tokens": 45592671.0,
      "step": 20815
    },
    {
      "entropy": 5.489886903762818,
      "epoch": 2.0662961492655816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004575729170248327,
      "loss": 4.984,
      "mean_token_accuracy": 0.18286297470331192,
      "num_tokens": 45605002.0,
      "step": 20820
    },
    {
      "entropy": 5.570701694488525,
      "epoch": 2.066792377927749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004575522070920305,
      "loss": 5.0543,
      "mean_token_accuracy": 0.1770429104566574,
      "num_tokens": 45615238.0,
      "step": 20825
    },
    {
      "entropy": 5.600820827484131,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17416812032461165,
      "num_tokens": 45625272.0,
      "step": 20830
    },
    {
      "entropy": 5.623308324813843,
      "epoch": 2.067784835252084,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.0243,
      "mean_token_accuracy": 0.18472252786159515,
      "num_tokens": 45638591.0,
      "step": 20835
    },
    {
      "entropy": 5.5503161430358885,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.0633,
      "mean_token_accuracy": 0.18072776049375533,
      "num_tokens": 45648802.0,
      "step": 20840
    },
    {
      "entropy": 5.611609840393067,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1717084154486656,
      "num_tokens": 45660297.0,
      "step": 20845
    },
    {
      "entropy": 5.562689876556396,
      "epoch": 2.069273521238587,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004574485895347738,
      "loss": 4.9756,
      "mean_token_accuracy": 0.1773783802986145,
      "num_tokens": 45671782.0,
      "step": 20850
    },
    {
      "entropy": 5.595685434341431,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045742785244827066,
      "loss": 5.0603,
      "mean_token_accuracy": 0.1789800316095352,
      "num_tokens": 45682018.0,
      "step": 20855
    },
    {
      "entropy": 5.541380834579468,
      "epoch": 2.070265978562922,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17742883861064912,
      "num_tokens": 45692746.0,
      "step": 20860
    },
    {
      "entropy": 5.578357172012329,
      "epoch": 2.0707622072250893,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004573863647043259,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17887574136257173,
      "num_tokens": 45702452.0,
      "step": 20865
    },
    {
      "entropy": 5.596489381790161,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004573656140479129,
      "loss": 5.0096,
      "mean_token_accuracy": 0.17478578984737397,
      "num_tokens": 45712931.0,
      "step": 20870
    },
    {
      "entropy": 5.547520732879638,
      "epoch": 2.0717546645494243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004573448588692258,
      "loss": 4.9781,
      "mean_token_accuracy": 0.1919335350394249,
      "num_tokens": 45723677.0,
      "step": 20875
    },
    {
      "entropy": 5.541289377212524,
      "epoch": 2.0722508932115917,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000457324099168779,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18495735973119737,
      "num_tokens": 45734835.0,
      "step": 20880
    },
    {
      "entropy": 5.554993534088135,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.0024,
      "mean_token_accuracy": 0.1832018941640854,
      "num_tokens": 45746900.0,
      "step": 20885
    },
    {
      "entropy": 5.641939163208008,
      "epoch": 2.073243350535927,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17736797630786896,
      "num_tokens": 45757565.0,
      "step": 20890
    },
    {
      "entropy": 5.643324518203736,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.1159,
      "mean_token_accuracy": 0.16987851709127427,
      "num_tokens": 45770468.0,
      "step": 20895
    },
    {
      "entropy": 5.5834081172943115,
      "epoch": 2.074235807860262,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.0141,
      "mean_token_accuracy": 0.1871139109134674,
      "num_tokens": 45781245.0,
      "step": 20900
    },
    {
      "entropy": 5.593252515792846,
      "epoch": 2.0747320365224295,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.0538,
      "mean_token_accuracy": 0.18454191386699675,
      "num_tokens": 45793190.0,
      "step": 20905
    },
    {
      "entropy": 5.5415246963500975,
      "epoch": 2.075228265184597,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004571994460379837,
      "loss": 4.9483,
      "mean_token_accuracy": 0.18502604067325593,
      "num_tokens": 45804268.0,
      "step": 20910
    },
    {
      "entropy": 5.541912364959717,
      "epoch": 2.0757244938467645,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1818566545844078,
      "num_tokens": 45816107.0,
      "step": 20915
    },
    {
      "entropy": 5.6492702007293705,
      "epoch": 2.076220722508932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.092,
      "mean_token_accuracy": 0.17455961406230927,
      "num_tokens": 45827919.0,
      "step": 20920
    },
    {
      "entropy": 5.634049654006958,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.0,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.1381,
      "mean_token_accuracy": 0.173937426507473,
      "num_tokens": 45839343.0,
      "step": 20925
    },
    {
      "entropy": 5.613475275039673,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.073,
      "mean_token_accuracy": 0.1724502146244049,
      "num_tokens": 45849903.0,
      "step": 20930
    },
    {
      "entropy": 5.590167474746704,
      "epoch": 2.077709408495435,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18266820460557937,
      "num_tokens": 45860115.0,
      "step": 20935
    },
    {
      "entropy": 5.614244031906128,
      "epoch": 2.0782056371576023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17775402516126632,
      "num_tokens": 45870772.0,
      "step": 20940
    },
    {
      "entropy": 5.6555849552154545,
      "epoch": 2.07870186581977,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17480594366788865,
      "num_tokens": 45882498.0,
      "step": 20945
    },
    {
      "entropy": 5.61206579208374,
      "epoch": 2.0791980944819373,
      "grad_norm": 1.0,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.055,
      "mean_token_accuracy": 0.18366037011146547,
      "num_tokens": 45893829.0,
      "step": 20950
    },
    {
      "entropy": 5.509068727493286,
      "epoch": 2.0796943231441047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18237219154834747,
      "num_tokens": 45902941.0,
      "step": 20955
    },
    {
      "entropy": 5.530039691925049,
      "epoch": 2.080190551806272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045699132942274425,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18107244819402696,
      "num_tokens": 45915024.0,
      "step": 20960
    },
    {
      "entropy": 5.62871561050415,
      "epoch": 2.0806867804684397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.0592,
      "mean_token_accuracy": 0.17330416291952133,
      "num_tokens": 45925019.0,
      "step": 20965
    },
    {
      "entropy": 5.553193235397339,
      "epoch": 2.0811830091306076,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004569496519272593,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18426849991083144,
      "num_tokens": 45936824.0,
      "step": 20970
    },
    {
      "entropy": 5.52666597366333,
      "epoch": 2.081679237792775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045692880641131945,
      "loss": 4.9677,
      "mean_token_accuracy": 0.17992454618215561,
      "num_tokens": 45947513.0,
      "step": 20975
    },
    {
      "entropy": 5.530027770996094,
      "epoch": 2.0821754664549426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004569079563839372,
      "loss": 4.933,
      "mean_token_accuracy": 0.18599275201559068,
      "num_tokens": 45957190.0,
      "step": 20980
    },
    {
      "entropy": 5.574507427215576,
      "epoch": 2.08267169511711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.0734,
      "mean_token_accuracy": 0.18213292509317397,
      "num_tokens": 45968461.0,
      "step": 20985
    },
    {
      "entropy": 5.510564184188842,
      "epoch": 2.0831679237792775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045686624279691387,
      "loss": 4.989,
      "mean_token_accuracy": 0.180607670545578,
      "num_tokens": 45978782.0,
      "step": 20990
    },
    {
      "entropy": 5.574177026748657,
      "epoch": 2.083664152441445,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004568453792383072,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18708662390708924,
      "num_tokens": 45988042.0,
      "step": 20995
    },
    {
      "entropy": 5.580483055114746,
      "epoch": 2.0841603811036125,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000456824511170327,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1738030344247818,
      "num_tokens": 46000250.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.388416813909087,
      "eval_loss": 5.183629512786865,
      "eval_mean_token_accuracy": 0.17737327473772255,
      "eval_num_tokens": 46000250.0,
      "eval_runtime": 20.1429,
      "eval_samples_per_second": 1671.51,
      "eval_steps_per_second": 208.957,
      "step": 21000
    },
    {
      "entropy": 5.6331085681915285,
      "epoch": 2.08465660976578,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004568036385934906,
      "loss": 5.0724,
      "mean_token_accuracy": 0.17614651173353196,
      "num_tokens": 46011707.0,
      "step": 21005
    },
    {
      "entropy": 5.628687334060669,
      "epoch": 2.0851528384279474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.0218,
      "mean_token_accuracy": 0.17650180906057358,
      "num_tokens": 46022673.0,
      "step": 21010
    },
    {
      "entropy": 5.663446044921875,
      "epoch": 2.0856490670901153,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.1697,
      "mean_token_accuracy": 0.16567262560129165,
      "num_tokens": 46033073.0,
      "step": 21015
    },
    {
      "entropy": 5.572248363494873,
      "epoch": 2.086145295752283,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18563256710767745,
      "num_tokens": 46043395.0,
      "step": 21020
    },
    {
      "entropy": 5.5763201236724855,
      "epoch": 2.0866415244144503,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.134,
      "mean_token_accuracy": 0.16803362518548964,
      "num_tokens": 46054808.0,
      "step": 21025
    },
    {
      "entropy": 5.629660701751709,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.0985,
      "mean_token_accuracy": 0.17439703196287154,
      "num_tokens": 46065726.0,
      "step": 21030
    },
    {
      "entropy": 5.594673776626587,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.051,
      "mean_token_accuracy": 0.1791357532143593,
      "num_tokens": 46076470.0,
      "step": 21035
    },
    {
      "entropy": 5.525838565826416,
      "epoch": 2.0881302104009527,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045665740435115815,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18153254240751265,
      "num_tokens": 46087348.0,
      "step": 21040
    },
    {
      "entropy": 5.574073696136475,
      "epoch": 2.08862643906312,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000456636495722116,
      "loss": 5.0577,
      "mean_token_accuracy": 0.17647084593772888,
      "num_tokens": 46099225.0,
      "step": 21045
    },
    {
      "entropy": 5.6643797874450685,
      "epoch": 2.0891226677252877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.1158,
      "mean_token_accuracy": 0.17712267786264418,
      "num_tokens": 46110212.0,
      "step": 21050
    },
    {
      "entropy": 5.603739070892334,
      "epoch": 2.089618896387455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.0284,
      "mean_token_accuracy": 0.1830102726817131,
      "num_tokens": 46120197.0,
      "step": 21055
    },
    {
      "entropy": 5.553951597213745,
      "epoch": 2.090115125049623,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045657374281190177,
      "loss": 5.0319,
      "mean_token_accuracy": 0.18166479617357253,
      "num_tokens": 46133043.0,
      "step": 21060
    },
    {
      "entropy": 5.597258758544922,
      "epoch": 2.0906113537117905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.0424,
      "mean_token_accuracy": 0.17711450159549713,
      "num_tokens": 46144594.0,
      "step": 21065
    },
    {
      "entropy": 5.598935985565186,
      "epoch": 2.091107582373958,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.0739,
      "mean_token_accuracy": 0.17810887843370438,
      "num_tokens": 46155400.0,
      "step": 21070
    },
    {
      "entropy": 5.574726104736328,
      "epoch": 2.0916038110361255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004565109493779502,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18048541992902756,
      "num_tokens": 46166354.0,
      "step": 21075
    },
    {
      "entropy": 5.594369888305664,
      "epoch": 2.092100039698293,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18531812131404876,
      "num_tokens": 46177472.0,
      "step": 21080
    },
    {
      "entropy": 5.576369953155518,
      "epoch": 2.0925962683604604,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1762381374835968,
      "num_tokens": 46189625.0,
      "step": 21085
    },
    {
      "entropy": 5.643133687973022,
      "epoch": 2.093092497022628,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.0356,
      "mean_token_accuracy": 0.17636981159448623,
      "num_tokens": 46201639.0,
      "step": 21090
    },
    {
      "entropy": 5.595928287506103,
      "epoch": 2.0935887256847954,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004564271617866467,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18112955391407012,
      "num_tokens": 46212248.0,
      "step": 21095
    },
    {
      "entropy": 5.627478790283203,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.074,
      "mean_token_accuracy": 0.17154680341482162,
      "num_tokens": 46223171.0,
      "step": 21100
    },
    {
      "entropy": 5.629086208343506,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.054,
      "mean_token_accuracy": 0.1751398742198944,
      "num_tokens": 46233782.0,
      "step": 21105
    },
    {
      "entropy": 5.616465759277344,
      "epoch": 2.0950774116712982,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.0976,
      "mean_token_accuracy": 0.17073156982660292,
      "num_tokens": 46245983.0,
      "step": 21110
    },
    {
      "entropy": 5.574044704437256,
      "epoch": 2.0955736403334657,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18368615955114365,
      "num_tokens": 46256855.0,
      "step": 21115
    },
    {
      "entropy": 5.577168941497803,
      "epoch": 2.096069868995633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004563223260738589,
      "loss": 4.9932,
      "mean_token_accuracy": 0.19207434803247453,
      "num_tokens": 46268103.0,
      "step": 21120
    },
    {
      "entropy": 5.619946002960205,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.0625,
      "learning_rate": 0.000456301345440012,
      "loss": 5.1197,
      "mean_token_accuracy": 0.17436318695545197,
      "num_tokens": 46278647.0,
      "step": 21125
    },
    {
      "entropy": 5.566387128829956,
      "epoch": 2.097062326319968,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1769214853644371,
      "num_tokens": 46289018.0,
      "step": 21130
    },
    {
      "entropy": 5.543808460235596,
      "epoch": 2.0975585549821356,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045625937068519075,
      "loss": 4.9939,
      "mean_token_accuracy": 0.17980262488126755,
      "num_tokens": 46300491.0,
      "step": 21135
    },
    {
      "entropy": 5.570202875137329,
      "epoch": 2.098054783644303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.0819,
      "mean_token_accuracy": 0.17575387954711913,
      "num_tokens": 46313322.0,
      "step": 21140
    },
    {
      "entropy": 5.650679445266723,
      "epoch": 2.098551012306471,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.1749,
      "mean_token_accuracy": 0.17181223928928374,
      "num_tokens": 46326581.0,
      "step": 21145
    },
    {
      "entropy": 5.628621387481689,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.140625,
      "learning_rate": 0.000456196374842947,
      "loss": 5.0351,
      "mean_token_accuracy": 0.18371749967336654,
      "num_tokens": 46337723.0,
      "step": 21150
    },
    {
      "entropy": 5.570526456832885,
      "epoch": 2.099543469630806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045617536724161203,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18395262509584426,
      "num_tokens": 46347290.0,
      "step": 21155
    },
    {
      "entropy": 5.553693342208862,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045615435514751834,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18420406877994538,
      "num_tokens": 46358040.0,
      "step": 21160
    },
    {
      "entropy": 5.569168186187744,
      "epoch": 2.100535926955141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000456133338561187,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17748283743858337,
      "num_tokens": 46369692.0,
      "step": 21165
    },
    {
      "entropy": 5.5698024272918705,
      "epoch": 2.1010321556173084,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1811805933713913,
      "num_tokens": 46381030.0,
      "step": 21170
    },
    {
      "entropy": 5.583285617828369,
      "epoch": 2.101528384279476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004560912919138962,
      "loss": 4.9772,
      "mean_token_accuracy": 0.17753509283065796,
      "num_tokens": 46391321.0,
      "step": 21175
    },
    {
      "entropy": 5.563186359405518,
      "epoch": 2.1020246129416433,
      "grad_norm": 0.890625,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1750202402472496,
      "num_tokens": 46403750.0,
      "step": 21180
    },
    {
      "entropy": 5.551441621780396,
      "epoch": 2.102520841603811,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004560492273039101,
      "loss": 4.981,
      "mean_token_accuracy": 0.17947891503572463,
      "num_tokens": 46413619.0,
      "step": 21185
    },
    {
      "entropy": 5.534088373184204,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18204739242792128,
      "num_tokens": 46423757.0,
      "step": 21190
    },
    {
      "entropy": 5.533443593978882,
      "epoch": 2.103513298928146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004560071447354013,
      "loss": 4.9711,
      "mean_token_accuracy": 0.19064454287290572,
      "num_tokens": 46433885.0,
      "step": 21195
    },
    {
      "entropy": 5.542391157150268,
      "epoch": 2.1040095275903137,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004559860967180053,
      "loss": 4.9629,
      "mean_token_accuracy": 0.1848060294985771,
      "num_tokens": 46445149.0,
      "step": 21200
    },
    {
      "entropy": 5.549589061737061,
      "epoch": 2.104505756252481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004559650442125441,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18324183374643327,
      "num_tokens": 46456417.0,
      "step": 21205
    },
    {
      "entropy": 5.526774263381958,
      "epoch": 2.1050019849146486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.0407,
      "mean_token_accuracy": 0.17967095077037812,
      "num_tokens": 46469093.0,
      "step": 21210
    },
    {
      "entropy": 5.612925815582275,
      "epoch": 2.105498213576816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17225219458341598,
      "num_tokens": 46479940.0,
      "step": 21215
    },
    {
      "entropy": 5.564649486541748,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045590185977299025,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18245019167661666,
      "num_tokens": 46490231.0,
      "step": 21220
    },
    {
      "entropy": 5.55569863319397,
      "epoch": 2.106490670901151,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004558807893204899,
      "loss": 5.0273,
      "mean_token_accuracy": 0.1754243329167366,
      "num_tokens": 46501229.0,
      "step": 21225
    },
    {
      "entropy": 5.62276873588562,
      "epoch": 2.106986899563319,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17252779752016068,
      "num_tokens": 46513657.0,
      "step": 21230
    },
    {
      "entropy": 5.589501476287841,
      "epoch": 2.1074831282254864,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18651823401451112,
      "num_tokens": 46525655.0,
      "step": 21235
    },
    {
      "entropy": 5.58717041015625,
      "epoch": 2.107979356887654,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17517500072717668,
      "num_tokens": 46538313.0,
      "step": 21240
    },
    {
      "entropy": 5.677880144119262,
      "epoch": 2.1084755855498214,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00045579646266117696,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17579368352890015,
      "num_tokens": 46551311.0,
      "step": 21245
    },
    {
      "entropy": 5.589240312576294,
      "epoch": 2.108971814211989,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045577536978663475,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18473295420408248,
      "num_tokens": 46561468.0,
      "step": 21250
    },
    {
      "entropy": 5.559021425247193,
      "epoch": 2.1094680428741563,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004557542724292529,
      "loss": 4.9647,
      "mean_token_accuracy": 0.18336673229932784,
      "num_tokens": 46572447.0,
      "step": 21255
    },
    {
      "entropy": 5.546955204010009,
      "epoch": 2.109964271536324,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004557331705895545,
      "loss": 4.9715,
      "mean_token_accuracy": 0.1853867381811142,
      "num_tokens": 46583736.0,
      "step": 21260
    },
    {
      "entropy": 5.5032905578613285,
      "epoch": 2.1104605001984913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004557120642680631,
      "loss": 4.9503,
      "mean_token_accuracy": 0.1831448808312416,
      "num_tokens": 46595182.0,
      "step": 21265
    },
    {
      "entropy": 5.600337123870849,
      "epoch": 2.1109567288606588,
      "grad_norm": 1.0,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18005235493183136,
      "num_tokens": 46606058.0,
      "step": 21270
    },
    {
      "entropy": 5.565517997741699,
      "epoch": 2.1114529575228267,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004556698381817942,
      "loss": 4.9758,
      "mean_token_accuracy": 0.18061963021755217,
      "num_tokens": 46617862.0,
      "step": 21275
    },
    {
      "entropy": 5.536686038970947,
      "epoch": 2.111949186184994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000455648718418064,
      "loss": 5.0247,
      "mean_token_accuracy": 0.17909787744283676,
      "num_tokens": 46627708.0,
      "step": 21280
    },
    {
      "entropy": 5.528559827804566,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004556275941746348,
      "loss": 4.9745,
      "mean_token_accuracy": 0.18387251496315002,
      "num_tokens": 46638263.0,
      "step": 21285
    },
    {
      "entropy": 5.645162296295166,
      "epoch": 2.112941643509329,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045560646545203044,
      "loss": 5.0507,
      "mean_token_accuracy": 0.17351210117340088,
      "num_tokens": 46648554.0,
      "step": 21290
    },
    {
      "entropy": 5.580668115615845,
      "epoch": 2.1134378721714966,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045558533225077496,
      "loss": 5.0845,
      "mean_token_accuracy": 0.18203630149364472,
      "num_tokens": 46659081.0,
      "step": 21295
    },
    {
      "entropy": 5.599754571914673,
      "epoch": 2.113934100833664,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.0704,
      "mean_token_accuracy": 0.1790687695145607,
      "num_tokens": 46670452.0,
      "step": 21300
    },
    {
      "entropy": 5.605826377868652,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.1199,
      "mean_token_accuracy": 0.17312029451131822,
      "num_tokens": 46680911.0,
      "step": 21305
    },
    {
      "entropy": 5.580679893493652,
      "epoch": 2.114926558157999,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004555219057803426,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18871504068374634,
      "num_tokens": 46692367.0,
      "step": 21310
    },
    {
      "entropy": 5.558912467956543,
      "epoch": 2.115422786820167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004555007546697242,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18644898384809494,
      "num_tokens": 46702198.0,
      "step": 21315
    },
    {
      "entropy": 5.584960079193115,
      "epoch": 2.1159190154823344,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18682713210582733,
      "num_tokens": 46714887.0,
      "step": 21320
    },
    {
      "entropy": 5.556647634506225,
      "epoch": 2.116415244144502,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.0343,
      "mean_token_accuracy": 0.17494019567966462,
      "num_tokens": 46725725.0,
      "step": 21325
    },
    {
      "entropy": 5.57613525390625,
      "epoch": 2.1169114728066694,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17991283237934114,
      "num_tokens": 46737105.0,
      "step": 21330
    },
    {
      "entropy": 5.5828619480133055,
      "epoch": 2.117407701468837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.063,
      "mean_token_accuracy": 0.17729820907115937,
      "num_tokens": 46747287.0,
      "step": 21335
    },
    {
      "entropy": 5.5602131366729735,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004553949319866791,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18351313471794128,
      "num_tokens": 46757818.0,
      "step": 21340
    },
    {
      "entropy": 5.564404582977295,
      "epoch": 2.118400158793172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045537375402775334,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18078820258378983,
      "num_tokens": 46768086.0,
      "step": 21345
    },
    {
      "entropy": 5.488570499420166,
      "epoch": 2.1188963874553393,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045535257159594724,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18829632699489593,
      "num_tokens": 46779591.0,
      "step": 21350
    },
    {
      "entropy": 5.577497863769532,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.0,
      "learning_rate": 0.000455331384691786,
      "loss": 5.0507,
      "mean_token_accuracy": 0.1823122173547745,
      "num_tokens": 46790843.0,
      "step": 21355
    },
    {
      "entropy": 5.591254377365113,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1785653978586197,
      "num_tokens": 46802046.0,
      "step": 21360
    },
    {
      "entropy": 5.534264755249024,
      "epoch": 2.120385073441842,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045528899746849965,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18397294282913207,
      "num_tokens": 46812955.0,
      "step": 21365
    },
    {
      "entropy": 5.622903299331665,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1791974797844887,
      "num_tokens": 46824206.0,
      "step": 21370
    },
    {
      "entropy": 5.544225120544434,
      "epoch": 2.121377530766177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004552465923620988,
      "loss": 4.9816,
      "mean_token_accuracy": 0.17998237609863282,
      "num_tokens": 46835352.0,
      "step": 21375
    },
    {
      "entropy": 5.582840299606323,
      "epoch": 2.1218737594283446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1839810937643051,
      "num_tokens": 46847071.0,
      "step": 21380
    },
    {
      "entropy": 5.566541624069214,
      "epoch": 2.122369988090512,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045520416937678973,
      "loss": 4.9679,
      "mean_token_accuracy": 0.1802217349410057,
      "num_tokens": 46856802.0,
      "step": 21385
    },
    {
      "entropy": 5.549986267089844,
      "epoch": 2.1228662167526795,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045518295118085955,
      "loss": 4.9574,
      "mean_token_accuracy": 0.18468967378139495,
      "num_tokens": 46867549.0,
      "step": 21390
    },
    {
      "entropy": 5.580000209808349,
      "epoch": 2.123362445414847,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.1161,
      "mean_token_accuracy": 0.17451453506946563,
      "num_tokens": 46877162.0,
      "step": 21395
    },
    {
      "entropy": 5.520930576324463,
      "epoch": 2.123858674077015,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045514050138507857,
      "loss": 4.979,
      "mean_token_accuracy": 0.18713399320840834,
      "num_tokens": 46889418.0,
      "step": 21400
    },
    {
      "entropy": 5.56641321182251,
      "epoch": 2.1243549027391824,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17733988910913467,
      "num_tokens": 46900778.0,
      "step": 21405
    },
    {
      "entropy": 5.535918188095093,
      "epoch": 2.12485113140135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004550980337209127,
      "loss": 4.9958,
      "mean_token_accuracy": 0.1872819557785988,
      "num_tokens": 46911383.0,
      "step": 21410
    },
    {
      "entropy": 5.56452522277832,
      "epoch": 2.1253473600635173,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.0662,
      "mean_token_accuracy": 0.17651623189449311,
      "num_tokens": 46922066.0,
      "step": 21415
    },
    {
      "entropy": 5.60972957611084,
      "epoch": 2.125843588725685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16915528625249862,
      "num_tokens": 46933317.0,
      "step": 21420
    },
    {
      "entropy": 5.615018129348755,
      "epoch": 2.1263398173878523,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045503429873065723,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18791090399026872,
      "num_tokens": 46945437.0,
      "step": 21425
    },
    {
      "entropy": 5.529301023483276,
      "epoch": 2.1268360460500197,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004550130448042775,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1843731001019478,
      "num_tokens": 46957455.0,
      "step": 21430
    },
    {
      "entropy": 5.600944471359253,
      "epoch": 2.127332274712187,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1752164751291275,
      "num_tokens": 46967626.0,
      "step": 21435
    },
    {
      "entropy": 5.568463945388794,
      "epoch": 2.1278285033743547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.0061,
      "mean_token_accuracy": 0.180928510427475,
      "num_tokens": 46979492.0,
      "step": 21440
    },
    {
      "entropy": 5.571028327941894,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18590018898248672,
      "num_tokens": 46990803.0,
      "step": 21445
    },
    {
      "entropy": 5.564924049377441,
      "epoch": 2.12882096069869,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045492798446467465,
      "loss": 5.026,
      "mean_token_accuracy": 0.1869608834385872,
      "num_tokens": 47000771.0,
      "step": 21450
    },
    {
      "entropy": 5.55486216545105,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045490670822388957,
      "loss": 5.0357,
      "mean_token_accuracy": 0.18229570388793945,
      "num_tokens": 47011461.0,
      "step": 21455
    },
    {
      "entropy": 5.477197742462158,
      "epoch": 2.129813418023025,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004548854275218059,
      "loss": 4.8696,
      "mean_token_accuracy": 0.1931580513715744,
      "num_tokens": 47021164.0,
      "step": 21460
    },
    {
      "entropy": 5.570854663848877,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17669197022914887,
      "num_tokens": 47032150.0,
      "step": 21465
    },
    {
      "entropy": 5.569223213195801,
      "epoch": 2.13080587534736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18006097674369811,
      "num_tokens": 47042853.0,
      "step": 21470
    },
    {
      "entropy": 5.574472856521607,
      "epoch": 2.1313021040095275,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004548215586530401,
      "loss": 4.946,
      "mean_token_accuracy": 0.18208712339401245,
      "num_tokens": 47053197.0,
      "step": 21475
    },
    {
      "entropy": 5.593827152252198,
      "epoch": 2.131798332671695,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045480026011103975,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17827381044626237,
      "num_tokens": 47063234.0,
      "step": 21480
    },
    {
      "entropy": 5.570671463012696,
      "epoch": 2.132294561333863,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045477895711038045,
      "loss": 4.9983,
      "mean_token_accuracy": 0.1831948474049568,
      "num_tokens": 47074291.0,
      "step": 21485
    },
    {
      "entropy": 5.541709613800049,
      "epoch": 2.1327907899960303,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.0153,
      "mean_token_accuracy": 0.17766523063182832,
      "num_tokens": 47084647.0,
      "step": 21490
    },
    {
      "entropy": 5.55436692237854,
      "epoch": 2.133287018658198,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18301872313022613,
      "num_tokens": 47096590.0,
      "step": 21495
    },
    {
      "entropy": 5.641123914718628,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.0888,
      "mean_token_accuracy": 0.1762508660554886,
      "num_tokens": 47107740.0,
      "step": 21500
    },
    {
      "entropy": 5.635685777664184,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004546937005317201,
      "loss": 4.9926,
      "mean_token_accuracy": 0.1841651439666748,
      "num_tokens": 47117665.0,
      "step": 21505
    },
    {
      "entropy": 5.684440183639526,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.1241,
      "mean_token_accuracy": 0.16997685432434081,
      "num_tokens": 47129018.0,
      "step": 21510
    },
    {
      "entropy": 5.587864208221435,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045465104550417604,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18730106800794602,
      "num_tokens": 47139882.0,
      "step": 21515
    },
    {
      "entropy": 5.551690340042114,
      "epoch": 2.135768161969035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045462971130770123,
      "loss": 4.9872,
      "mean_token_accuracy": 0.18111253678798675,
      "num_tokens": 47149294.0,
      "step": 21520
    },
    {
      "entropy": 5.681732177734375,
      "epoch": 2.1362643906312027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004546083726567968,
      "loss": 5.2396,
      "mean_token_accuracy": 0.1688072070479393,
      "num_tokens": 47161604.0,
      "step": 21525
    },
    {
      "entropy": 5.642612934112549,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.0786,
      "mean_token_accuracy": 0.18277953267097474,
      "num_tokens": 47171730.0,
      "step": 21530
    },
    {
      "entropy": 5.502249479293823,
      "epoch": 2.137256847955538,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045456568199381524,
      "loss": 4.9482,
      "mean_token_accuracy": 0.18713319748640062,
      "num_tokens": 47181167.0,
      "step": 21535
    },
    {
      "entropy": 5.574695014953614,
      "epoch": 2.1377530766177055,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.0738,
      "mean_token_accuracy": 0.18218696415424346,
      "num_tokens": 47192904.0,
      "step": 21540
    },
    {
      "entropy": 5.6826389789581295,
      "epoch": 2.138249305279873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.1336,
      "mean_token_accuracy": 0.17276070415973663,
      "num_tokens": 47203653.0,
      "step": 21545
    },
    {
      "entropy": 5.588252925872803,
      "epoch": 2.1387455339420405,
      "grad_norm": 1.0,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18849244564771653,
      "num_tokens": 47215835.0,
      "step": 21550
    },
    {
      "entropy": 5.58931975364685,
      "epoch": 2.139241762604208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.0999,
      "mean_token_accuracy": 0.1767088443040848,
      "num_tokens": 47227588.0,
      "step": 21555
    },
    {
      "entropy": 5.583798408508301,
      "epoch": 2.1397379912663754,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.0765,
      "mean_token_accuracy": 0.17306662648916243,
      "num_tokens": 47239119.0,
      "step": 21560
    },
    {
      "entropy": 5.601566743850708,
      "epoch": 2.140234219928543,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.0449,
      "mean_token_accuracy": 0.18937018662691116,
      "num_tokens": 47251047.0,
      "step": 21565
    },
    {
      "entropy": 5.63416428565979,
      "epoch": 2.140730448590711,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.1026,
      "mean_token_accuracy": 0.177091021835804,
      "num_tokens": 47260811.0,
      "step": 21570
    },
    {
      "entropy": 5.608564996719361,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004543947412705826,
      "loss": 4.978,
      "mean_token_accuracy": 0.18115118891000748,
      "num_tokens": 47271175.0,
      "step": 21575
    },
    {
      "entropy": 5.521393966674805,
      "epoch": 2.1417229059150458,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045437335365589905,
      "loss": 4.9163,
      "mean_token_accuracy": 0.18423741459846496,
      "num_tokens": 47282191.0,
      "step": 21580
    },
    {
      "entropy": 5.484629678726196,
      "epoch": 2.1422191345772132,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00045435196159314276,
      "loss": 4.9394,
      "mean_token_accuracy": 0.18814811259508132,
      "num_tokens": 47292222.0,
      "step": 21585
    },
    {
      "entropy": 5.542283964157105,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17631214410066604,
      "num_tokens": 47302936.0,
      "step": 21590
    },
    {
      "entropy": 5.605771732330322,
      "epoch": 2.143211591901548,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004543091641255342,
      "loss": 4.9632,
      "mean_token_accuracy": 0.18619616478681564,
      "num_tokens": 47314126.0,
      "step": 21595
    },
    {
      "entropy": 5.583200120925904,
      "epoch": 2.1437078205637157,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004542877587217432,
      "loss": 4.964,
      "mean_token_accuracy": 0.1847236841917038,
      "num_tokens": 47324038.0,
      "step": 21600
    },
    {
      "entropy": 5.4859599590301515,
      "epoch": 2.144204049225883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045426634887200224,
      "loss": 4.9841,
      "mean_token_accuracy": 0.1838211491703987,
      "num_tokens": 47334820.0,
      "step": 21605
    },
    {
      "entropy": 5.51772108078003,
      "epoch": 2.1447002778880506,
      "grad_norm": 1.109375,
      "learning_rate": 0.000454244934576842,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17926013320684434,
      "num_tokens": 47345039.0,
      "step": 21610
    },
    {
      "entropy": 5.629975175857544,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1784510374069214,
      "num_tokens": 47354668.0,
      "step": 21615
    },
    {
      "entropy": 5.628005599975586,
      "epoch": 2.145692735212386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.0558,
      "mean_token_accuracy": 0.17981281727552414,
      "num_tokens": 47366002.0,
      "step": 21620
    },
    {
      "entropy": 5.583718967437744,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.0389,
      "mean_token_accuracy": 0.18171764314174652,
      "num_tokens": 47377052.0,
      "step": 21625
    },
    {
      "entropy": 5.565111351013184,
      "epoch": 2.146685192536721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1763979598879814,
      "num_tokens": 47387598.0,
      "step": 21630
    },
    {
      "entropy": 5.560182428359985,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.0387,
      "mean_token_accuracy": 0.17998067289590836,
      "num_tokens": 47398565.0,
      "step": 21635
    },
    {
      "entropy": 5.590123462677002,
      "epoch": 2.147677649861056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.0085,
      "mean_token_accuracy": 0.17949577867984773,
      "num_tokens": 47408715.0,
      "step": 21640
    },
    {
      "entropy": 5.666389608383179,
      "epoch": 2.1481738785232234,
      "grad_norm": 1.0,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.108,
      "mean_token_accuracy": 0.17688941061496735,
      "num_tokens": 47418690.0,
      "step": 21645
    },
    {
      "entropy": 5.579273986816406,
      "epoch": 2.148670107185391,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.0468,
      "mean_token_accuracy": 0.17626574784517288,
      "num_tokens": 47429415.0,
      "step": 21650
    },
    {
      "entropy": 5.565044021606445,
      "epoch": 2.149166335847559,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004540520059641937,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18150714337825774,
      "num_tokens": 47440852.0,
      "step": 21655
    },
    {
      "entropy": 5.546222782135009,
      "epoch": 2.1496625645097263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004540305472440666,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18244267255067825,
      "num_tokens": 47450800.0,
      "step": 21660
    },
    {
      "entropy": 5.623450565338135,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17845799177885055,
      "num_tokens": 47460900.0,
      "step": 21665
    },
    {
      "entropy": 5.5883355140686035,
      "epoch": 2.150655021834061,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004539876164856292,
      "loss": 5.0274,
      "mean_token_accuracy": 0.17965126037597656,
      "num_tokens": 47471182.0,
      "step": 21670
    },
    {
      "entropy": 5.546105909347534,
      "epoch": 2.1511512504962287,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004539661444483835,
      "loss": 4.977,
      "mean_token_accuracy": 0.18448082208633423,
      "num_tokens": 47482348.0,
      "step": 21675
    },
    {
      "entropy": 5.643081378936768,
      "epoch": 2.151647479158396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.0653,
      "mean_token_accuracy": 0.18210738003253937,
      "num_tokens": 47494642.0,
      "step": 21680
    },
    {
      "entropy": 5.615840435028076,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539231870605005,
      "loss": 4.9633,
      "mean_token_accuracy": 0.18270669728517533,
      "num_tokens": 47504965.0,
      "step": 21685
    },
    {
      "entropy": 5.585027408599854,
      "epoch": 2.152639936482731,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.0415,
      "mean_token_accuracy": 0.18191924840211868,
      "num_tokens": 47516071.0,
      "step": 21690
    },
    {
      "entropy": 5.628939533233643,
      "epoch": 2.1531361651448986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.1147,
      "mean_token_accuracy": 0.1680321991443634,
      "num_tokens": 47527765.0,
      "step": 21695
    },
    {
      "entropy": 5.5559446811676025,
      "epoch": 2.1536323938070665,
      "grad_norm": 0.96875,
      "learning_rate": 0.000453858717703187,
      "loss": 4.9812,
      "mean_token_accuracy": 0.17975877076387406,
      "num_tokens": 47538526.0,
      "step": 21700
    },
    {
      "entropy": 5.657262992858887,
      "epoch": 2.154128622469234,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.1242,
      "mean_token_accuracy": 0.17412192076444627,
      "num_tokens": 47550061.0,
      "step": 21705
    },
    {
      "entropy": 5.5452313899993895,
      "epoch": 2.1546248511314015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004538157159542025,
      "loss": 4.9686,
      "mean_token_accuracy": 0.18930543214082718,
      "num_tokens": 47559955.0,
      "step": 21710
    },
    {
      "entropy": 5.608613681793213,
      "epoch": 2.155121079793569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004537942084280771,
      "loss": 5.0479,
      "mean_token_accuracy": 0.1750270828604698,
      "num_tokens": 47570768.0,
      "step": 21715
    },
    {
      "entropy": 5.558120059967041,
      "epoch": 2.1556173084557364,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045377269646824036,
      "loss": 4.9686,
      "mean_token_accuracy": 0.17997040897607802,
      "num_tokens": 47581296.0,
      "step": 21720
    },
    {
      "entropy": 5.615666723251342,
      "epoch": 2.156113537117904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.0969,
      "mean_token_accuracy": 0.17480258792638778,
      "num_tokens": 47591481.0,
      "step": 21725
    },
    {
      "entropy": 5.563283061981201,
      "epoch": 2.1566097657800714,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004537296592495678,
      "loss": 5.0571,
      "mean_token_accuracy": 0.18116059601306916,
      "num_tokens": 47601920.0,
      "step": 21730
    },
    {
      "entropy": 5.53978967666626,
      "epoch": 2.157105994442239,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000453708133991799,
      "loss": 4.947,
      "mean_token_accuracy": 0.18790972530841826,
      "num_tokens": 47612454.0,
      "step": 21735
    },
    {
      "entropy": 5.557951211929321,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.074,
      "mean_token_accuracy": 0.17484133392572404,
      "num_tokens": 47624134.0,
      "step": 21740
    },
    {
      "entropy": 5.5661884307861325,
      "epoch": 2.158098451766574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18084953278303145,
      "num_tokens": 47635789.0,
      "step": 21745
    },
    {
      "entropy": 5.563689422607422,
      "epoch": 2.1585946804287417,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004536435316311685,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18461095690727233,
      "num_tokens": 47645738.0,
      "step": 21750
    },
    {
      "entropy": 5.601718282699585,
      "epoch": 2.159090909090909,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.159,
      "mean_token_accuracy": 0.17255968004465103,
      "num_tokens": 47657027.0,
      "step": 21755
    },
    {
      "entropy": 5.583934926986695,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.0375,
      "mean_token_accuracy": 0.1857856273651123,
      "num_tokens": 47667257.0,
      "step": 21760
    },
    {
      "entropy": 5.561339044570923,
      "epoch": 2.160083366415244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.0415,
      "mean_token_accuracy": 0.17930829674005508,
      "num_tokens": 47677247.0,
      "step": 21765
    },
    {
      "entropy": 5.611834239959717,
      "epoch": 2.1605795950774116,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004535573331331771,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17423049062490464,
      "num_tokens": 47688751.0,
      "step": 21770
    },
    {
      "entropy": 5.599397945404053,
      "epoch": 2.161075823739579,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17797626405954362,
      "num_tokens": 47700347.0,
      "step": 21775
    },
    {
      "entropy": 5.5775465965271,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004535142073150211,
      "loss": 5.031,
      "mean_token_accuracy": 0.1859830617904663,
      "num_tokens": 47710187.0,
      "step": 21780
    },
    {
      "entropy": 5.629618453979492,
      "epoch": 2.162068281063914,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004534926377655241,
      "loss": 5.0755,
      "mean_token_accuracy": 0.1793506845831871,
      "num_tokens": 47720461.0,
      "step": 21785
    },
    {
      "entropy": 5.6041899681091305,
      "epoch": 2.162564509726082,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045347106378979435,
      "loss": 4.9821,
      "mean_token_accuracy": 0.18743932545185088,
      "num_tokens": 47730457.0,
      "step": 21790
    },
    {
      "entropy": 5.591742706298828,
      "epoch": 2.1630607383882494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18068692833185196,
      "num_tokens": 47742390.0,
      "step": 21795
    },
    {
      "entropy": 5.593025732040405,
      "epoch": 2.163556967050417,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.1444,
      "mean_token_accuracy": 0.17076237350702286,
      "num_tokens": 47752783.0,
      "step": 21800
    },
    {
      "entropy": 5.544629859924316,
      "epoch": 2.1640531957125844,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.0213,
      "mean_token_accuracy": 0.1894751861691475,
      "num_tokens": 47762997.0,
      "step": 21805
    },
    {
      "entropy": 5.577791213989258,
      "epoch": 2.164549424374752,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004533847236352487,
      "loss": 5.0912,
      "mean_token_accuracy": 0.1765584245324135,
      "num_tokens": 47773152.0,
      "step": 21810
    },
    {
      "entropy": 5.560214185714722,
      "epoch": 2.1650456530369193,
      "grad_norm": 1.03125,
      "learning_rate": 0.000453363127536382,
      "loss": 4.9881,
      "mean_token_accuracy": 0.1857873886823654,
      "num_tokens": 47784392.0,
      "step": 21815
    },
    {
      "entropy": 5.563153028488159,
      "epoch": 2.165541881699087,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045334152701449407,
      "loss": 4.9685,
      "mean_token_accuracy": 0.18468028157949448,
      "num_tokens": 47795775.0,
      "step": 21820
    },
    {
      "entropy": 5.657825040817261,
      "epoch": 2.1660381103612543,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.1689,
      "mean_token_accuracy": 0.17426373809576035,
      "num_tokens": 47808340.0,
      "step": 21825
    },
    {
      "entropy": 5.584088516235352,
      "epoch": 2.166534339023422,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17665714025497437,
      "num_tokens": 47819591.0,
      "step": 21830
    },
    {
      "entropy": 5.587255382537842,
      "epoch": 2.1670305676855897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.069,
      "mean_token_accuracy": 0.17484642714262008,
      "num_tokens": 47830551.0,
      "step": 21835
    },
    {
      "entropy": 5.596146965026856,
      "epoch": 2.167526796347757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.0304,
      "mean_token_accuracy": 0.17648040056228637,
      "num_tokens": 47839553.0,
      "step": 21840
    },
    {
      "entropy": 5.553736829757691,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.0347,
      "mean_token_accuracy": 0.17848556637763976,
      "num_tokens": 47850449.0,
      "step": 21845
    },
    {
      "entropy": 5.599539136886596,
      "epoch": 2.168519253672092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.0549,
      "mean_token_accuracy": 0.176257486641407,
      "num_tokens": 47861751.0,
      "step": 21850
    },
    {
      "entropy": 5.554778289794922,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004531901995616921,
      "loss": 4.9574,
      "mean_token_accuracy": 0.18373635709285735,
      "num_tokens": 47872019.0,
      "step": 21855
    },
    {
      "entropy": 5.591323518753052,
      "epoch": 2.169511710996427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.0349,
      "mean_token_accuracy": 0.17383376657962799,
      "num_tokens": 47882271.0,
      "step": 21860
    },
    {
      "entropy": 5.574695444107055,
      "epoch": 2.1700079396585945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17775145620107652,
      "num_tokens": 47892036.0,
      "step": 21865
    },
    {
      "entropy": 5.589202499389648,
      "epoch": 2.170504168320762,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1781063750386238,
      "num_tokens": 47902760.0,
      "step": 21870
    },
    {
      "entropy": 5.592151641845703,
      "epoch": 2.17100039698293,
      "grad_norm": 1.0,
      "learning_rate": 0.0004531036295075989,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17894388288259505,
      "num_tokens": 47913954.0,
      "step": 21875
    },
    {
      "entropy": 5.612679529190063,
      "epoch": 2.1714966256450974,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045308197595126797,
      "loss": 5.053,
      "mean_token_accuracy": 0.17946455925703048,
      "num_tokens": 47925167.0,
      "step": 21880
    },
    {
      "entropy": 5.5864228248596195,
      "epoch": 2.171992854307265,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004530603179788879,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18453045934438705,
      "num_tokens": 47934828.0,
      "step": 21885
    },
    {
      "entropy": 5.5671101093292235,
      "epoch": 2.1724890829694323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.0258,
      "mean_token_accuracy": 0.17946331053972245,
      "num_tokens": 47945716.0,
      "step": 21890
    },
    {
      "entropy": 5.644271564483643,
      "epoch": 2.1729853116316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.1268,
      "mean_token_accuracy": 0.17631648033857344,
      "num_tokens": 47956653.0,
      "step": 21895
    },
    {
      "entropy": 5.522090721130371,
      "epoch": 2.1734815402937673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004529953175708231,
      "loss": 4.8702,
      "mean_token_accuracy": 0.19161672443151473,
      "num_tokens": 47967404.0,
      "step": 21900
    },
    {
      "entropy": 5.5518683910369875,
      "epoch": 2.1739777689559348,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045297364193961763,
      "loss": 4.9422,
      "mean_token_accuracy": 0.18889790624380112,
      "num_tokens": 47977281.0,
      "step": 21905
    },
    {
      "entropy": 5.583648300170898,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1790000393986702,
      "num_tokens": 47988009.0,
      "step": 21910
    },
    {
      "entropy": 5.576822280883789,
      "epoch": 2.17497022628027,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045293027743765595,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1749187722802162,
      "num_tokens": 47999403.0,
      "step": 21915
    },
    {
      "entropy": 5.600507545471191,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.0,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.0577,
      "mean_token_accuracy": 0.18203659653663634,
      "num_tokens": 48011377.0,
      "step": 21920
    },
    {
      "entropy": 5.582033681869507,
      "epoch": 2.175962683604605,
      "grad_norm": 1.0,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.0302,
      "mean_token_accuracy": 0.17665254175662995,
      "num_tokens": 48023425.0,
      "step": 21925
    },
    {
      "entropy": 5.639842081069946,
      "epoch": 2.1764589122667726,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.1348,
      "mean_token_accuracy": 0.17536880671977997,
      "num_tokens": 48035743.0,
      "step": 21930
    },
    {
      "entropy": 5.582977819442749,
      "epoch": 2.17695514092894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004528434954905863,
      "loss": 4.9046,
      "mean_token_accuracy": 0.1872315913438797,
      "num_tokens": 48045799.0,
      "step": 21935
    },
    {
      "entropy": 5.572813177108765,
      "epoch": 2.1774513695911075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.0825,
      "mean_token_accuracy": 0.17682400345802307,
      "num_tokens": 48056514.0,
      "step": 21940
    },
    {
      "entropy": 5.53058443069458,
      "epoch": 2.177947598253275,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1767354890704155,
      "num_tokens": 48067493.0,
      "step": 21945
    },
    {
      "entropy": 5.697188472747802,
      "epoch": 2.1784438269154425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.1633,
      "mean_token_accuracy": 0.17621775418519975,
      "num_tokens": 48078593.0,
      "step": 21950
    },
    {
      "entropy": 5.592310094833374,
      "epoch": 2.17894005557761,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18119787126779557,
      "num_tokens": 48089840.0,
      "step": 21955
    },
    {
      "entropy": 5.554585647583008,
      "epoch": 2.179436284239778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.0248,
      "mean_token_accuracy": 0.1782837152481079,
      "num_tokens": 48100965.0,
      "step": 21960
    },
    {
      "entropy": 5.559949922561645,
      "epoch": 2.1799325129019453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.0299,
      "mean_token_accuracy": 0.1747749075293541,
      "num_tokens": 48111391.0,
      "step": 21965
    },
    {
      "entropy": 5.56065034866333,
      "epoch": 2.180428741564113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.051,
      "mean_token_accuracy": 0.18067461997270584,
      "num_tokens": 48122213.0,
      "step": 21970
    },
    {
      "entropy": 5.633478450775146,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.094,
      "mean_token_accuracy": 0.1780737206339836,
      "num_tokens": 48132966.0,
      "step": 21975
    },
    {
      "entropy": 5.5715484619140625,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000452647978169263,
      "loss": 5.0089,
      "mean_token_accuracy": 0.17728296518325806,
      "num_tokens": 48144253.0,
      "step": 21980
    },
    {
      "entropy": 5.570482349395752,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045262623198882797,
      "loss": 5.0118,
      "mean_token_accuracy": 0.17844044417142868,
      "num_tokens": 48154512.0,
      "step": 21985
    },
    {
      "entropy": 5.551167726516724,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.0444,
      "mean_token_accuracy": 0.18118836730718613,
      "num_tokens": 48165296.0,
      "step": 21990
    },
    {
      "entropy": 5.5991551876068115,
      "epoch": 2.18290988487495,
      "grad_norm": 1.0,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.0804,
      "mean_token_accuracy": 0.1742137849330902,
      "num_tokens": 48175878.0,
      "step": 21995
    },
    {
      "entropy": 5.592577219009399,
      "epoch": 2.183406113537118,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.0651,
      "mean_token_accuracy": 0.17924188524484636,
      "num_tokens": 48187400.0,
      "step": 22000
    },
    {
      "entropy": 5.674735975265503,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.1929,
      "mean_token_accuracy": 0.1757264405488968,
      "num_tokens": 48196956.0,
      "step": 22005
    },
    {
      "entropy": 5.577967691421509,
      "epoch": 2.184398570861453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.0398,
      "mean_token_accuracy": 0.182955402135849,
      "num_tokens": 48208947.0,
      "step": 22010
    },
    {
      "entropy": 5.654741430282593,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.1338,
      "mean_token_accuracy": 0.17420834302902222,
      "num_tokens": 48220370.0,
      "step": 22015
    },
    {
      "entropy": 5.585988759994507,
      "epoch": 2.185391028185788,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.0424,
      "mean_token_accuracy": 0.1743196040391922,
      "num_tokens": 48231977.0,
      "step": 22020
    },
    {
      "entropy": 5.554697036743164,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1739214166998863,
      "num_tokens": 48242791.0,
      "step": 22025
    },
    {
      "entropy": 5.535006475448609,
      "epoch": 2.186383485510123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18582326918840408,
      "num_tokens": 48253683.0,
      "step": 22030
    },
    {
      "entropy": 5.625677013397217,
      "epoch": 2.1868797141722904,
      "grad_norm": 1.0,
      "learning_rate": 0.0004524085280123795,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17416258454322814,
      "num_tokens": 48264681.0,
      "step": 22035
    },
    {
      "entropy": 5.613886308670044,
      "epoch": 2.187375942834458,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1837730422616005,
      "num_tokens": 48275766.0,
      "step": 22040
    },
    {
      "entropy": 5.631250715255737,
      "epoch": 2.187872171496626,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004523649344076154,
      "loss": 5.1862,
      "mean_token_accuracy": 0.17449237853288652,
      "num_tokens": 48288273.0,
      "step": 22045
    },
    {
      "entropy": 5.538258981704712,
      "epoch": 2.1883684001587933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045234313100756077,
      "loss": 4.9412,
      "mean_token_accuracy": 0.19809087812900544,
      "num_tokens": 48299668.0,
      "step": 22050
    },
    {
      "entropy": 5.597110462188721,
      "epoch": 2.188864628820961,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045232132320977854,
      "loss": 5.0126,
      "mean_token_accuracy": 0.18537514209747313,
      "num_tokens": 48311296.0,
      "step": 22055
    },
    {
      "entropy": 5.550709342956543,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004522995110148097,
      "loss": 4.9985,
      "mean_token_accuracy": 0.18255802541971206,
      "num_tokens": 48321549.0,
      "step": 22060
    },
    {
      "entropy": 5.612544870376587,
      "epoch": 2.1898570861452957,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000452277694423195,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1765699341893196,
      "num_tokens": 48334391.0,
      "step": 22065
    },
    {
      "entropy": 5.606517553329468,
      "epoch": 2.190353314807463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004522558734354755,
      "loss": 5.0478,
      "mean_token_accuracy": 0.17971969097852708,
      "num_tokens": 48345102.0,
      "step": 22070
    },
    {
      "entropy": 5.607900953292846,
      "epoch": 2.1908495434696307,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17461160868406295,
      "num_tokens": 48357175.0,
      "step": 22075
    },
    {
      "entropy": 5.583129739761352,
      "epoch": 2.191345772131798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17902783453464508,
      "num_tokens": 48369597.0,
      "step": 22080
    },
    {
      "entropy": 5.641428375244141,
      "epoch": 2.191842000793966,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004521903841010997,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17990517169237136,
      "num_tokens": 48380339.0,
      "step": 22085
    },
    {
      "entropy": 5.605413198471069,
      "epoch": 2.1923382294561335,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17769980877637864,
      "num_tokens": 48393001.0,
      "step": 22090
    },
    {
      "entropy": 5.540935611724853,
      "epoch": 2.192834458118301,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.027,
      "mean_token_accuracy": 0.18477399945259093,
      "num_tokens": 48403089.0,
      "step": 22095
    },
    {
      "entropy": 5.6534019947052006,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.0733,
      "mean_token_accuracy": 0.18110419809818268,
      "num_tokens": 48414017.0,
      "step": 22100
    },
    {
      "entropy": 5.6180808544158936,
      "epoch": 2.193826915442636,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.0878,
      "mean_token_accuracy": 0.18605958372354509,
      "num_tokens": 48426812.0,
      "step": 22105
    },
    {
      "entropy": 5.5099162578582765,
      "epoch": 2.1943231441048034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004520811473389023,
      "loss": 4.9414,
      "mean_token_accuracy": 0.18803781569004058,
      "num_tokens": 48437373.0,
      "step": 22110
    },
    {
      "entropy": 5.564790344238281,
      "epoch": 2.194819372766971,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.004,
      "mean_token_accuracy": 0.18187412172555922,
      "num_tokens": 48448250.0,
      "step": 22115
    },
    {
      "entropy": 5.606889152526856,
      "epoch": 2.1953156014291384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000452037421891614,
      "loss": 5.0778,
      "mean_token_accuracy": 0.17819123715162277,
      "num_tokens": 48459026.0,
      "step": 22120
    },
    {
      "entropy": 5.528990983963013,
      "epoch": 2.195811830091306,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045201555258247896,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18706662952899933,
      "num_tokens": 48469319.0,
      "step": 22125
    },
    {
      "entropy": 5.599331283569336,
      "epoch": 2.196308058753474,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.078,
      "mean_token_accuracy": 0.1825847789645195,
      "num_tokens": 48479777.0,
      "step": 22130
    },
    {
      "entropy": 5.6326963901519775,
      "epoch": 2.1968042874156413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.0945,
      "mean_token_accuracy": 0.1791132315993309,
      "num_tokens": 48491434.0,
      "step": 22135
    },
    {
      "entropy": 5.665299606323242,
      "epoch": 2.1973005160778087,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.1357,
      "mean_token_accuracy": 0.17376846373081206,
      "num_tokens": 48502526.0,
      "step": 22140
    },
    {
      "entropy": 5.604603719711304,
      "epoch": 2.197796744739976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.0429,
      "mean_token_accuracy": 0.17622127681970595,
      "num_tokens": 48512897.0,
      "step": 22145
    },
    {
      "entropy": 5.594691514968872,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.0229,
      "mean_token_accuracy": 0.1771448329091072,
      "num_tokens": 48523659.0,
      "step": 22150
    },
    {
      "entropy": 5.589489984512329,
      "epoch": 2.198789202064311,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.0483,
      "mean_token_accuracy": 0.17994043380022048,
      "num_tokens": 48533899.0,
      "step": 22155
    },
    {
      "entropy": 5.552674627304077,
      "epoch": 2.1992854307264786,
      "grad_norm": 1.0,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18455394953489304,
      "num_tokens": 48545693.0,
      "step": 22160
    },
    {
      "entropy": 5.642074537277222,
      "epoch": 2.199781659388646,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045184044012922154,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18118453174829482,
      "num_tokens": 48556411.0,
      "step": 22165
    },
    {
      "entropy": 5.63690447807312,
      "epoch": 2.200277888050814,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.0509,
      "mean_token_accuracy": 0.1764154002070427,
      "num_tokens": 48569385.0,
      "step": 22170
    },
    {
      "entropy": 5.6801872730255125,
      "epoch": 2.2007741167129815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.1363,
      "mean_token_accuracy": 0.1719836711883545,
      "num_tokens": 48580586.0,
      "step": 22175
    },
    {
      "entropy": 5.582874393463134,
      "epoch": 2.201270345375149,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.038,
      "mean_token_accuracy": 0.17950402349233627,
      "num_tokens": 48589557.0,
      "step": 22180
    },
    {
      "entropy": 5.621746683120728,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004517527786389349,
      "loss": 5.0189,
      "mean_token_accuracy": 0.1824157014489174,
      "num_tokens": 48601880.0,
      "step": 22185
    },
    {
      "entropy": 5.615210294723511,
      "epoch": 2.202262802699484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.0149,
      "mean_token_accuracy": 0.1842949017882347,
      "num_tokens": 48612469.0,
      "step": 22190
    },
    {
      "entropy": 5.63696403503418,
      "epoch": 2.2027590313616514,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.1101,
      "mean_token_accuracy": 0.18591355085372924,
      "num_tokens": 48622852.0,
      "step": 22195
    },
    {
      "entropy": 5.609521532058716,
      "epoch": 2.203255260023819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.009,
      "mean_token_accuracy": 0.1832035794854164,
      "num_tokens": 48634062.0,
      "step": 22200
    },
    {
      "entropy": 5.565167999267578,
      "epoch": 2.2037514886859864,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045166504701924176,
      "loss": 4.9901,
      "mean_token_accuracy": 0.17664708495140075,
      "num_tokens": 48644538.0,
      "step": 22205
    },
    {
      "entropy": 5.595347166061401,
      "epoch": 2.204247717348154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.0515,
      "mean_token_accuracy": 0.1781026691198349,
      "num_tokens": 48656053.0,
      "step": 22210
    },
    {
      "entropy": 5.57747631072998,
      "epoch": 2.2047439460103218,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004516211549217444,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17841352075338363,
      "num_tokens": 48667846.0,
      "step": 22215
    },
    {
      "entropy": 5.610389852523804,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.067,
      "mean_token_accuracy": 0.17821394950151442,
      "num_tokens": 48678087.0,
      "step": 22220
    },
    {
      "entropy": 5.508975219726563,
      "epoch": 2.2057364033346567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000451577245304951,
      "loss": 4.9738,
      "mean_token_accuracy": 0.189738130569458,
      "num_tokens": 48688727.0,
      "step": 22225
    },
    {
      "entropy": 5.515401220321655,
      "epoch": 2.206232631996824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515552839281792,
      "loss": 4.9086,
      "mean_token_accuracy": 0.184029284119606,
      "num_tokens": 48700370.0,
      "step": 22230
    },
    {
      "entropy": 5.547165584564209,
      "epoch": 2.2067288606589917,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004515333181732169,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18625444322824478,
      "num_tokens": 48710988.0,
      "step": 22235
    },
    {
      "entropy": 5.589523029327393,
      "epoch": 2.207225089321159,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045151134804060863,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18270520716905594,
      "num_tokens": 48721133.0,
      "step": 22240
    },
    {
      "entropy": 5.6289811611175535,
      "epoch": 2.2077213179833266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004514893735308993,
      "loss": 4.9709,
      "mean_token_accuracy": 0.19267261624336243,
      "num_tokens": 48732051.0,
      "step": 22245
    },
    {
      "entropy": 5.619219875335693,
      "epoch": 2.208217546645494,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.0757,
      "mean_token_accuracy": 0.17786315828561783,
      "num_tokens": 48742847.0,
      "step": 22250
    },
    {
      "entropy": 5.600520515441895,
      "epoch": 2.208713775307662,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.0611,
      "mean_token_accuracy": 0.18133440017700195,
      "num_tokens": 48754773.0,
      "step": 22255
    },
    {
      "entropy": 5.569276332855225,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004514234237446146,
      "loss": 4.999,
      "mean_token_accuracy": 0.17975999861955644,
      "num_tokens": 48765387.0,
      "step": 22260
    },
    {
      "entropy": 5.5128427028656,
      "epoch": 2.209706232631997,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045140143173195117,
      "loss": 4.9876,
      "mean_token_accuracy": 0.18637581169605255,
      "num_tokens": 48776532.0,
      "step": 22265
    },
    {
      "entropy": 5.625549983978272,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.1349,
      "mean_token_accuracy": 0.17634645998477935,
      "num_tokens": 48788690.0,
      "step": 22270
    },
    {
      "entropy": 5.64700984954834,
      "epoch": 2.210698689956332,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.0905,
      "mean_token_accuracy": 0.17987393736839294,
      "num_tokens": 48799375.0,
      "step": 22275
    },
    {
      "entropy": 5.6634604930877686,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.0718,
      "mean_token_accuracy": 0.17613878548145295,
      "num_tokens": 48810918.0,
      "step": 22280
    },
    {
      "entropy": 5.550295019149781,
      "epoch": 2.211691147280667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.0442,
      "mean_token_accuracy": 0.1795092761516571,
      "num_tokens": 48821225.0,
      "step": 22285
    },
    {
      "entropy": 5.6070112705230715,
      "epoch": 2.2121873759428343,
      "grad_norm": 1.125,
      "learning_rate": 0.0004512914060639143,
      "loss": 5.0484,
      "mean_token_accuracy": 0.18544497936964036,
      "num_tokens": 48831457.0,
      "step": 22290
    },
    {
      "entropy": 5.551849365234375,
      "epoch": 2.212683604605002,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004512693878131827,
      "loss": 4.9526,
      "mean_token_accuracy": 0.19135776460170745,
      "num_tokens": 48841795.0,
      "step": 22295
    },
    {
      "entropy": 5.618838500976563,
      "epoch": 2.2131798332671697,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.112,
      "mean_token_accuracy": 0.17136074155569075,
      "num_tokens": 48852921.0,
      "step": 22300
    },
    {
      "entropy": 5.5841169357299805,
      "epoch": 2.213676061929337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.026,
      "mean_token_accuracy": 0.17689517587423326,
      "num_tokens": 48863456.0,
      "step": 22305
    },
    {
      "entropy": 5.728140640258789,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.1391,
      "mean_token_accuracy": 0.17878762781620025,
      "num_tokens": 48873295.0,
      "step": 22310
    },
    {
      "entropy": 5.6040925025939945,
      "epoch": 2.214668519253672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18013451397418975,
      "num_tokens": 48884701.0,
      "step": 22315
    },
    {
      "entropy": 5.533820915222168,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17891496270895005,
      "num_tokens": 48896393.0,
      "step": 22320
    },
    {
      "entropy": 5.563909101486206,
      "epoch": 2.215660976578007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045113718653478984,
      "loss": 5.0257,
      "mean_token_accuracy": 0.17662313133478164,
      "num_tokens": 48907710.0,
      "step": 22325
    },
    {
      "entropy": 5.627797222137451,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.0879,
      "mean_token_accuracy": 0.1726559117436409,
      "num_tokens": 48918665.0,
      "step": 22330
    },
    {
      "entropy": 5.573796033859253,
      "epoch": 2.216653433902342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045109308449358114,
      "loss": 4.9786,
      "mean_token_accuracy": 0.18670932501554488,
      "num_tokens": 48931057.0,
      "step": 22335
    },
    {
      "entropy": 5.568115901947022,
      "epoch": 2.21714966256451,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004510710269226098,
      "loss": 5.0291,
      "mean_token_accuracy": 0.1754863902926445,
      "num_tokens": 48941470.0,
      "step": 22340
    },
    {
      "entropy": 5.536735868453979,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004510489649854564,
      "loss": 4.9885,
      "mean_token_accuracy": 0.18301912546157836,
      "num_tokens": 48953278.0,
      "step": 22345
    },
    {
      "entropy": 5.5474358081817625,
      "epoch": 2.218142119888845,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000451026898682668,
      "loss": 5.01,
      "mean_token_accuracy": 0.18348414301872254,
      "num_tokens": 48964265.0,
      "step": 22350
    },
    {
      "entropy": 5.623793745040894,
      "epoch": 2.2186383485510124,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.0965,
      "mean_token_accuracy": 0.17568760067224504,
      "num_tokens": 48975307.0,
      "step": 22355
    },
    {
      "entropy": 5.630650329589844,
      "epoch": 2.21913457721318,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.0708,
      "mean_token_accuracy": 0.18035320937633514,
      "num_tokens": 48985770.0,
      "step": 22360
    },
    {
      "entropy": 5.55918607711792,
      "epoch": 2.2196308058753473,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045096067358596536,
      "loss": 4.9697,
      "mean_token_accuracy": 0.1847071170806885,
      "num_tokens": 48996638.0,
      "step": 22365
    },
    {
      "entropy": 5.609777784347534,
      "epoch": 2.220127034537515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1754095122218132,
      "num_tokens": 49006974.0,
      "step": 22370
    },
    {
      "entropy": 5.53361554145813,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004509165017033569,
      "loss": 4.9996,
      "mean_token_accuracy": 0.18079848289489747,
      "num_tokens": 49018344.0,
      "step": 22375
    },
    {
      "entropy": 5.627915334701538,
      "epoch": 2.2211194918618498,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045089440921825354,
      "loss": 5.095,
      "mean_token_accuracy": 0.17400488406419753,
      "num_tokens": 49028720.0,
      "step": 22380
    },
    {
      "entropy": 5.592286872863769,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045087231237134787,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18224816620349885,
      "num_tokens": 49038884.0,
      "step": 22385
    },
    {
      "entropy": 5.621495771408081,
      "epoch": 2.222111949186185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045085021116318783,
      "loss": 5.0362,
      "mean_token_accuracy": 0.1806903064250946,
      "num_tokens": 49050031.0,
      "step": 22390
    },
    {
      "entropy": 5.680980777740478,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.151,
      "mean_token_accuracy": 0.16742316782474517,
      "num_tokens": 49060770.0,
      "step": 22395
    },
    {
      "entropy": 5.649902725219727,
      "epoch": 2.22310440651052,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.0138,
      "mean_token_accuracy": 0.18099208027124405,
      "num_tokens": 49070621.0,
      "step": 22400
    },
    {
      "entropy": 5.647969961166382,
      "epoch": 2.2236006351726876,
      "grad_norm": 1.25,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17687080055475235,
      "num_tokens": 49080452.0,
      "step": 22405
    },
    {
      "entropy": 5.640028238296509,
      "epoch": 2.224096863834855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1765598773956299,
      "num_tokens": 49091344.0,
      "step": 22410
    },
    {
      "entropy": 5.6083104610443115,
      "epoch": 2.2245930924970225,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17772969007492065,
      "num_tokens": 49103590.0,
      "step": 22415
    },
    {
      "entropy": 5.541432666778564,
      "epoch": 2.22508932115919,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045071751235858454,
      "loss": 4.9585,
      "mean_token_accuracy": 0.1833804130554199,
      "num_tokens": 49113923.0,
      "step": 22420
    },
    {
      "entropy": 5.648481655120849,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004506953806369956,
      "loss": 5.0897,
      "mean_token_accuracy": 0.17167526483535767,
      "num_tokens": 49124690.0,
      "step": 22425
    },
    {
      "entropy": 5.678071546554565,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.1049,
      "mean_token_accuracy": 0.173943392932415,
      "num_tokens": 49135546.0,
      "step": 22430
    },
    {
      "entropy": 5.638536691665649,
      "epoch": 2.226578007145693,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.0137,
      "mean_token_accuracy": 0.18680070638656615,
      "num_tokens": 49147446.0,
      "step": 22435
    },
    {
      "entropy": 5.477818250656128,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.0051,
      "mean_token_accuracy": 0.17926720827817916,
      "num_tokens": 49157224.0,
      "step": 22440
    },
    {
      "entropy": 5.540500164031982,
      "epoch": 2.227570464470028,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004506068101874614,
      "loss": 5.0762,
      "mean_token_accuracy": 0.17741056233644487,
      "num_tokens": 49169961.0,
      "step": 22445
    },
    {
      "entropy": 5.687937164306641,
      "epoch": 2.2280666931321953,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.0571,
      "mean_token_accuracy": 0.18221801072359084,
      "num_tokens": 49180486.0,
      "step": 22450
    },
    {
      "entropy": 5.591024017333984,
      "epoch": 2.2285629217943628,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045056249883247457,
      "loss": 4.983,
      "mean_token_accuracy": 0.1755169674754143,
      "num_tokens": 49191792.0,
      "step": 22455
    },
    {
      "entropy": 5.570288228988647,
      "epoch": 2.2290591504565302,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045054033662434903,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18707623928785325,
      "num_tokens": 49202268.0,
      "step": 22460
    },
    {
      "entropy": 5.603441572189331,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045051817006320127,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1763371780514717,
      "num_tokens": 49212677.0,
      "step": 22465
    },
    {
      "entropy": 5.554003620147705,
      "epoch": 2.230051607780865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.0339,
      "mean_token_accuracy": 0.1833575889468193,
      "num_tokens": 49224543.0,
      "step": 22470
    },
    {
      "entropy": 5.583997297286987,
      "epoch": 2.230547836443033,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.015,
      "mean_token_accuracy": 0.17375893592834474,
      "num_tokens": 49235495.0,
      "step": 22475
    },
    {
      "entropy": 5.643943500518799,
      "epoch": 2.2310440651052006,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004504516442671231,
      "loss": 4.984,
      "mean_token_accuracy": 0.18331888020038606,
      "num_tokens": 49248378.0,
      "step": 22480
    },
    {
      "entropy": 5.528069686889649,
      "epoch": 2.231540293767368,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045042946029938495,
      "loss": 4.9486,
      "mean_token_accuracy": 0.1879992365837097,
      "num_tokens": 49258519.0,
      "step": 22485
    },
    {
      "entropy": 5.626123857498169,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.0,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.0805,
      "mean_token_accuracy": 0.17352667599916458,
      "num_tokens": 49268418.0,
      "step": 22490
    },
    {
      "entropy": 5.628327894210815,
      "epoch": 2.232532751091703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.0668,
      "mean_token_accuracy": 0.18135936707258224,
      "num_tokens": 49278290.0,
      "step": 22495
    },
    {
      "entropy": 5.601427316665649,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045036288229673583,
      "loss": 4.9996,
      "mean_token_accuracy": 0.17824541181325912,
      "num_tokens": 49288523.0,
      "step": 22500
    },
    {
      "entropy": 5.575342655181885,
      "epoch": 2.233525208416038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045034068093120907,
      "loss": 5.0013,
      "mean_token_accuracy": 0.18380950540304183,
      "num_tokens": 49299123.0,
      "step": 22505
    },
    {
      "entropy": 5.626219606399536,
      "epoch": 2.2340214370782054,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.0789,
      "mean_token_accuracy": 0.17713601142168045,
      "num_tokens": 49310674.0,
      "step": 22510
    },
    {
      "entropy": 5.6521543025970455,
      "epoch": 2.2345176657403734,
      "grad_norm": 1.0,
      "learning_rate": 0.000450296265156493,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18119064569473267,
      "num_tokens": 49322606.0,
      "step": 22515
    },
    {
      "entropy": 5.564062547683716,
      "epoch": 2.235013894402541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045027405074840506,
      "loss": 4.9776,
      "mean_token_accuracy": 0.18648817092180253,
      "num_tokens": 49332817.0,
      "step": 22520
    },
    {
      "entropy": 5.536906957626343,
      "epoch": 2.2355101230647083,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.0318,
      "mean_token_accuracy": 0.17506596744060515,
      "num_tokens": 49344276.0,
      "step": 22525
    },
    {
      "entropy": 5.629969263076783,
      "epoch": 2.236006351726876,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.0785,
      "mean_token_accuracy": 0.17987145483493805,
      "num_tokens": 49354179.0,
      "step": 22530
    },
    {
      "entropy": 5.668520116806031,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17801381051540374,
      "num_tokens": 49364940.0,
      "step": 22535
    },
    {
      "entropy": 5.588645124435425,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004501851496573761,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18531485348939897,
      "num_tokens": 49374992.0,
      "step": 22540
    },
    {
      "entropy": 5.584466171264649,
      "epoch": 2.237495037713378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.0136,
      "mean_token_accuracy": 0.17764276415109634,
      "num_tokens": 49386521.0,
      "step": 22545
    },
    {
      "entropy": 5.558242464065552,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004501406730443714,
      "loss": 4.9316,
      "mean_token_accuracy": 0.1839029774069786,
      "num_tokens": 49396246.0,
      "step": 22550
    },
    {
      "entropy": 5.554070663452149,
      "epoch": 2.238487495037713,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045011842822292636,
      "loss": 4.9579,
      "mean_token_accuracy": 0.1798984095454216,
      "num_tokens": 49406158.0,
      "step": 22555
    },
    {
      "entropy": 5.642992734909058,
      "epoch": 2.238983723699881,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.0908,
      "mean_token_accuracy": 0.18003874868154526,
      "num_tokens": 49418189.0,
      "step": 22560
    },
    {
      "entropy": 5.598552227020264,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000450073925552909,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1799916595220566,
      "num_tokens": 49430385.0,
      "step": 22565
    },
    {
      "entropy": 5.601214361190796,
      "epoch": 2.239976181024216,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.0249,
      "mean_token_accuracy": 0.18009404391050338,
      "num_tokens": 49440279.0,
      "step": 22570
    },
    {
      "entropy": 5.577130556106567,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.0695,
      "mean_token_accuracy": 0.17490374147892,
      "num_tokens": 49450895.0,
      "step": 22575
    },
    {
      "entropy": 5.613034868240357,
      "epoch": 2.240968638348551,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1747937649488449,
      "num_tokens": 49462113.0,
      "step": 22580
    },
    {
      "entropy": 5.587055683135986,
      "epoch": 2.2414648670107185,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004499848681198173,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1722656235098839,
      "num_tokens": 49473972.0,
      "step": 22585
    },
    {
      "entropy": 5.610272359848023,
      "epoch": 2.241961095672886,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.0648,
      "mean_token_accuracy": 0.1795056238770485,
      "num_tokens": 49484706.0,
      "step": 22590
    },
    {
      "entropy": 5.653530740737915,
      "epoch": 2.2424573243350534,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.102,
      "mean_token_accuracy": 0.17893620133399962,
      "num_tokens": 49496732.0,
      "step": 22595
    },
    {
      "entropy": 5.550813961029053,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18146783262491226,
      "num_tokens": 49508615.0,
      "step": 22600
    },
    {
      "entropy": 5.5866106986999515,
      "epoch": 2.243449781659389,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044989574125876516,
      "loss": 5.0118,
      "mean_token_accuracy": 0.17501483857631683,
      "num_tokens": 49520081.0,
      "step": 22605
    },
    {
      "entropy": 5.531628274917603,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004498734486995267,
      "loss": 4.9321,
      "mean_token_accuracy": 0.18496207594871522,
      "num_tokens": 49531537.0,
      "step": 22610
    },
    {
      "entropy": 5.527130460739135,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.125,
      "learning_rate": 0.00044985115180380355,
      "loss": 4.9662,
      "mean_token_accuracy": 0.18681156635284424,
      "num_tokens": 49541829.0,
      "step": 22615
    },
    {
      "entropy": 5.525015544891358,
      "epoch": 2.2449384676458912,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.0439,
      "mean_token_accuracy": 0.18371747732162474,
      "num_tokens": 49554056.0,
      "step": 22620
    },
    {
      "entropy": 5.645674800872802,
      "epoch": 2.2454346963080587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044980654500511485,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17755382508039474,
      "num_tokens": 49565427.0,
      "step": 22625
    },
    {
      "entropy": 5.646420001983643,
      "epoch": 2.245930924970226,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.0401,
      "mean_token_accuracy": 0.18149691224098205,
      "num_tokens": 49574903.0,
      "step": 22630
    },
    {
      "entropy": 5.666992282867431,
      "epoch": 2.2464271536323936,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.1015,
      "mean_token_accuracy": 0.18116783052682878,
      "num_tokens": 49586721.0,
      "step": 22635
    },
    {
      "entropy": 5.612128591537475,
      "epoch": 2.246923382294561,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.0951,
      "mean_token_accuracy": 0.18403806835412978,
      "num_tokens": 49597212.0,
      "step": 22640
    },
    {
      "entropy": 5.584489727020264,
      "epoch": 2.247419610956729,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18435066491365432,
      "num_tokens": 49608235.0,
      "step": 22645
    },
    {
      "entropy": 5.572566843032837,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004496949521586279,
      "loss": 4.9826,
      "mean_token_accuracy": 0.189041630923748,
      "num_tokens": 49618460.0,
      "step": 22650
    },
    {
      "entropy": 5.614906311035156,
      "epoch": 2.248412068281064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18279426842927932,
      "num_tokens": 49629704.0,
      "step": 22655
    },
    {
      "entropy": 5.562216281890869,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1772376537322998,
      "num_tokens": 49640307.0,
      "step": 22660
    },
    {
      "entropy": 5.593123865127564,
      "epoch": 2.249404525605399,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1741635322570801,
      "num_tokens": 49651119.0,
      "step": 22665
    },
    {
      "entropy": 5.632040500640869,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1760463982820511,
      "num_tokens": 49661822.0,
      "step": 22670
    },
    {
      "entropy": 5.673927927017212,
      "epoch": 2.250396982929734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.0872,
      "mean_token_accuracy": 0.18454292863607408,
      "num_tokens": 49672864.0,
      "step": 22675
    },
    {
      "entropy": 5.552701616287232,
      "epoch": 2.2508932115919014,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.0417,
      "mean_token_accuracy": 0.1850741595029831,
      "num_tokens": 49683570.0,
      "step": 22680
    },
    {
      "entropy": 5.5706836700439455,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004495385402426128,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1907603159546852,
      "num_tokens": 49693297.0,
      "step": 22685
    },
    {
      "entropy": 5.566359901428223,
      "epoch": 2.2518856689162368,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000449516178366029,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18134918063879013,
      "num_tokens": 49705506.0,
      "step": 22690
    },
    {
      "entropy": 5.623693227767944,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044949381216182,
      "loss": 5.089,
      "mean_token_accuracy": 0.17399301677942275,
      "num_tokens": 49715594.0,
      "step": 22695
    },
    {
      "entropy": 5.616727638244629,
      "epoch": 2.2528781262405717,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18185445070266723,
      "num_tokens": 49725355.0,
      "step": 22700
    },
    {
      "entropy": 5.620272207260132,
      "epoch": 2.253374354902739,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17360759079456328,
      "num_tokens": 49735912.0,
      "step": 22705
    },
    {
      "entropy": 5.6132513046264645,
      "epoch": 2.2538705835649067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.0972,
      "mean_token_accuracy": 0.18550612777471542,
      "num_tokens": 49746807.0,
      "step": 22710
    },
    {
      "entropy": 5.607242202758789,
      "epoch": 2.254366812227074,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.103,
      "mean_token_accuracy": 0.17038858085870742,
      "num_tokens": 49757505.0,
      "step": 22715
    },
    {
      "entropy": 5.635884571075439,
      "epoch": 2.2548630408892416,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.1367,
      "mean_token_accuracy": 0.17256742417812349,
      "num_tokens": 49768689.0,
      "step": 22720
    },
    {
      "entropy": 5.682335615158081,
      "epoch": 2.255359269551409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.0967,
      "mean_token_accuracy": 0.16602427810430526,
      "num_tokens": 49781069.0,
      "step": 22725
    },
    {
      "entropy": 5.617661571502685,
      "epoch": 2.255855498213577,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.0356,
      "mean_token_accuracy": 0.1787629619240761,
      "num_tokens": 49792820.0,
      "step": 22730
    },
    {
      "entropy": 5.586894369125366,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1823272719979286,
      "num_tokens": 49802640.0,
      "step": 22735
    },
    {
      "entropy": 5.601014852523804,
      "epoch": 2.256847955537912,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044929232167235613,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17519067525863646,
      "num_tokens": 49813578.0,
      "step": 22740
    },
    {
      "entropy": 5.55746111869812,
      "epoch": 2.2573441842000794,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004492699122224179,
      "loss": 4.9958,
      "mean_token_accuracy": 0.18534359633922576,
      "num_tokens": 49824289.0,
      "step": 22745
    },
    {
      "entropy": 5.6322167873382565,
      "epoch": 2.257840412862247,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044924749845096144,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18271940797567368,
      "num_tokens": 49835235.0,
      "step": 22750
    },
    {
      "entropy": 5.619686603546143,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.1031,
      "mean_token_accuracy": 0.17193905264139175,
      "num_tokens": 49846266.0,
      "step": 22755
    },
    {
      "entropy": 5.574302053451538,
      "epoch": 2.258832870186582,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.0159,
      "mean_token_accuracy": 0.18278345912694932,
      "num_tokens": 49856642.0,
      "step": 22760
    },
    {
      "entropy": 5.692113971710205,
      "epoch": 2.2593290988487493,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.1669,
      "mean_token_accuracy": 0.17203570157289505,
      "num_tokens": 49868170.0,
      "step": 22765
    },
    {
      "entropy": 5.662557888031006,
      "epoch": 2.2598253275109172,
      "grad_norm": 1.125,
      "learning_rate": 0.0004491578001610703,
      "loss": 5.0933,
      "mean_token_accuracy": 0.18147562444210052,
      "num_tokens": 49878485.0,
      "step": 22770
    },
    {
      "entropy": 5.6299408912658695,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044913536479036163,
      "loss": 5.0891,
      "mean_token_accuracy": 0.17578182071447374,
      "num_tokens": 49889564.0,
      "step": 22775
    },
    {
      "entropy": 5.553850317001343,
      "epoch": 2.260817784835252,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044911292510147117,
      "loss": 5.0852,
      "mean_token_accuracy": 0.1779729902744293,
      "num_tokens": 49901081.0,
      "step": 22780
    },
    {
      "entropy": 5.579539442062378,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.042,
      "mean_token_accuracy": 0.1707979053258896,
      "num_tokens": 49910917.0,
      "step": 22785
    },
    {
      "entropy": 5.656940174102783,
      "epoch": 2.261810242159587,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.1172,
      "mean_token_accuracy": 0.17182111889123916,
      "num_tokens": 49923793.0,
      "step": 22790
    },
    {
      "entropy": 5.623887634277343,
      "epoch": 2.2623064708217546,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1783459261059761,
      "num_tokens": 49934989.0,
      "step": 22795
    },
    {
      "entropy": 5.607205724716186,
      "epoch": 2.262802699483922,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.1681,
      "mean_token_accuracy": 0.16612186431884765,
      "num_tokens": 49945961.0,
      "step": 22800
    },
    {
      "entropy": 5.680399465560913,
      "epoch": 2.2632989281460896,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004490006619037711,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1764674037694931,
      "num_tokens": 49958200.0,
      "step": 22805
    },
    {
      "entropy": 5.661529493331909,
      "epoch": 2.263795156808257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.0959,
      "mean_token_accuracy": 0.1779355600476265,
      "num_tokens": 49969962.0,
      "step": 22810
    },
    {
      "entropy": 5.547876024246216,
      "epoch": 2.2642913854704245,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044895572641690256,
      "loss": 5.024,
      "mean_token_accuracy": 0.18082612603902817,
      "num_tokens": 49979855.0,
      "step": 22815
    },
    {
      "entropy": 5.649654817581177,
      "epoch": 2.2647876141325924,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000448933252202599,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1808880612254143,
      "num_tokens": 49990905.0,
      "step": 22820
    },
    {
      "entropy": 5.65143084526062,
      "epoch": 2.26528384279476,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044891077367512544,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17827021777629853,
      "num_tokens": 50003188.0,
      "step": 22825
    },
    {
      "entropy": 5.605825805664063,
      "epoch": 2.2657800714569274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.0037,
      "mean_token_accuracy": 0.17808572053909302,
      "num_tokens": 50014538.0,
      "step": 22830
    },
    {
      "entropy": 5.560335731506347,
      "epoch": 2.266276300119095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.0501,
      "mean_token_accuracy": 0.17355345487594603,
      "num_tokens": 50025899.0,
      "step": 22835
    },
    {
      "entropy": 5.576333236694336,
      "epoch": 2.2667725287812623,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.0473,
      "mean_token_accuracy": 0.175111985206604,
      "num_tokens": 50036828.0,
      "step": 22840
    },
    {
      "entropy": 5.63362832069397,
      "epoch": 2.26726875744343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.078,
      "mean_token_accuracy": 0.1858227014541626,
      "num_tokens": 50047166.0,
      "step": 22845
    },
    {
      "entropy": 5.596086406707764,
      "epoch": 2.2677649861055973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.0239,
      "mean_token_accuracy": 0.18587104976177216,
      "num_tokens": 50058635.0,
      "step": 22850
    },
    {
      "entropy": 5.536551094055175,
      "epoch": 2.268261214767765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004487758119649403,
      "loss": 4.9366,
      "mean_token_accuracy": 0.18452418744564056,
      "num_tokens": 50068304.0,
      "step": 22855
    },
    {
      "entropy": 5.542117691040039,
      "epoch": 2.2687574434299327,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044875330326089153,
      "loss": 5.0251,
      "mean_token_accuracy": 0.18384317457675933,
      "num_tokens": 50079785.0,
      "step": 22860
    },
    {
      "entropy": 5.654488229751587,
      "epoch": 2.2692536720921,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004487307902481352,
      "loss": 5.0701,
      "mean_token_accuracy": 0.18243562132120134,
      "num_tokens": 50089637.0,
      "step": 22865
    },
    {
      "entropy": 5.6139976978302,
      "epoch": 2.2697499007542676,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044870827292722967,
      "loss": 4.9921,
      "mean_token_accuracy": 0.1787286415696144,
      "num_tokens": 50101447.0,
      "step": 22870
    },
    {
      "entropy": 5.627714347839356,
      "epoch": 2.270246129416435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.06,
      "mean_token_accuracy": 0.17460907995700836,
      "num_tokens": 50111488.0,
      "step": 22875
    },
    {
      "entropy": 5.560762071609497,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.0427,
      "mean_token_accuracy": 0.177193783223629,
      "num_tokens": 50121696.0,
      "step": 22880
    },
    {
      "entropy": 5.638032007217407,
      "epoch": 2.27123858674077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.2066,
      "mean_token_accuracy": 0.17209581434726715,
      "num_tokens": 50133106.0,
      "step": 22885
    },
    {
      "entropy": 5.6503184795379635,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18024697601795198,
      "num_tokens": 50143670.0,
      "step": 22890
    },
    {
      "entropy": 5.596884679794312,
      "epoch": 2.272231044065105,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18666804283857347,
      "num_tokens": 50155475.0,
      "step": 22895
    },
    {
      "entropy": 5.544524717330932,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004485730785619391,
      "loss": 4.9618,
      "mean_token_accuracy": 0.17822163105010985,
      "num_tokens": 50165830.0,
      "step": 22900
    },
    {
      "entropy": 5.569457054138184,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004485505310996295,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18877363502979277,
      "num_tokens": 50176101.0,
      "step": 22905
    },
    {
      "entropy": 5.577553462982178,
      "epoch": 2.273719730051608,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.0214,
      "mean_token_accuracy": 0.17834053039550782,
      "num_tokens": 50187021.0,
      "step": 22910
    },
    {
      "entropy": 5.544042110443115,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.0107,
      "mean_token_accuracy": 0.18148203045129777,
      "num_tokens": 50198646.0,
      "step": 22915
    },
    {
      "entropy": 5.578164863586426,
      "epoch": 2.274712187375943,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.0075,
      "mean_token_accuracy": 0.1827496588230133,
      "num_tokens": 50210088.0,
      "step": 22920
    },
    {
      "entropy": 5.563963985443115,
      "epoch": 2.2752084160381103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044846029821919193,
      "loss": 4.9776,
      "mean_token_accuracy": 0.18674665689468384,
      "num_tokens": 50221188.0,
      "step": 22925
    },
    {
      "entropy": 5.533077144622803,
      "epoch": 2.275704644700278,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.011,
      "mean_token_accuracy": 0.17924693673849107,
      "num_tokens": 50233225.0,
      "step": 22930
    },
    {
      "entropy": 5.574255847930909,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044841515596808553,
      "loss": 5.0026,
      "mean_token_accuracy": 0.1805005505681038,
      "num_tokens": 50244324.0,
      "step": 22935
    },
    {
      "entropy": 5.603653860092163,
      "epoch": 2.276697102024613,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044839257839176914,
      "loss": 4.8963,
      "mean_token_accuracy": 0.1899267926812172,
      "num_tokens": 50254060.0,
      "step": 22940
    },
    {
      "entropy": 5.665879583358764,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.1608,
      "mean_token_accuracy": 0.171154448390007,
      "num_tokens": 50264165.0,
      "step": 22945
    },
    {
      "entropy": 5.568433570861816,
      "epoch": 2.277689559348948,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.0644,
      "mean_token_accuracy": 0.18120863735675813,
      "num_tokens": 50274751.0,
      "step": 22950
    },
    {
      "entropy": 5.644810438156128,
      "epoch": 2.2781857880111156,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044832481986648633,
      "loss": 5.1647,
      "mean_token_accuracy": 0.17691460847854615,
      "num_tokens": 50286992.0,
      "step": 22955
    },
    {
      "entropy": 5.736995458602905,
      "epoch": 2.278682016673283,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000448302225094481,
      "loss": 5.1658,
      "mean_token_accuracy": 0.1699075311422348,
      "num_tokens": 50298462.0,
      "step": 22960
    },
    {
      "entropy": 5.6464659690856935,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.028,
      "mean_token_accuracy": 0.17416316121816636,
      "num_tokens": 50309717.0,
      "step": 22965
    },
    {
      "entropy": 5.573827266693115,
      "epoch": 2.279674473997618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044825702265846557,
      "loss": 4.9771,
      "mean_token_accuracy": 0.18042363673448564,
      "num_tokens": 50320334.0,
      "step": 22970
    },
    {
      "entropy": 5.549642610549927,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18132832646369934,
      "num_tokens": 50332028.0,
      "step": 22975
    },
    {
      "entropy": 5.529637432098388,
      "epoch": 2.280666931321953,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004482118030368469,
      "loss": 4.9419,
      "mean_token_accuracy": 0.18387202620506288,
      "num_tokens": 50343673.0,
      "step": 22980
    },
    {
      "entropy": 5.60453462600708,
      "epoch": 2.2811631599841204,
      "grad_norm": 1.09375,
      "learning_rate": 0.000448189186782838,
      "loss": 5.0802,
      "mean_token_accuracy": 0.16910840719938278,
      "num_tokens": 50353542.0,
      "step": 22985
    },
    {
      "entropy": 5.567365980148315,
      "epoch": 2.2816593886462884,
      "grad_norm": 1.0,
      "learning_rate": 0.00044816656623411044,
      "loss": 4.9947,
      "mean_token_accuracy": 0.17969014942646028,
      "num_tokens": 50365317.0,
      "step": 22990
    },
    {
      "entropy": 5.55487847328186,
      "epoch": 2.282155617308456,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004481439413912252,
      "loss": 4.997,
      "mean_token_accuracy": 0.18280373960733415,
      "num_tokens": 50376105.0,
      "step": 22995
    },
    {
      "entropy": 5.5574384212493895,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.125,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1773567259311676,
      "num_tokens": 50385854.0,
      "step": 23000
    },
    {
      "entropy": 5.563249158859253,
      "epoch": 2.283148074632791,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17650260627269745,
      "num_tokens": 50396869.0,
      "step": 23005
    },
    {
      "entropy": 5.5855059146881105,
      "epoch": 2.2836443032949583,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004480760411032343,
      "loss": 4.977,
      "mean_token_accuracy": 0.19075382947921754,
      "num_tokens": 50407620.0,
      "step": 23010
    },
    {
      "entropy": 5.573118066787719,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480533990893298,
      "loss": 4.9966,
      "mean_token_accuracy": 0.17942266017198563,
      "num_tokens": 50418401.0,
      "step": 23015
    },
    {
      "entropy": 5.532725620269775,
      "epoch": 2.284636760619293,
      "grad_norm": 1.0,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.0231,
      "mean_token_accuracy": 0.1835502192378044,
      "num_tokens": 50429835.0,
      "step": 23020
    },
    {
      "entropy": 5.6355215549469,
      "epoch": 2.285132989281461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.1036,
      "mean_token_accuracy": 0.181842665374279,
      "num_tokens": 50440072.0,
      "step": 23025
    },
    {
      "entropy": 5.620820760726929,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.053,
      "mean_token_accuracy": 0.1778666764497757,
      "num_tokens": 50450829.0,
      "step": 23030
    },
    {
      "entropy": 5.613069868087768,
      "epoch": 2.286125446605796,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044796278812581417,
      "loss": 5.0091,
      "mean_token_accuracy": 0.18036057502031327,
      "num_tokens": 50460913.0,
      "step": 23035
    },
    {
      "entropy": 5.526015710830689,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004479401246607695,
      "loss": 4.9736,
      "mean_token_accuracy": 0.18269770741462707,
      "num_tokens": 50473140.0,
      "step": 23040
    },
    {
      "entropy": 5.622762441635132,
      "epoch": 2.287117903930131,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.0894,
      "mean_token_accuracy": 0.18213336765766144,
      "num_tokens": 50483147.0,
      "step": 23045
    },
    {
      "entropy": 5.569093704223633,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478947848656149,
      "loss": 4.9381,
      "mean_token_accuracy": 0.1889767587184906,
      "num_tokens": 50493955.0,
      "step": 23050
    },
    {
      "entropy": 5.59017391204834,
      "epoch": 2.288110361254466,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.0385,
      "mean_token_accuracy": 0.17362988591194153,
      "num_tokens": 50505143.0,
      "step": 23055
    },
    {
      "entropy": 5.6439331531524655,
      "epoch": 2.2886065899166335,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004478494279207881,
      "loss": 5.0656,
      "mean_token_accuracy": 0.18090253174304963,
      "num_tokens": 50515515.0,
      "step": 23060
    },
    {
      "entropy": 5.618187522888183,
      "epoch": 2.289102818578801,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.0405,
      "mean_token_accuracy": 0.1765442818403244,
      "num_tokens": 50525046.0,
      "step": 23065
    },
    {
      "entropy": 5.525027227401734,
      "epoch": 2.2895990472409684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004478040538307879,
      "loss": 4.958,
      "mean_token_accuracy": 0.18630353212356568,
      "num_tokens": 50535906.0,
      "step": 23070
    },
    {
      "entropy": 5.577449893951416,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.0662,
      "mean_token_accuracy": 0.17763956040143966,
      "num_tokens": 50546791.0,
      "step": 23075
    },
    {
      "entropy": 5.636306238174439,
      "epoch": 2.290591504565304,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.0305,
      "mean_token_accuracy": 0.17770813703536986,
      "num_tokens": 50558150.0,
      "step": 23080
    },
    {
      "entropy": 5.659216642379761,
      "epoch": 2.2910877332274713,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.0868,
      "mean_token_accuracy": 0.17589824199676513,
      "num_tokens": 50569030.0,
      "step": 23085
    },
    {
      "entropy": 5.5988054275512695,
      "epoch": 2.2915839618896388,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.001,
      "mean_token_accuracy": 0.19024017602205276,
      "num_tokens": 50579872.0,
      "step": 23090
    },
    {
      "entropy": 5.6208216667175295,
      "epoch": 2.2920801905518062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.0668,
      "mean_token_accuracy": 0.1755719542503357,
      "num_tokens": 50591389.0,
      "step": 23095
    },
    {
      "entropy": 5.577130651473999,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044766782873476333,
      "loss": 5.1076,
      "mean_token_accuracy": 0.18438721001148223,
      "num_tokens": 50603732.0,
      "step": 23100
    },
    {
      "entropy": 5.552527141571045,
      "epoch": 2.293072647876141,
      "grad_norm": 1.03125,
      "learning_rate": 0.000447645109562542,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1847940653562546,
      "num_tokens": 50614950.0,
      "step": 23105
    },
    {
      "entropy": 5.588282489776612,
      "epoch": 2.293568876538309,
      "grad_norm": 1.0,
      "learning_rate": 0.00044762238610909414,
      "loss": 4.9944,
      "mean_token_accuracy": 0.18073246628046036,
      "num_tokens": 50625387.0,
      "step": 23110
    },
    {
      "entropy": 5.678070688247681,
      "epoch": 2.2940651052004766,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17689589262008668,
      "num_tokens": 50637017.0,
      "step": 23115
    },
    {
      "entropy": 5.548126459121704,
      "epoch": 2.294561333862644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18421871215105057,
      "num_tokens": 50647291.0,
      "step": 23120
    },
    {
      "entropy": 5.675448942184448,
      "epoch": 2.2950575625248115,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.1163,
      "mean_token_accuracy": 0.17458135187625884,
      "num_tokens": 50659750.0,
      "step": 23125
    },
    {
      "entropy": 5.6865575313568115,
      "epoch": 2.295553791186979,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17923236042261123,
      "num_tokens": 50671489.0,
      "step": 23130
    },
    {
      "entropy": 5.608805847167969,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004475087046431798,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18938604444265367,
      "num_tokens": 50682377.0,
      "step": 23135
    },
    {
      "entropy": 5.594154930114746,
      "epoch": 2.296546248511314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18252121806144714,
      "num_tokens": 50693502.0,
      "step": 23140
    },
    {
      "entropy": 5.6022123336792,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044746320210795753,
      "loss": 5.0437,
      "mean_token_accuracy": 0.17671750336885453,
      "num_tokens": 50704182.0,
      "step": 23145
    },
    {
      "entropy": 5.559465551376343,
      "epoch": 2.297538705835649,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474404444249906,
      "loss": 4.9698,
      "mean_token_accuracy": 0.18677013218402863,
      "num_tokens": 50714209.0,
      "step": 23150
    },
    {
      "entropy": 5.561301279067993,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044741768246587226,
      "loss": 4.9894,
      "mean_token_accuracy": 0.18078578114509583,
      "num_tokens": 50724862.0,
      "step": 23155
    },
    {
      "entropy": 5.672218942642212,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473949162311669,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1741903692483902,
      "num_tokens": 50735009.0,
      "step": 23160
    },
    {
      "entropy": 5.517476224899292,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004473721457214391,
      "loss": 4.9778,
      "mean_token_accuracy": 0.1795119062066078,
      "num_tokens": 50745883.0,
      "step": 23165
    },
    {
      "entropy": 5.575741815567016,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18408472388982772,
      "num_tokens": 50755341.0,
      "step": 23170
    },
    {
      "entropy": 5.5590643882751465,
      "epoch": 2.3000198491464867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473265918791749,
      "loss": 4.9626,
      "mean_token_accuracy": 0.1806015059351921,
      "num_tokens": 50765891.0,
      "step": 23175
    },
    {
      "entropy": 5.65436429977417,
      "epoch": 2.300516077808654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.0591,
      "mean_token_accuracy": 0.1760549411177635,
      "num_tokens": 50777228.0,
      "step": 23180
    },
    {
      "entropy": 5.605205345153808,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.1125,
      "mean_token_accuracy": 0.17398609668016435,
      "num_tokens": 50787334.0,
      "step": 23185
    },
    {
      "entropy": 5.588559961318969,
      "epoch": 2.301508535132989,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1770724281668663,
      "num_tokens": 50798433.0,
      "step": 23190
    },
    {
      "entropy": 5.579701662063599,
      "epoch": 2.3020047637951566,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044723543291922936,
      "loss": 4.9624,
      "mean_token_accuracy": 0.18964965045452117,
      "num_tokens": 50808560.0,
      "step": 23195
    },
    {
      "entropy": 5.538507413864136,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.0249,
      "mean_token_accuracy": 0.1792350083589554,
      "num_tokens": 50819328.0,
      "step": 23200
    },
    {
      "entropy": 5.605437517166138,
      "epoch": 2.302997221119492,
      "grad_norm": 1.125,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.0501,
      "mean_token_accuracy": 0.17772773653268814,
      "num_tokens": 50830066.0,
      "step": 23205
    },
    {
      "entropy": 5.674561977386475,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1725653290748596,
      "num_tokens": 50840775.0,
      "step": 23210
    },
    {
      "entropy": 5.613191986083985,
      "epoch": 2.303989678443827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1785319060087204,
      "num_tokens": 50850936.0,
      "step": 23215
    },
    {
      "entropy": 5.5791100025177,
      "epoch": 2.3044859071059944,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.071,
      "mean_token_accuracy": 0.1759073704481125,
      "num_tokens": 50862800.0,
      "step": 23220
    },
    {
      "entropy": 5.6158466815948485,
      "epoch": 2.304982135768162,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000447098566358597,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18558995872735978,
      "num_tokens": 50873450.0,
      "step": 23225
    },
    {
      "entropy": 5.579421329498291,
      "epoch": 2.3054783644303294,
      "grad_norm": 1.125,
      "learning_rate": 0.00044707574032500005,
      "loss": 4.9952,
      "mean_token_accuracy": 0.17953396141529082,
      "num_tokens": 50885624.0,
      "step": 23230
    },
    {
      "entropy": 5.671551847457886,
      "epoch": 2.305974593092497,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004470529100242955,
      "loss": 4.9922,
      "mean_token_accuracy": 0.18757929801940917,
      "num_tokens": 50896464.0,
      "step": 23235
    },
    {
      "entropy": 5.652951431274414,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.0326,
      "mean_token_accuracy": 0.18301300555467606,
      "num_tokens": 50907062.0,
      "step": 23240
    },
    {
      "entropy": 5.5259801864624025,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004470072366238284,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18168897330760955,
      "num_tokens": 50916637.0,
      "step": 23245
    },
    {
      "entropy": 5.529639196395874,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044698439352519836,
      "loss": 4.932,
      "mean_token_accuracy": 0.1859893888235092,
      "num_tokens": 50926573.0,
      "step": 23250
    },
    {
      "entropy": 5.571539831161499,
      "epoch": 2.307959507741167,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18214814215898514,
      "num_tokens": 50936767.0,
      "step": 23255
    },
    {
      "entropy": 5.6359296321868895,
      "epoch": 2.3084557364033347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1786173015832901,
      "num_tokens": 50948884.0,
      "step": 23260
    },
    {
      "entropy": 5.511661958694458,
      "epoch": 2.308951965065502,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004469158386425206,
      "loss": 4.9587,
      "mean_token_accuracy": 0.1821444109082222,
      "num_tokens": 50959634.0,
      "step": 23265
    },
    {
      "entropy": 5.592338371276855,
      "epoch": 2.3094481937276696,
      "grad_norm": 1.03125,
      "learning_rate": 0.000446892978487921,
      "loss": 5.065,
      "mean_token_accuracy": 0.17201694697141648,
      "num_tokens": 50970420.0,
      "step": 23270
    },
    {
      "entropy": 5.5929272174835205,
      "epoch": 2.309944422389837,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.0442,
      "mean_token_accuracy": 0.17950789779424667,
      "num_tokens": 50981564.0,
      "step": 23275
    },
    {
      "entropy": 5.597288799285889,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004468472453915623,
      "loss": 4.9753,
      "mean_token_accuracy": 0.1777655526995659,
      "num_tokens": 50991523.0,
      "step": 23280
    },
    {
      "entropy": 5.566152667999267,
      "epoch": 2.3109368797141725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004468243724509373,
      "loss": 5.0147,
      "mean_token_accuracy": 0.17989614456892014,
      "num_tokens": 51002339.0,
      "step": 23285
    },
    {
      "entropy": 5.632096576690674,
      "epoch": 2.31143310837634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.1113,
      "mean_token_accuracy": 0.17798396050930024,
      "num_tokens": 51013973.0,
      "step": 23290
    },
    {
      "entropy": 5.6462455749511715,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.0714,
      "mean_token_accuracy": 0.17770685702562333,
      "num_tokens": 51024450.0,
      "step": 23295
    },
    {
      "entropy": 5.577877998352051,
      "epoch": 2.312425565700675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18116877228021622,
      "num_tokens": 51035098.0,
      "step": 23300
    },
    {
      "entropy": 5.575752401351929,
      "epoch": 2.3129217943628424,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.0239,
      "mean_token_accuracy": 0.17938361465930938,
      "num_tokens": 51045402.0,
      "step": 23305
    },
    {
      "entropy": 5.638475322723389,
      "epoch": 2.31341802302501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.0391,
      "mean_token_accuracy": 0.18434927612543106,
      "num_tokens": 51056939.0,
      "step": 23310
    },
    {
      "entropy": 5.600464963912964,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044668704534868886,
      "loss": 4.9686,
      "mean_token_accuracy": 0.1866610527038574,
      "num_tokens": 51066952.0,
      "step": 23315
    },
    {
      "entropy": 5.609276676177979,
      "epoch": 2.314410480349345,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044666414259386173,
      "loss": 4.9814,
      "mean_token_accuracy": 0.18652106821537018,
      "num_tokens": 51076907.0,
      "step": 23320
    },
    {
      "entropy": 5.600595760345459,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.121,
      "mean_token_accuracy": 0.1762348085641861,
      "num_tokens": 51087089.0,
      "step": 23325
    },
    {
      "entropy": 5.598427629470825,
      "epoch": 2.31540293767368,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004466183243140729,
      "loss": 5.0299,
      "mean_token_accuracy": 0.18490330576896669,
      "num_tokens": 51097977.0,
      "step": 23330
    },
    {
      "entropy": 5.701250457763672,
      "epoch": 2.3158991663358477,
      "grad_norm": 1.0,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17635368108749389,
      "num_tokens": 51108929.0,
      "step": 23335
    },
    {
      "entropy": 5.6188946723937985,
      "epoch": 2.316395394998015,
      "grad_norm": 1.125,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18018334954977036,
      "num_tokens": 51118860.0,
      "step": 23340
    },
    {
      "entropy": 5.573257541656494,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.0522,
      "mean_token_accuracy": 0.17729189842939377,
      "num_tokens": 51130321.0,
      "step": 23345
    },
    {
      "entropy": 5.540111112594604,
      "epoch": 2.31738785232235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044652663668986634,
      "loss": 4.9732,
      "mean_token_accuracy": 0.18156736195087433,
      "num_tokens": 51140400.0,
      "step": 23350
    },
    {
      "entropy": 5.631522226333618,
      "epoch": 2.3178840809845176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000446503704148666,
      "loss": 5.0029,
      "mean_token_accuracy": 0.1832767516374588,
      "num_tokens": 51149823.0,
      "step": 23355
    },
    {
      "entropy": 5.534630346298218,
      "epoch": 2.318380309646685,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004464807673545433,
      "loss": 4.9518,
      "mean_token_accuracy": 0.18816331028938293,
      "num_tokens": 51161193.0,
      "step": 23360
    },
    {
      "entropy": 5.5682759284973145,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.0373,
      "mean_token_accuracy": 0.17989009320735933,
      "num_tokens": 51172566.0,
      "step": 23365
    },
    {
      "entropy": 5.571518611907959,
      "epoch": 2.3193727669710205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044643488100980623,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18994133919477463,
      "num_tokens": 51182484.0,
      "step": 23370
    },
    {
      "entropy": 5.550550889968872,
      "epoch": 2.319868995633188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004464119314603298,
      "loss": 4.93,
      "mean_token_accuracy": 0.18282828778028487,
      "num_tokens": 51193081.0,
      "step": 23375
    },
    {
      "entropy": 5.595645236968994,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.0806,
      "mean_token_accuracy": 0.1857695013284683,
      "num_tokens": 51203540.0,
      "step": 23380
    },
    {
      "entropy": 5.576986455917359,
      "epoch": 2.320861452957523,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004463660196100064,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18709612935781478,
      "num_tokens": 51213482.0,
      "step": 23385
    },
    {
      "entropy": 5.557799005508423,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044634305731029803,
      "loss": 4.9956,
      "mean_token_accuracy": 0.1784244731068611,
      "num_tokens": 51223520.0,
      "step": 23390
    },
    {
      "entropy": 5.634769010543823,
      "epoch": 2.321853910281858,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044632009076165096,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17884373813867568,
      "num_tokens": 51233133.0,
      "step": 23395
    },
    {
      "entropy": 5.616444253921509,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000446297119964635,
      "loss": 5.0467,
      "mean_token_accuracy": 0.18345650136470795,
      "num_tokens": 51245048.0,
      "step": 23400
    },
    {
      "entropy": 5.617425060272216,
      "epoch": 2.322846367606193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004462741449198194,
      "loss": 5.1033,
      "mean_token_accuracy": 0.1710338056087494,
      "num_tokens": 51256394.0,
      "step": 23405
    },
    {
      "entropy": 5.6440788269042965,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18399025797843932,
      "num_tokens": 51268311.0,
      "step": 23410
    },
    {
      "entropy": 5.5833333969116214,
      "epoch": 2.323838824930528,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004462281820890689,
      "loss": 4.9911,
      "mean_token_accuracy": 0.17955262362957,
      "num_tokens": 51278386.0,
      "step": 23415
    },
    {
      "entropy": 5.640593099594116,
      "epoch": 2.3243350535926957,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044620519430427376,
      "loss": 5.1344,
      "mean_token_accuracy": 0.18218934834003447,
      "num_tokens": 51288544.0,
      "step": 23420
    },
    {
      "entropy": 5.648425769805908,
      "epoch": 2.324831282254863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004461822022739588,
      "loss": 5.0774,
      "mean_token_accuracy": 0.18100918382406234,
      "num_tokens": 51300006.0,
      "step": 23425
    },
    {
      "entropy": 5.630337333679199,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000446159205998694,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18207605928182602,
      "num_tokens": 51311206.0,
      "step": 23430
    },
    {
      "entropy": 5.563255643844604,
      "epoch": 2.325823739579198,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.0059,
      "mean_token_accuracy": 0.1782529279589653,
      "num_tokens": 51321540.0,
      "step": 23435
    },
    {
      "entropy": 5.6191144466400145,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17714738845825195,
      "num_tokens": 51333058.0,
      "step": 23440
    },
    {
      "entropy": 5.658822011947632,
      "epoch": 2.326816196903533,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044609019170890424,
      "loss": 5.0611,
      "mean_token_accuracy": 0.1828753426671028,
      "num_tokens": 51344200.0,
      "step": 23445
    },
    {
      "entropy": 5.5618157386779785,
      "epoch": 2.3273124255657005,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.0328,
      "mean_token_accuracy": 0.17627465575933457,
      "num_tokens": 51355365.0,
      "step": 23450
    },
    {
      "entropy": 5.590382242202759,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044604416096808656,
      "loss": 5.0208,
      "mean_token_accuracy": 0.1779606819152832,
      "num_tokens": 51365328.0,
      "step": 23455
    },
    {
      "entropy": 5.680452919006347,
      "epoch": 2.328304882890036,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044602113923510235,
      "loss": 5.1317,
      "mean_token_accuracy": 0.18125548213720322,
      "num_tokens": 51376922.0,
      "step": 23460
    },
    {
      "entropy": 5.6454462051391605,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.0939,
      "mean_token_accuracy": 0.18207112848758697,
      "num_tokens": 51388183.0,
      "step": 23465
    },
    {
      "entropy": 5.6016875267028805,
      "epoch": 2.329297340214371,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.0284,
      "mean_token_accuracy": 0.1812193959951401,
      "num_tokens": 51399323.0,
      "step": 23470
    },
    {
      "entropy": 5.648498010635376,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004459520485926995,
      "loss": 5.0516,
      "mean_token_accuracy": 0.18521451950073242,
      "num_tokens": 51409350.0,
      "step": 23475
    },
    {
      "entropy": 5.687201356887817,
      "epoch": 2.330289797538706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.1201,
      "mean_token_accuracy": 0.1780110329389572,
      "num_tokens": 51420064.0,
      "step": 23480
    },
    {
      "entropy": 5.583702564239502,
      "epoch": 2.3307860262008733,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18519877791404724,
      "num_tokens": 51432129.0,
      "step": 23485
    },
    {
      "entropy": 5.638684701919556,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17811211943626404,
      "num_tokens": 51442795.0,
      "step": 23490
    },
    {
      "entropy": 5.558371257781983,
      "epoch": 2.3317784835252082,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044585986838943556,
      "loss": 4.9721,
      "mean_token_accuracy": 0.1831974059343338,
      "num_tokens": 51453551.0,
      "step": 23495
    },
    {
      "entropy": 5.604446172714233,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17054487764835358,
      "num_tokens": 51464997.0,
      "step": 23500
    },
    {
      "entropy": 5.554299640655517,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004458137528637781,
      "loss": 4.9829,
      "mean_token_accuracy": 0.18111877292394638,
      "num_tokens": 51476750.0,
      "step": 23505
    },
    {
      "entropy": 5.634915399551391,
      "epoch": 2.333267169511711,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.1049,
      "mean_token_accuracy": 0.18029374331235887,
      "num_tokens": 51487348.0,
      "step": 23510
    },
    {
      "entropy": 5.636059188842774,
      "epoch": 2.3337633981738786,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.031,
      "mean_token_accuracy": 0.18569215387105942,
      "num_tokens": 51498646.0,
      "step": 23515
    },
    {
      "entropy": 5.534091234207153,
      "epoch": 2.334259626836046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.0108,
      "mean_token_accuracy": 0.18248410075902938,
      "num_tokens": 51508881.0,
      "step": 23520
    },
    {
      "entropy": 5.630548191070557,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17876344472169875,
      "num_tokens": 51518959.0,
      "step": 23525
    },
    {
      "entropy": 5.574586629867554,
      "epoch": 2.335252084160381,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044569838993291757,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18125225305557252,
      "num_tokens": 51530472.0,
      "step": 23530
    },
    {
      "entropy": 5.601261949539184,
      "epoch": 2.3357483128225485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18525516390800476,
      "num_tokens": 51542639.0,
      "step": 23535
    },
    {
      "entropy": 5.621382474899292,
      "epoch": 2.3362445414847164,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.0515,
      "mean_token_accuracy": 0.1792745038866997,
      "num_tokens": 51553341.0,
      "step": 23540
    },
    {
      "entropy": 5.589063215255737,
      "epoch": 2.336740770146884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000445629121374397,
      "loss": 5.0126,
      "mean_token_accuracy": 0.17950734496116638,
      "num_tokens": 51564787.0,
      "step": 23545
    },
    {
      "entropy": 5.66425518989563,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17837681621313095,
      "num_tokens": 51576497.0,
      "step": 23550
    },
    {
      "entropy": 5.652095079421997,
      "epoch": 2.337733227471219,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18008643686771392,
      "num_tokens": 51587940.0,
      "step": 23555
    },
    {
      "entropy": 5.589194679260254,
      "epoch": 2.3382294561333863,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004455598147347661,
      "loss": 5.004,
      "mean_token_accuracy": 0.1828180029988289,
      "num_tokens": 51599103.0,
      "step": 23560
    },
    {
      "entropy": 5.633094263076782,
      "epoch": 2.3387256847955538,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004455367040617601,
      "loss": 5.1264,
      "mean_token_accuracy": 0.17346659153699875,
      "num_tokens": 51610257.0,
      "step": 23565
    },
    {
      "entropy": 5.599783992767334,
      "epoch": 2.3392219134577212,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.0276,
      "mean_token_accuracy": 0.18351719826459884,
      "num_tokens": 51620846.0,
      "step": 23570
    },
    {
      "entropy": 5.669289064407349,
      "epoch": 2.3397181421198887,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18636049181222916,
      "num_tokens": 51632432.0,
      "step": 23575
    },
    {
      "entropy": 5.568924283981323,
      "epoch": 2.340214370782056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004454673466713783,
      "loss": 4.999,
      "mean_token_accuracy": 0.1866435557603836,
      "num_tokens": 51644086.0,
      "step": 23580
    },
    {
      "entropy": 5.578968334197998,
      "epoch": 2.3407105994442237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.0328,
      "mean_token_accuracy": 0.17902909815311432,
      "num_tokens": 51655660.0,
      "step": 23585
    },
    {
      "entropy": 5.62666540145874,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044542108727405386,
      "loss": 5.11,
      "mean_token_accuracy": 0.17099392414093018,
      "num_tokens": 51667628.0,
      "step": 23590
    },
    {
      "entropy": 5.55895471572876,
      "epoch": 2.341703056768559,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044539795123599117,
      "loss": 4.9709,
      "mean_token_accuracy": 0.178879314661026,
      "num_tokens": 51677426.0,
      "step": 23595
    },
    {
      "entropy": 5.605389642715454,
      "epoch": 2.3421992854307265,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044537481097242654,
      "loss": 5.0784,
      "mean_token_accuracy": 0.17848048359155655,
      "num_tokens": 51688254.0,
      "step": 23600
    },
    {
      "entropy": 5.6096444606781,
      "epoch": 2.342695514092894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.0516,
      "mean_token_accuracy": 0.18006632030010222,
      "num_tokens": 51699852.0,
      "step": 23605
    },
    {
      "entropy": 5.587094640731811,
      "epoch": 2.3431917427550615,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044532851777108663,
      "loss": 5.0404,
      "mean_token_accuracy": 0.1825939491391182,
      "num_tokens": 51710758.0,
      "step": 23610
    },
    {
      "entropy": 5.4918536186218265,
      "epoch": 2.343687971417229,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004453053648344593,
      "loss": 4.9081,
      "mean_token_accuracy": 0.18454095870256423,
      "num_tokens": 51721151.0,
      "step": 23615
    },
    {
      "entropy": 5.665028285980225,
      "epoch": 2.3441842000793964,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1768116071820259,
      "num_tokens": 51733160.0,
      "step": 23620
    },
    {
      "entropy": 5.667888259887695,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.0031,
      "mean_token_accuracy": 0.184529809653759,
      "num_tokens": 51743923.0,
      "step": 23625
    },
    {
      "entropy": 5.5797525405883786,
      "epoch": 2.345176657403732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.0049,
      "mean_token_accuracy": 0.181864894926548,
      "num_tokens": 51754601.0,
      "step": 23630
    },
    {
      "entropy": 5.619714546203613,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004452127108616324,
      "loss": 5.1475,
      "mean_token_accuracy": 0.17320909351110458,
      "num_tokens": 51766041.0,
      "step": 23635
    },
    {
      "entropy": 5.547415018081665,
      "epoch": 2.3461691147280668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044518953681471824,
      "loss": 5.0054,
      "mean_token_accuracy": 0.18929387181997298,
      "num_tokens": 51777541.0,
      "step": 23640
    },
    {
      "entropy": 5.64255051612854,
      "epoch": 2.3466653433902342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.0521,
      "mean_token_accuracy": 0.18083659559488297,
      "num_tokens": 51789914.0,
      "step": 23645
    },
    {
      "entropy": 5.608322811126709,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044514317606046336,
      "loss": 4.9807,
      "mean_token_accuracy": 0.1909684956073761,
      "num_tokens": 51801289.0,
      "step": 23650
    },
    {
      "entropy": 5.560509729385376,
      "epoch": 2.347657800714569,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004451199893542723,
      "loss": 4.9577,
      "mean_token_accuracy": 0.1860889568924904,
      "num_tokens": 51811781.0,
      "step": 23655
    },
    {
      "entropy": 5.575207757949829,
      "epoch": 2.3481540293767367,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18313563168048858,
      "num_tokens": 51822100.0,
      "step": 23660
    },
    {
      "entropy": 5.610750818252564,
      "epoch": 2.348650258038904,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450736032866377,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18579019457101822,
      "num_tokens": 51832495.0,
      "step": 23665
    },
    {
      "entropy": 5.637111186981201,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17716218680143356,
      "num_tokens": 51842915.0,
      "step": 23670
    },
    {
      "entropy": 5.667268323898315,
      "epoch": 2.3496427153632395,
      "grad_norm": 1.0,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.1612,
      "mean_token_accuracy": 0.17329785227775574,
      "num_tokens": 51855268.0,
      "step": 23675
    },
    {
      "entropy": 5.584625339508056,
      "epoch": 2.350138944025407,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044500399255568274,
      "loss": 4.995,
      "mean_token_accuracy": 0.18678926676511765,
      "num_tokens": 51865756.0,
      "step": 23680
    },
    {
      "entropy": 5.571317434310913,
      "epoch": 2.3506351726875745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044498078054646507,
      "loss": 4.9442,
      "mean_token_accuracy": 0.1923266664147377,
      "num_tokens": 51875079.0,
      "step": 23685
    },
    {
      "entropy": 5.6054295063018795,
      "epoch": 2.351131401349742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.1059,
      "mean_token_accuracy": 0.17656407356262208,
      "num_tokens": 51886607.0,
      "step": 23690
    },
    {
      "entropy": 5.643296432495117,
      "epoch": 2.3516276300119094,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044493434388313423,
      "loss": 5.0661,
      "mean_token_accuracy": 0.1794283702969551,
      "num_tokens": 51897553.0,
      "step": 23695
    },
    {
      "entropy": 5.592515134811402,
      "epoch": 2.352123858674077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044491111923017254,
      "loss": 4.9791,
      "mean_token_accuracy": 0.18487852513790132,
      "num_tokens": 51908395.0,
      "step": 23700
    },
    {
      "entropy": 5.591972494125367,
      "epoch": 2.3526200873362444,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18437820822000503,
      "num_tokens": 51919764.0,
      "step": 23705
    },
    {
      "entropy": 5.526127338409424,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004448646572845363,
      "loss": 5.0226,
      "mean_token_accuracy": 0.17876419872045518,
      "num_tokens": 51930904.0,
      "step": 23710
    },
    {
      "entropy": 5.580008459091187,
      "epoch": 2.35361254466058,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18546722829341888,
      "num_tokens": 51942373.0,
      "step": 23715
    },
    {
      "entropy": 5.681336355209351,
      "epoch": 2.3541087733227473,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.0724,
      "mean_token_accuracy": 0.18706974089145662,
      "num_tokens": 51953770.0,
      "step": 23720
    },
    {
      "entropy": 5.600088119506836,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004447949327754418,
      "loss": 5.0573,
      "mean_token_accuracy": 0.17546446323394777,
      "num_tokens": 51964047.0,
      "step": 23725
    },
    {
      "entropy": 5.598326539993286,
      "epoch": 2.355101230647082,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.061,
      "mean_token_accuracy": 0.17683582305908202,
      "num_tokens": 51974813.0,
      "step": 23730
    },
    {
      "entropy": 5.566554069519043,
      "epoch": 2.3555974593092497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18178702890872955,
      "num_tokens": 51986321.0,
      "step": 23735
    },
    {
      "entropy": 5.584995222091675,
      "epoch": 2.356093687971417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.0515,
      "mean_token_accuracy": 0.17287956327199935,
      "num_tokens": 51996294.0,
      "step": 23740
    },
    {
      "entropy": 5.542879867553711,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.03125,
      "learning_rate": 0.000444701907818329,
      "loss": 5.0384,
      "mean_token_accuracy": 0.18626135885715484,
      "num_tokens": 52008359.0,
      "step": 23745
    },
    {
      "entropy": 5.629446411132813,
      "epoch": 2.357086145295752,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004446786410570048,
      "loss": 4.9957,
      "mean_token_accuracy": 0.1857800751924515,
      "num_tokens": 52019604.0,
      "step": 23750
    },
    {
      "entropy": 5.596062278747558,
      "epoch": 2.3575823739579196,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044465537008801575,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18314288556575775,
      "num_tokens": 52031238.0,
      "step": 23755
    },
    {
      "entropy": 5.556994152069092,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18140938729047776,
      "num_tokens": 52042108.0,
      "step": 23760
    },
    {
      "entropy": 5.610235261917114,
      "epoch": 2.358574831282255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044460881552935215,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18313467651605606,
      "num_tokens": 52052894.0,
      "step": 23765
    },
    {
      "entropy": 5.589718246459961,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044458553194083204,
      "loss": 4.989,
      "mean_token_accuracy": 0.18542774766683578,
      "num_tokens": 52063622.0,
      "step": 23770
    },
    {
      "entropy": 5.532653093338013,
      "epoch": 2.35956728860659,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004445622441469562,
      "loss": 4.9504,
      "mean_token_accuracy": 0.1850046619772911,
      "num_tokens": 52074457.0,
      "step": 23775
    },
    {
      "entropy": 5.58157377243042,
      "epoch": 2.3600635172687574,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044453895214830195,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18183585852384568,
      "num_tokens": 52086067.0,
      "step": 23780
    },
    {
      "entropy": 5.539091300964356,
      "epoch": 2.360559745930925,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004445156559454471,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18000859022140503,
      "num_tokens": 52096449.0,
      "step": 23785
    },
    {
      "entropy": 5.5982917785644535,
      "epoch": 2.3610559745930924,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.0558,
      "mean_token_accuracy": 0.1829450324177742,
      "num_tokens": 52107930.0,
      "step": 23790
    },
    {
      "entropy": 5.660968399047851,
      "epoch": 2.36155220325526,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17613216936588288,
      "num_tokens": 52119132.0,
      "step": 23795
    },
    {
      "entropy": 5.624952936172486,
      "epoch": 2.3620484319174277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004444457421174558,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18311311453580856,
      "num_tokens": 52129634.0,
      "step": 23800
    },
    {
      "entropy": 5.578297805786133,
      "epoch": 2.3625446605795952,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004444224291035761,
      "loss": 4.9661,
      "mean_token_accuracy": 0.18162842392921447,
      "num_tokens": 52140218.0,
      "step": 23805
    },
    {
      "entropy": 5.65509066581726,
      "epoch": 2.3630408892417627,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044439911188838533,
      "loss": 5.0885,
      "mean_token_accuracy": 0.182084758579731,
      "num_tokens": 52150795.0,
      "step": 23810
    },
    {
      "entropy": 5.526303291320801,
      "epoch": 2.36353711790393,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004443757904724615,
      "loss": 4.9381,
      "mean_token_accuracy": 0.18595149964094163,
      "num_tokens": 52161240.0,
      "step": 23815
    },
    {
      "entropy": 5.577584457397461,
      "epoch": 2.3640333465660976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1864042803645134,
      "num_tokens": 52172389.0,
      "step": 23820
    },
    {
      "entropy": 5.618075370788574,
      "epoch": 2.364529575228265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004443291350407285,
      "loss": 5.0181,
      "mean_token_accuracy": 0.18166959881782532,
      "num_tokens": 52183799.0,
      "step": 23825
    },
    {
      "entropy": 5.558178186416626,
      "epoch": 2.3650258038904326,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004443058010260762,
      "loss": 4.9524,
      "mean_token_accuracy": 0.1886661559343338,
      "num_tokens": 52195291.0,
      "step": 23830
    },
    {
      "entropy": 5.564321613311767,
      "epoch": 2.3655220325526,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044428246281300493,
      "loss": 5.0781,
      "mean_token_accuracy": 0.17610710710287095,
      "num_tokens": 52205952.0,
      "step": 23835
    },
    {
      "entropy": 5.636321401596069,
      "epoch": 2.3660182612147675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.0929,
      "mean_token_accuracy": 0.18116055577993392,
      "num_tokens": 52216995.0,
      "step": 23840
    },
    {
      "entropy": 5.632485198974609,
      "epoch": 2.3665144898769355,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.0556,
      "mean_token_accuracy": 0.18110557794570922,
      "num_tokens": 52228297.0,
      "step": 23845
    },
    {
      "entropy": 5.638520050048828,
      "epoch": 2.367010718539103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.014,
      "mean_token_accuracy": 0.18110352009534836,
      "num_tokens": 52238918.0,
      "step": 23850
    },
    {
      "entropy": 5.591239786148071,
      "epoch": 2.3675069472012704,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044418906798810554,
      "loss": 4.9977,
      "mean_token_accuracy": 0.18289946317672728,
      "num_tokens": 52249277.0,
      "step": 23855
    },
    {
      "entropy": 5.612111234664917,
      "epoch": 2.368003175863438,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1756119206547737,
      "num_tokens": 52259404.0,
      "step": 23860
    },
    {
      "entropy": 5.596198987960816,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.0453,
      "mean_token_accuracy": 0.1849292755126953,
      "num_tokens": 52271368.0,
      "step": 23865
    },
    {
      "entropy": 5.592835474014282,
      "epoch": 2.368995633187773,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1793601855635643,
      "num_tokens": 52282176.0,
      "step": 23870
    },
    {
      "entropy": 5.6596146583557125,
      "epoch": 2.3694918618499403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004440956060348191,
      "loss": 5.1223,
      "mean_token_accuracy": 0.1747632160782814,
      "num_tokens": 52292593.0,
      "step": 23875
    },
    {
      "entropy": 5.605229139328003,
      "epoch": 2.369988090512108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1788872927427292,
      "num_tokens": 52302509.0,
      "step": 23880
    },
    {
      "entropy": 5.639726972579956,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18061091452836991,
      "num_tokens": 52314200.0,
      "step": 23885
    },
    {
      "entropy": 5.587806367874146,
      "epoch": 2.370980547836443,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004440254655391568,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17950550317764283,
      "num_tokens": 52325506.0,
      "step": 23890
    },
    {
      "entropy": 5.501121234893799,
      "epoch": 2.3714767764986107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004440020769902282,
      "loss": 4.8805,
      "mean_token_accuracy": 0.18822346031665801,
      "num_tokens": 52336466.0,
      "step": 23895
    },
    {
      "entropy": 5.655187797546387,
      "epoch": 2.371973005160778,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.0603,
      "mean_token_accuracy": 0.1817791298031807,
      "num_tokens": 52347320.0,
      "step": 23900
    },
    {
      "entropy": 5.621414136886597,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.025,
      "mean_token_accuracy": 0.19543628245592118,
      "num_tokens": 52358355.0,
      "step": 23905
    },
    {
      "entropy": 5.567061471939087,
      "epoch": 2.372965462485113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044393188620043893,
      "loss": 5.0017,
      "mean_token_accuracy": 0.1829056531190872,
      "num_tokens": 52369639.0,
      "step": 23910
    },
    {
      "entropy": 5.6227391242980955,
      "epoch": 2.3734616911472806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004439084808914419,
      "loss": 5.0006,
      "mean_token_accuracy": 0.1868258908390999,
      "num_tokens": 52381691.0,
      "step": 23915
    },
    {
      "entropy": 5.622168970108032,
      "epoch": 2.373957919809448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17904310375452043,
      "num_tokens": 52392927.0,
      "step": 23920
    },
    {
      "entropy": 5.591922330856323,
      "epoch": 2.3744541484716155,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.058,
      "mean_token_accuracy": 0.17826517969369887,
      "num_tokens": 52404812.0,
      "step": 23925
    },
    {
      "entropy": 5.688504266738891,
      "epoch": 2.3749503771337834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000443838239835375,
      "loss": 5.1141,
      "mean_token_accuracy": 0.1782968059182167,
      "num_tokens": 52417035.0,
      "step": 23930
    },
    {
      "entropy": 5.670331859588623,
      "epoch": 2.375446605795951,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18285551965236663,
      "num_tokens": 52428482.0,
      "step": 23935
    },
    {
      "entropy": 5.611729764938355,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.0971,
      "mean_token_accuracy": 0.17992724627256393,
      "num_tokens": 52438471.0,
      "step": 23940
    },
    {
      "entropy": 5.536352825164795,
      "epoch": 2.376439063120286,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044376796109788777,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18031635582447053,
      "num_tokens": 52450418.0,
      "step": 23945
    },
    {
      "entropy": 5.6206746101379395,
      "epoch": 2.3769352917824533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.0578,
      "mean_token_accuracy": 0.1843432292342186,
      "num_tokens": 52461774.0,
      "step": 23950
    },
    {
      "entropy": 5.636575222015381,
      "epoch": 2.377431520444621,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1796657606959343,
      "num_tokens": 52473232.0,
      "step": 23955
    },
    {
      "entropy": 5.619772815704346,
      "epoch": 2.3779277491067883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.062,
      "mean_token_accuracy": 0.1814310908317566,
      "num_tokens": 52484683.0,
      "step": 23960
    },
    {
      "entropy": 5.688262987136841,
      "epoch": 2.3784239777689558,
      "grad_norm": 1.0,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.1561,
      "mean_token_accuracy": 0.1738188698887825,
      "num_tokens": 52496578.0,
      "step": 23965
    },
    {
      "entropy": 5.556980991363526,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004436507461748577,
      "loss": 4.9392,
      "mean_token_accuracy": 0.18904527872800828,
      "num_tokens": 52506658.0,
      "step": 23970
    },
    {
      "entropy": 5.663687992095947,
      "epoch": 2.379416435093291,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004436272906414007,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17536018192768096,
      "num_tokens": 52516443.0,
      "step": 23975
    },
    {
      "entropy": 5.6468287944793705,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.0853,
      "mean_token_accuracy": 0.17348237037658693,
      "num_tokens": 52526858.0,
      "step": 23980
    },
    {
      "entropy": 5.644855165481568,
      "epoch": 2.380408892417626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.0722,
      "mean_token_accuracy": 0.17327545583248138,
      "num_tokens": 52538263.0,
      "step": 23985
    },
    {
      "entropy": 5.580693769454956,
      "epoch": 2.3809051210797936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18637015372514726,
      "num_tokens": 52548722.0,
      "step": 23990
    },
    {
      "entropy": 5.662294387817383,
      "epoch": 2.381401349741961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.123,
      "mean_token_accuracy": 0.17567546367645265,
      "num_tokens": 52560189.0,
      "step": 23995
    },
    {
      "entropy": 5.574150705337525,
      "epoch": 2.3818975784041285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18244971483945846,
      "num_tokens": 52569505.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.3898249991179705,
      "eval_loss": 5.161362171173096,
      "eval_mean_token_accuracy": 0.18073683235343802,
      "eval_num_tokens": 52569505.0,
      "eval_runtime": 20.3378,
      "eval_samples_per_second": 1655.486,
      "eval_steps_per_second": 206.954,
      "step": 24000
    },
    {
      "entropy": 5.621862983703613,
      "epoch": 2.382393807066296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17546323388814927,
      "num_tokens": 52580665.0,
      "step": 24005
    },
    {
      "entropy": 5.6208192825317385,
      "epoch": 2.3828900357284635,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.0924,
      "mean_token_accuracy": 0.1804695814847946,
      "num_tokens": 52591506.0,
      "step": 24010
    },
    {
      "entropy": 5.667171049118042,
      "epoch": 2.3833862643906314,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16945031583309172,
      "num_tokens": 52602004.0,
      "step": 24015
    },
    {
      "entropy": 5.600239944458008,
      "epoch": 2.383882493052799,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1818201169371605,
      "num_tokens": 52613035.0,
      "step": 24020
    },
    {
      "entropy": 5.584380769729615,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044339250541521264,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17927720993757248,
      "num_tokens": 52623732.0,
      "step": 24025
    },
    {
      "entropy": 5.659279203414917,
      "epoch": 2.384874950377134,
      "grad_norm": 1.125,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.1126,
      "mean_token_accuracy": 0.1693638637661934,
      "num_tokens": 52634325.0,
      "step": 24030
    },
    {
      "entropy": 5.6286383152008055,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004433454982420808,
      "loss": 4.995,
      "mean_token_accuracy": 0.1899133086204529,
      "num_tokens": 52645214.0,
      "step": 24035
    },
    {
      "entropy": 5.6201972484588625,
      "epoch": 2.3858674077014688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000443321988393315,
      "loss": 5.0303,
      "mean_token_accuracy": 0.18483551144599913,
      "num_tokens": 52655726.0,
      "step": 24040
    },
    {
      "entropy": 5.613071489334106,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004432984743705265,
      "loss": 5.1604,
      "mean_token_accuracy": 0.17940750420093537,
      "num_tokens": 52667129.0,
      "step": 24045
    },
    {
      "entropy": 5.630047798156738,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044327495617429866,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17934710085391997,
      "num_tokens": 52678707.0,
      "step": 24050
    },
    {
      "entropy": 5.662850570678711,
      "epoch": 2.3873560936879716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004432514338052144,
      "loss": 5.0329,
      "mean_token_accuracy": 0.17762040793895723,
      "num_tokens": 52688456.0,
      "step": 24055
    },
    {
      "entropy": 5.547899770736694,
      "epoch": 2.387852322350139,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004432279072638571,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19160602986812592,
      "num_tokens": 52699591.0,
      "step": 24060
    },
    {
      "entropy": 5.599808692932129,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.0428,
      "mean_token_accuracy": 0.17773233652114867,
      "num_tokens": 52710115.0,
      "step": 24065
    },
    {
      "entropy": 5.583066749572754,
      "epoch": 2.388844779674474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044318084166665737,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18074205368757248,
      "num_tokens": 52720214.0,
      "step": 24070
    },
    {
      "entropy": 5.630034065246582,
      "epoch": 2.3893410083366415,
      "grad_norm": 1.1875,
      "learning_rate": 0.000443157302611982,
      "loss": 4.9746,
      "mean_token_accuracy": 0.1861799106001854,
      "num_tokens": 52731523.0,
      "step": 24075
    },
    {
      "entropy": 5.630649614334106,
      "epoch": 2.389837236998809,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.035,
      "mean_token_accuracy": 0.183871828019619,
      "num_tokens": 52742770.0,
      "step": 24080
    },
    {
      "entropy": 5.568564224243164,
      "epoch": 2.3903334656609765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18095093667507173,
      "num_tokens": 52753939.0,
      "step": 24085
    },
    {
      "entropy": 5.65234112739563,
      "epoch": 2.390829694323144,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17421250194311141,
      "num_tokens": 52764446.0,
      "step": 24090
    },
    {
      "entropy": 5.586223554611206,
      "epoch": 2.3913259229853114,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.036,
      "mean_token_accuracy": 0.1829386383295059,
      "num_tokens": 52773933.0,
      "step": 24095
    },
    {
      "entropy": 5.554648303985596,
      "epoch": 2.391822151647479,
      "grad_norm": 1.125,
      "learning_rate": 0.00044303954480120155,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18748168349266053,
      "num_tokens": 52784481.0,
      "step": 24100
    },
    {
      "entropy": 5.571626758575439,
      "epoch": 2.392318380309647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18969246000051498,
      "num_tokens": 52795866.0,
      "step": 24105
    },
    {
      "entropy": 5.556217336654663,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18109638541936873,
      "num_tokens": 52806617.0,
      "step": 24110
    },
    {
      "entropy": 5.649314546585083,
      "epoch": 2.393310837633982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1808510884642601,
      "num_tokens": 52819246.0,
      "step": 24115
    },
    {
      "entropy": 5.560991048812866,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004429452635427394,
      "loss": 4.9893,
      "mean_token_accuracy": 0.1881849318742752,
      "num_tokens": 52829730.0,
      "step": 24120
    },
    {
      "entropy": 5.624324464797974,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044292168281496086,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1776331126689911,
      "num_tokens": 52840010.0,
      "step": 24125
    },
    {
      "entropy": 5.643479871749878,
      "epoch": 2.394799523620484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.0596,
      "mean_token_accuracy": 0.1809418499469757,
      "num_tokens": 52850619.0,
      "step": 24130
    },
    {
      "entropy": 5.66481294631958,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044287450886770075,
      "loss": 5.1153,
      "mean_token_accuracy": 0.17758038341999055,
      "num_tokens": 52862560.0,
      "step": 24135
    },
    {
      "entropy": 5.602255582809448,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004428509156493891,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18234355747699738,
      "num_tokens": 52872741.0,
      "step": 24140
    },
    {
      "entropy": 5.557645416259765,
      "epoch": 2.396288209606987,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004428273182687364,
      "loss": 5.0079,
      "mean_token_accuracy": 0.18092976808547973,
      "num_tokens": 52884026.0,
      "step": 24145
    },
    {
      "entropy": 5.623815393447876,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004428037167263279,
      "loss": 4.9896,
      "mean_token_accuracy": 0.19091151356697084,
      "num_tokens": 52894429.0,
      "step": 24150
    },
    {
      "entropy": 5.60607123374939,
      "epoch": 2.397280666931322,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044278011102274887,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1795090913772583,
      "num_tokens": 52907157.0,
      "step": 24155
    },
    {
      "entropy": 5.56346321105957,
      "epoch": 2.3977768955934895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.0233,
      "mean_token_accuracy": 0.18434469252824784,
      "num_tokens": 52918563.0,
      "step": 24160
    },
    {
      "entropy": 5.584967517852784,
      "epoch": 2.398273124255657,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004427328871344204,
      "loss": 4.9749,
      "mean_token_accuracy": 0.18580025136470796,
      "num_tokens": 52929879.0,
      "step": 24165
    },
    {
      "entropy": 5.520228290557862,
      "epoch": 2.3987693529178244,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004427092689508422,
      "loss": 4.9728,
      "mean_token_accuracy": 0.18313192576169968,
      "num_tokens": 52941474.0,
      "step": 24170
    },
    {
      "entropy": 5.594864845275879,
      "epoch": 2.399265581579992,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18165349811315537,
      "num_tokens": 52952326.0,
      "step": 24175
    },
    {
      "entropy": 5.611116647720337,
      "epoch": 2.3997618102421594,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000442662020107786,
      "loss": 5.0504,
      "mean_token_accuracy": 0.1801069289445877,
      "num_tokens": 52962546.0,
      "step": 24180
    },
    {
      "entropy": 5.599023151397705,
      "epoch": 2.400258038904327,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.0499,
      "mean_token_accuracy": 0.1802912786602974,
      "num_tokens": 52972619.0,
      "step": 24185
    },
    {
      "entropy": 5.629446125030517,
      "epoch": 2.400754267566495,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004426147546341027,
      "loss": 5.0561,
      "mean_token_accuracy": 0.1807327926158905,
      "num_tokens": 52982536.0,
      "step": 24190
    },
    {
      "entropy": 5.556130075454712,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.0703,
      "mean_token_accuracy": 0.1802510917186737,
      "num_tokens": 52994390.0,
      "step": 24195
    },
    {
      "entropy": 5.613465738296509,
      "epoch": 2.4017467248908297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044256747253448045,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18112504184246064,
      "num_tokens": 53004204.0,
      "step": 24200
    },
    {
      "entropy": 5.617527770996094,
      "epoch": 2.402242953552997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044254382525140784,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18815142512321473,
      "num_tokens": 53014045.0,
      "step": 24205
    },
    {
      "entropy": 5.6227819442749025,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044252017381360935,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17806114852428437,
      "num_tokens": 53024945.0,
      "step": 24210
    },
    {
      "entropy": 5.626154136657715,
      "epoch": 2.403235410877332,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004424965182216716,
      "loss": 4.9696,
      "mean_token_accuracy": 0.1844913586974144,
      "num_tokens": 53035203.0,
      "step": 24215
    },
    {
      "entropy": 5.56763801574707,
      "epoch": 2.4037316395394996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.0149,
      "mean_token_accuracy": 0.18382316678762436,
      "num_tokens": 53047359.0,
      "step": 24220
    },
    {
      "entropy": 5.627724981307983,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.0584,
      "mean_token_accuracy": 0.1791076809167862,
      "num_tokens": 53057716.0,
      "step": 24225
    },
    {
      "entropy": 5.644625425338745,
      "epoch": 2.404724096863835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044242552652688885,
      "loss": 5.1312,
      "mean_token_accuracy": 0.17430538237094878,
      "num_tokens": 53069087.0,
      "step": 24230
    },
    {
      "entropy": 5.627597093582153,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004424018543242608,
      "loss": 5.0007,
      "mean_token_accuracy": 0.17960938066244125,
      "num_tokens": 53079462.0,
      "step": 24235
    },
    {
      "entropy": 5.609508848190307,
      "epoch": 2.40571655418817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004423781779704275,
      "loss": 5.0276,
      "mean_token_accuracy": 0.18097779154777527,
      "num_tokens": 53090398.0,
      "step": 24240
    },
    {
      "entropy": 5.5495054721832275,
      "epoch": 2.4062127828503375,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004423544974659761,
      "loss": 4.9508,
      "mean_token_accuracy": 0.18420374840497972,
      "num_tokens": 53101250.0,
      "step": 24245
    },
    {
      "entropy": 5.561471080780029,
      "epoch": 2.406709011512505,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044233081281149373,
      "loss": 5.0204,
      "mean_token_accuracy": 0.18117234408855437,
      "num_tokens": 53112039.0,
      "step": 24250
    },
    {
      "entropy": 5.5501913070678714,
      "epoch": 2.4072052401746724,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044230712400756774,
      "loss": 4.9815,
      "mean_token_accuracy": 0.19275325983762742,
      "num_tokens": 53123348.0,
      "step": 24255
    },
    {
      "entropy": 5.6000988483428955,
      "epoch": 2.40770146883684,
      "grad_norm": 1.0,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.0297,
      "mean_token_accuracy": 0.18085428923368455,
      "num_tokens": 53133878.0,
      "step": 24260
    },
    {
      "entropy": 5.60840482711792,
      "epoch": 2.4081976974990074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.0267,
      "mean_token_accuracy": 0.1828615888953209,
      "num_tokens": 53144763.0,
      "step": 24265
    },
    {
      "entropy": 5.664021587371826,
      "epoch": 2.408693926161175,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.031,
      "mean_token_accuracy": 0.17810414880514144,
      "num_tokens": 53155550.0,
      "step": 24270
    },
    {
      "entropy": 5.5709240436553955,
      "epoch": 2.4091901548233428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044221232730917823,
      "loss": 4.9412,
      "mean_token_accuracy": 0.1935721904039383,
      "num_tokens": 53166719.0,
      "step": 24275
    },
    {
      "entropy": 5.530984020233154,
      "epoch": 2.4096863834855102,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.0153,
      "mean_token_accuracy": 0.18341570645570754,
      "num_tokens": 53177923.0,
      "step": 24280
    },
    {
      "entropy": 5.60233097076416,
      "epoch": 2.4101826121476777,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044216490407859995,
      "loss": 5.0413,
      "mean_token_accuracy": 0.18423030227422715,
      "num_tokens": 53188628.0,
      "step": 24285
    },
    {
      "entropy": 5.555488204956054,
      "epoch": 2.410678840809845,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044214118624502267,
      "loss": 4.9664,
      "mean_token_accuracy": 0.1881862536072731,
      "num_tokens": 53198617.0,
      "step": 24290
    },
    {
      "entropy": 5.6869261264801025,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.1293,
      "mean_token_accuracy": 0.17294076681137086,
      "num_tokens": 53210490.0,
      "step": 24295
    },
    {
      "entropy": 5.616922378540039,
      "epoch": 2.41167129813418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.1115,
      "mean_token_accuracy": 0.17635497450828552,
      "num_tokens": 53221309.0,
      "step": 24300
    },
    {
      "entropy": 5.675211811065674,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.0774,
      "mean_token_accuracy": 0.17812766283750534,
      "num_tokens": 53233342.0,
      "step": 24305
    },
    {
      "entropy": 5.610787105560303,
      "epoch": 2.4126637554585155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044204627346918407,
      "loss": 5.046,
      "mean_token_accuracy": 0.17989957630634307,
      "num_tokens": 53243557.0,
      "step": 24310
    },
    {
      "entropy": 5.624948024749756,
      "epoch": 2.413159984120683,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044202253491778404,
      "loss": 5.0023,
      "mean_token_accuracy": 0.18178834468126298,
      "num_tokens": 53253274.0,
      "step": 24315
    },
    {
      "entropy": 5.553084373474121,
      "epoch": 2.4136562127828505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.0037,
      "mean_token_accuracy": 0.18660663962364196,
      "num_tokens": 53263955.0,
      "step": 24320
    },
    {
      "entropy": 5.655791854858398,
      "epoch": 2.414152441445018,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000441975045390176,
      "loss": 5.1084,
      "mean_token_accuracy": 0.1816877841949463,
      "num_tokens": 53274538.0,
      "step": 24325
    },
    {
      "entropy": 5.670346879959107,
      "epoch": 2.4146486701071854,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17725464105606079,
      "num_tokens": 53284260.0,
      "step": 24330
    },
    {
      "entropy": 5.63478717803955,
      "epoch": 2.415144898769353,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044192753930008267,
      "loss": 5.0433,
      "mean_token_accuracy": 0.17561573386192322,
      "num_tokens": 53295818.0,
      "step": 24335
    },
    {
      "entropy": 5.583434295654297,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004419037800455765,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18114859461784363,
      "num_tokens": 53305576.0,
      "step": 24340
    },
    {
      "entropy": 5.528741121292114,
      "epoch": 2.416137356093688,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18434623926877974,
      "num_tokens": 53317218.0,
      "step": 24345
    },
    {
      "entropy": 5.598852491378784,
      "epoch": 2.4166335847558553,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.0168,
      "mean_token_accuracy": 0.18097465187311174,
      "num_tokens": 53328688.0,
      "step": 24350
    },
    {
      "entropy": 5.595192384719849,
      "epoch": 2.417129813418023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17692476660013198,
      "num_tokens": 53339806.0,
      "step": 24355
    },
    {
      "entropy": 5.634525203704834,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.0007,
      "mean_token_accuracy": 0.1859672635793686,
      "num_tokens": 53350088.0,
      "step": 24360
    },
    {
      "entropy": 5.547395706176758,
      "epoch": 2.418122270742358,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044178492170202153,
      "loss": 4.9603,
      "mean_token_accuracy": 0.19142677783966064,
      "num_tokens": 53360390.0,
      "step": 24365
    },
    {
      "entropy": 5.622951889038086,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.0592,
      "mean_token_accuracy": 0.18180756121873856,
      "num_tokens": 53371250.0,
      "step": 24370
    },
    {
      "entropy": 5.652459907531738,
      "epoch": 2.419114728066693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.0319,
      "mean_token_accuracy": 0.17401161789894104,
      "num_tokens": 53382228.0,
      "step": 24375
    },
    {
      "entropy": 5.6427665710449215,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044171355706029247,
      "loss": 5.1773,
      "mean_token_accuracy": 0.1743377298116684,
      "num_tokens": 53393383.0,
      "step": 24380
    },
    {
      "entropy": 5.556003713607788,
      "epoch": 2.420107185391028,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044168976057732225,
      "loss": 5.0489,
      "mean_token_accuracy": 0.18015406280755997,
      "num_tokens": 53404311.0,
      "step": 24385
    },
    {
      "entropy": 5.614069938659668,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004416659599608051,
      "loss": 5.0168,
      "mean_token_accuracy": 0.18537463545799254,
      "num_tokens": 53415730.0,
      "step": 24390
    },
    {
      "entropy": 5.602495765686035,
      "epoch": 2.4210996427153635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004416421552113312,
      "loss": 4.987,
      "mean_token_accuracy": 0.18438416868448257,
      "num_tokens": 53426190.0,
      "step": 24395
    },
    {
      "entropy": 5.610986280441284,
      "epoch": 2.421595871377531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000441618346329491,
      "loss": 5.0279,
      "mean_token_accuracy": 0.18196206539869308,
      "num_tokens": 53437430.0,
      "step": 24400
    },
    {
      "entropy": 5.60591139793396,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.0346,
      "mean_token_accuracy": 0.1869465172290802,
      "num_tokens": 53446645.0,
      "step": 24405
    },
    {
      "entropy": 5.564341926574707,
      "epoch": 2.422588328701866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044157071617107325,
      "loss": 4.9861,
      "mean_token_accuracy": 0.18414940536022187,
      "num_tokens": 53456609.0,
      "step": 24410
    },
    {
      "entropy": 5.554924154281617,
      "epoch": 2.4230845573640334,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044154689489567684,
      "loss": 4.9933,
      "mean_token_accuracy": 0.1790814831852913,
      "num_tokens": 53467235.0,
      "step": 24415
    },
    {
      "entropy": 5.646943092346191,
      "epoch": 2.423580786026201,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17824016362428666,
      "num_tokens": 53477447.0,
      "step": 24420
    },
    {
      "entropy": 5.642285585403442,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17858196347951888,
      "num_tokens": 53491104.0,
      "step": 24425
    },
    {
      "entropy": 5.711981678009034,
      "epoch": 2.424573243350536,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004414754062918262,
      "loss": 5.1095,
      "mean_token_accuracy": 0.17438221126794815,
      "num_tokens": 53501906.0,
      "step": 24430
    },
    {
      "entropy": 5.6084692001342775,
      "epoch": 2.4250694720127033,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.0957,
      "mean_token_accuracy": 0.18655532747507095,
      "num_tokens": 53512125.0,
      "step": 24435
    },
    {
      "entropy": 5.6700725078582765,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004414277265804508,
      "loss": 5.1523,
      "mean_token_accuracy": 0.1713351145386696,
      "num_tokens": 53524142.0,
      "step": 24440
    },
    {
      "entropy": 5.6796801567077635,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441403880533894,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18209469467401504,
      "num_tokens": 53534852.0,
      "step": 24445
    },
    {
      "entropy": 5.606662178039551,
      "epoch": 2.426558157999206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044138003036087956,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1773296818137169,
      "num_tokens": 53545452.0,
      "step": 24450
    },
    {
      "entropy": 5.58747968673706,
      "epoch": 2.4270543866613736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.0833,
      "mean_token_accuracy": 0.18347130715847015,
      "num_tokens": 53558101.0,
      "step": 24455
    },
    {
      "entropy": 5.665705871582031,
      "epoch": 2.427550615323541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004413323176378434,
      "loss": 5.0755,
      "mean_token_accuracy": 0.18738810271024703,
      "num_tokens": 53568142.0,
      "step": 24460
    },
    {
      "entropy": 5.606198358535766,
      "epoch": 2.4280468439857086,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17938447445631028,
      "num_tokens": 53579175.0,
      "step": 24465
    },
    {
      "entropy": 5.648486518859864,
      "epoch": 2.428543072647876,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.0537,
      "mean_token_accuracy": 0.18310323655605315,
      "num_tokens": 53589861.0,
      "step": 24470
    },
    {
      "entropy": 5.683528757095337,
      "epoch": 2.4290393013100435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.1402,
      "mean_token_accuracy": 0.17515061646699906,
      "num_tokens": 53601385.0,
      "step": 24475
    },
    {
      "entropy": 5.667871570587158,
      "epoch": 2.429535529972211,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.0557,
      "mean_token_accuracy": 0.1809667930006981,
      "num_tokens": 53611473.0,
      "step": 24480
    },
    {
      "entropy": 5.586551332473755,
      "epoch": 2.430031758634379,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.0648,
      "mean_token_accuracy": 0.1828792691230774,
      "num_tokens": 53623232.0,
      "step": 24485
    },
    {
      "entropy": 5.544074487686157,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044118908049528105,
      "loss": 4.9795,
      "mean_token_accuracy": 0.18715317845344542,
      "num_tokens": 53634127.0,
      "step": 24490
    },
    {
      "entropy": 5.615525388717652,
      "epoch": 2.431024215958714,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044116519321077435,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18532055020332336,
      "num_tokens": 53644471.0,
      "step": 24495
    },
    {
      "entropy": 5.6094917297363285,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.0,
      "learning_rate": 0.00044114130180572883,
      "loss": 5.1134,
      "mean_token_accuracy": 0.1688984140753746,
      "num_tokens": 53656241.0,
      "step": 24500
    },
    {
      "entropy": 5.590101480484009,
      "epoch": 2.432016673283049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.043,
      "mean_token_accuracy": 0.17973359376192094,
      "num_tokens": 53666916.0,
      "step": 24505
    },
    {
      "entropy": 5.61839599609375,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004410935066363916,
      "loss": 4.9746,
      "mean_token_accuracy": 0.18533049374818802,
      "num_tokens": 53677473.0,
      "step": 24510
    },
    {
      "entropy": 5.566709852218628,
      "epoch": 2.4330091306073838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.0079,
      "mean_token_accuracy": 0.18328261971473694,
      "num_tokens": 53688424.0,
      "step": 24515
    },
    {
      "entropy": 5.587289237976075,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.125,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.0296,
      "mean_token_accuracy": 0.18646875768899918,
      "num_tokens": 53700390.0,
      "step": 24520
    },
    {
      "entropy": 5.549588680267334,
      "epoch": 2.4340015879317187,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044102178299315994,
      "loss": 4.9913,
      "mean_token_accuracy": 0.18009667098522186,
      "num_tokens": 53711986.0,
      "step": 24525
    },
    {
      "entropy": 5.674306774139405,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.0778,
      "mean_token_accuracy": 0.18024307936429979,
      "num_tokens": 53723657.0,
      "step": 24530
    },
    {
      "entropy": 5.733411073684692,
      "epoch": 2.434994045256054,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.2074,
      "mean_token_accuracy": 0.175213523209095,
      "num_tokens": 53736352.0,
      "step": 24535
    },
    {
      "entropy": 5.615485906600952,
      "epoch": 2.4354902739182216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.0635,
      "mean_token_accuracy": 0.18037253320217134,
      "num_tokens": 53747799.0,
      "step": 24540
    },
    {
      "entropy": 5.591141223907471,
      "epoch": 2.435986502580389,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044092609383386517,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1764320820569992,
      "num_tokens": 53759263.0,
      "step": 24545
    },
    {
      "entropy": 5.631829452514649,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.1228,
      "mean_token_accuracy": 0.17238491773605347,
      "num_tokens": 53770589.0,
      "step": 24550
    },
    {
      "entropy": 5.617319869995117,
      "epoch": 2.436978959904724,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004408782245641872,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18752662986516952,
      "num_tokens": 53782851.0,
      "step": 24555
    },
    {
      "entropy": 5.611665058135986,
      "epoch": 2.4374751885668915,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004408542837589176,
      "loss": 4.9985,
      "mean_token_accuracy": 0.1851370006799698,
      "num_tokens": 53793685.0,
      "step": 24560
    },
    {
      "entropy": 5.5514562129974365,
      "epoch": 2.437971417229059,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1770240008831024,
      "num_tokens": 53805263.0,
      "step": 24565
    },
    {
      "entropy": 5.53056960105896,
      "epoch": 2.438467645891227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004408063898104856,
      "loss": 4.935,
      "mean_token_accuracy": 0.18838650584220887,
      "num_tokens": 53816146.0,
      "step": 24570
    },
    {
      "entropy": 5.618579626083374,
      "epoch": 2.4389638745533944,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.0384,
      "mean_token_accuracy": 0.18189147859811783,
      "num_tokens": 53826823.0,
      "step": 24575
    },
    {
      "entropy": 5.600149250030517,
      "epoch": 2.439460103215562,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044075847941548905,
      "loss": 5.0277,
      "mean_token_accuracy": 0.17829255014657974,
      "num_tokens": 53837132.0,
      "step": 24580
    },
    {
      "entropy": 5.599223613739014,
      "epoch": 2.4399563318777293,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.0139,
      "mean_token_accuracy": 0.18634710162878038,
      "num_tokens": 53847817.0,
      "step": 24585
    },
    {
      "entropy": 5.5392418384552,
      "epoch": 2.440452560539897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044071055257868,
      "loss": 4.9323,
      "mean_token_accuracy": 0.19061242043972015,
      "num_tokens": 53858678.0,
      "step": 24590
    },
    {
      "entropy": 5.575757122039795,
      "epoch": 2.4409487892020643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.021,
      "mean_token_accuracy": 0.17706422805786132,
      "num_tokens": 53868289.0,
      "step": 24595
    },
    {
      "entropy": 5.566511058807373,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044066260930481264,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1891264095902443,
      "num_tokens": 53880167.0,
      "step": 24600
    },
    {
      "entropy": 5.5844379425048825,
      "epoch": 2.441941246526399,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004406386315054681,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18826788514852524,
      "num_tokens": 53890176.0,
      "step": 24605
    },
    {
      "entropy": 5.624113130569458,
      "epoch": 2.4424374751885667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.0343,
      "mean_token_accuracy": 0.17987263202667236,
      "num_tokens": 53900620.0,
      "step": 24610
    },
    {
      "entropy": 5.572382688522339,
      "epoch": 2.4429337038507346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004405906635849302,
      "loss": 4.9836,
      "mean_token_accuracy": 0.18562400043010713,
      "num_tokens": 53911591.0,
      "step": 24615
    },
    {
      "entropy": 5.568274593353271,
      "epoch": 2.443429932512902,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004405666734649265,
      "loss": 4.9812,
      "mean_token_accuracy": 0.1882295310497284,
      "num_tokens": 53921425.0,
      "step": 24620
    },
    {
      "entropy": 5.6520920753479,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.0392,
      "mean_token_accuracy": 0.1835281729698181,
      "num_tokens": 53931150.0,
      "step": 24625
    },
    {
      "entropy": 5.5672520160675045,
      "epoch": 2.444422389837237,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.0163,
      "mean_token_accuracy": 0.18498409688472747,
      "num_tokens": 53942162.0,
      "step": 24630
    },
    {
      "entropy": 5.551008510589599,
      "epoch": 2.4449186184994045,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044049467847311486,
      "loss": 4.9681,
      "mean_token_accuracy": 0.18585221469402313,
      "num_tokens": 53953681.0,
      "step": 24635
    },
    {
      "entropy": 5.62298994064331,
      "epoch": 2.445414847161572,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.0435,
      "mean_token_accuracy": 0.1797647625207901,
      "num_tokens": 53965146.0,
      "step": 24640
    },
    {
      "entropy": 5.565868091583252,
      "epoch": 2.4459110758237395,
      "grad_norm": 1.0,
      "learning_rate": 0.00044044666129135825,
      "loss": 4.9575,
      "mean_token_accuracy": 0.18179396241903306,
      "num_tokens": 53976445.0,
      "step": 24645
    },
    {
      "entropy": 5.634640312194824,
      "epoch": 2.446407304485907,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00044042264654610115,
      "loss": 4.9867,
      "mean_token_accuracy": 0.19210554361343385,
      "num_tokens": 53987318.0,
      "step": 24650
    },
    {
      "entropy": 5.550256395339966,
      "epoch": 2.446903533148075,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18598165810108186,
      "num_tokens": 53998154.0,
      "step": 24655
    },
    {
      "entropy": 5.608504056930542,
      "epoch": 2.4473997618102423,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.0439,
      "mean_token_accuracy": 0.1842258542776108,
      "num_tokens": 54009488.0,
      "step": 24660
    },
    {
      "entropy": 5.560416889190674,
      "epoch": 2.44789599047241,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1798657387495041,
      "num_tokens": 54020199.0,
      "step": 24665
    },
    {
      "entropy": 5.618944120407105,
      "epoch": 2.4483922191345773,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044032654654977766,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18299289047718048,
      "num_tokens": 54031204.0,
      "step": 24670
    },
    {
      "entropy": 5.555148363113403,
      "epoch": 2.4488884477967447,
      "grad_norm": 1.0,
      "learning_rate": 0.0004403025112998519,
      "loss": 4.935,
      "mean_token_accuracy": 0.19107364416122435,
      "num_tokens": 54042223.0,
      "step": 24675
    },
    {
      "entropy": 5.537606430053711,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004402784719507799,
      "loss": 4.9481,
      "mean_token_accuracy": 0.1917503520846367,
      "num_tokens": 54053307.0,
      "step": 24680
    },
    {
      "entropy": 5.630038118362426,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.0486,
      "mean_token_accuracy": 0.17599643319845198,
      "num_tokens": 54062821.0,
      "step": 24685
    },
    {
      "entropy": 5.715665817260742,
      "epoch": 2.450377133783247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004402303809575825,
      "loss": 5.2182,
      "mean_token_accuracy": 0.16986918598413467,
      "num_tokens": 54074070.0,
      "step": 24690
    },
    {
      "entropy": 5.546520948410034,
      "epoch": 2.4508733624454146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004402063293146495,
      "loss": 4.9923,
      "mean_token_accuracy": 0.18053488582372665,
      "num_tokens": 54085399.0,
      "step": 24695
    },
    {
      "entropy": 5.64898157119751,
      "epoch": 2.451369591107582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17522631585597992,
      "num_tokens": 54096102.0,
      "step": 24700
    },
    {
      "entropy": 5.606035232543945,
      "epoch": 2.45186581976975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044015821373909707,
      "loss": 5.0454,
      "mean_token_accuracy": 0.1812588170170784,
      "num_tokens": 54106878.0,
      "step": 24705
    },
    {
      "entropy": 5.645937061309814,
      "epoch": 2.4523620484319175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17618949115276336,
      "num_tokens": 54119164.0,
      "step": 24710
    },
    {
      "entropy": 5.632322359085083,
      "epoch": 2.452858277094085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.0068,
      "mean_token_accuracy": 0.18338830024003983,
      "num_tokens": 54129863.0,
      "step": 24715
    },
    {
      "entropy": 5.676909589767456,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.0,
      "learning_rate": 0.00044008600966050196,
      "loss": 5.1544,
      "mean_token_accuracy": 0.17161230444908143,
      "num_tokens": 54141042.0,
      "step": 24720
    },
    {
      "entropy": 5.622179269790649,
      "epoch": 2.45385073441842,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004400619334459531,
      "loss": 4.9979,
      "mean_token_accuracy": 0.17483333349227906,
      "num_tokens": 54151720.0,
      "step": 24725
    },
    {
      "entropy": 5.538681268692017,
      "epoch": 2.4543469630805874,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.0329,
      "mean_token_accuracy": 0.1833190843462944,
      "num_tokens": 54162711.0,
      "step": 24730
    },
    {
      "entropy": 5.602310371398926,
      "epoch": 2.454843191742755,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18217579871416092,
      "num_tokens": 54173735.0,
      "step": 24735
    },
    {
      "entropy": 5.615494680404663,
      "epoch": 2.455339420404923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.0528,
      "mean_token_accuracy": 0.17518673241138458,
      "num_tokens": 54185251.0,
      "step": 24740
    },
    {
      "entropy": 5.629410934448242,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.125,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.0465,
      "mean_token_accuracy": 0.180657559633255,
      "num_tokens": 54195915.0,
      "step": 24745
    },
    {
      "entropy": 5.6873410701751705,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.1759,
      "mean_token_accuracy": 0.17317407876253127,
      "num_tokens": 54207553.0,
      "step": 24750
    },
    {
      "entropy": 5.594030809402466,
      "epoch": 2.4568281063914252,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.018,
      "mean_token_accuracy": 0.18093765527009964,
      "num_tokens": 54218327.0,
      "step": 24755
    },
    {
      "entropy": 5.545742988586426,
      "epoch": 2.4573243350535927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043989328536851014,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19146056324243546,
      "num_tokens": 54229256.0,
      "step": 24760
    },
    {
      "entropy": 5.634311962127685,
      "epoch": 2.45782056371576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043986917642524465,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18169931471347808,
      "num_tokens": 54241241.0,
      "step": 24765
    },
    {
      "entropy": 5.618762445449829,
      "epoch": 2.4583167923779277,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17749235332012175,
      "num_tokens": 54252392.0,
      "step": 24770
    },
    {
      "entropy": 5.612084913253784,
      "epoch": 2.458813021040095,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.0034,
      "mean_token_accuracy": 0.18373685628175734,
      "num_tokens": 54262885.0,
      "step": 24775
    },
    {
      "entropy": 5.621937704086304,
      "epoch": 2.4593092497022626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17864258587360382,
      "num_tokens": 54273963.0,
      "step": 24780
    },
    {
      "entropy": 5.609233093261719,
      "epoch": 2.45980547836443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397726997741597,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18354406505823134,
      "num_tokens": 54283837.0,
      "step": 24785
    },
    {
      "entropy": 5.593465518951416,
      "epoch": 2.460301707026598,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004397485703948732,
      "loss": 4.9859,
      "mean_token_accuracy": 0.17949619740247727,
      "num_tokens": 54295752.0,
      "step": 24790
    },
    {
      "entropy": 5.692979860305786,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17461605966091157,
      "num_tokens": 54306853.0,
      "step": 24795
    },
    {
      "entropy": 5.5568922996521,
      "epoch": 2.461294164350933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.0223,
      "mean_token_accuracy": 0.1854265585541725,
      "num_tokens": 54318624.0,
      "step": 24800
    },
    {
      "entropy": 5.534363222122193,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.0449,
      "mean_token_accuracy": 0.18377525061368943,
      "num_tokens": 54330417.0,
      "step": 24805
    },
    {
      "entropy": 5.577790546417236,
      "epoch": 2.462286621675268,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043965201202961673,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18471207618713378,
      "num_tokens": 54340559.0,
      "step": 24810
    },
    {
      "entropy": 5.580596780776977,
      "epoch": 2.4627828503374354,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043962786222926807,
      "loss": 4.9174,
      "mean_token_accuracy": 0.1876722678542137,
      "num_tokens": 54350907.0,
      "step": 24815
    },
    {
      "entropy": 5.533764457702636,
      "epoch": 2.463279078999603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043960370834650325,
      "loss": 4.9012,
      "mean_token_accuracy": 0.19530907273292542,
      "num_tokens": 54361742.0,
      "step": 24820
    },
    {
      "entropy": 5.526121377944946,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1824989825487137,
      "num_tokens": 54372979.0,
      "step": 24825
    },
    {
      "entropy": 5.555322217941284,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004395553883361207,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17563277035951613,
      "num_tokens": 54383557.0,
      "step": 24830
    },
    {
      "entropy": 5.620133924484253,
      "epoch": 2.4647677649861057,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043953122220970127,
      "loss": 4.9744,
      "mean_token_accuracy": 0.18336397409439087,
      "num_tokens": 54395542.0,
      "step": 24835
    },
    {
      "entropy": 5.5982341289520265,
      "epoch": 2.465263993648273,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17614607959985734,
      "num_tokens": 54407540.0,
      "step": 24840
    },
    {
      "entropy": 5.645266580581665,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.1011,
      "mean_token_accuracy": 0.177418452501297,
      "num_tokens": 54419308.0,
      "step": 24845
    },
    {
      "entropy": 5.5902611255645756,
      "epoch": 2.466256450972608,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043945869935272194,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18274253755807876,
      "num_tokens": 54431104.0,
      "step": 24850
    },
    {
      "entropy": 5.5562177181243895,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043943451690982006,
      "loss": 4.9756,
      "mean_token_accuracy": 0.1868402451276779,
      "num_tokens": 54441501.0,
      "step": 24855
    },
    {
      "entropy": 5.561993741989136,
      "epoch": 2.467248908296943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.0093,
      "mean_token_accuracy": 0.18908584862947464,
      "num_tokens": 54452370.0,
      "step": 24860
    },
    {
      "entropy": 5.583293151855469,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043938613979175087,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17770748287439347,
      "num_tokens": 54463378.0,
      "step": 24865
    },
    {
      "entropy": 5.62570366859436,
      "epoch": 2.468241365621278,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17691177278757095,
      "num_tokens": 54474878.0,
      "step": 24870
    },
    {
      "entropy": 5.629361534118653,
      "epoch": 2.468737594283446,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043933774636799365,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18738810420036317,
      "num_tokens": 54487173.0,
      "step": 24875
    },
    {
      "entropy": 5.627760219573974,
      "epoch": 2.4692338229456134,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.0623,
      "mean_token_accuracy": 0.18058975040912628,
      "num_tokens": 54498856.0,
      "step": 24880
    },
    {
      "entropy": 5.518284845352173,
      "epoch": 2.469730051607781,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.0029,
      "mean_token_accuracy": 0.1843298852443695,
      "num_tokens": 54508988.0,
      "step": 24885
    },
    {
      "entropy": 5.617498826980591,
      "epoch": 2.4702262802699484,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.1383,
      "mean_token_accuracy": 0.16677508801221846,
      "num_tokens": 54521500.0,
      "step": 24890
    },
    {
      "entropy": 5.643680095672607,
      "epoch": 2.470722508932116,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043924091062261743,
      "loss": 5.0771,
      "mean_token_accuracy": 0.17512870877981185,
      "num_tokens": 54532656.0,
      "step": 24895
    },
    {
      "entropy": 5.704025077819824,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.125,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.0593,
      "mean_token_accuracy": 0.1801831007003784,
      "num_tokens": 54542621.0,
      "step": 24900
    },
    {
      "entropy": 5.603852558135986,
      "epoch": 2.471714966256451,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004391924683106037,
      "loss": 4.9926,
      "mean_token_accuracy": 0.18592435419559478,
      "num_tokens": 54554585.0,
      "step": 24905
    },
    {
      "entropy": 5.631532001495361,
      "epoch": 2.4722111949186187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.1584,
      "mean_token_accuracy": 0.1692544624209404,
      "num_tokens": 54566247.0,
      "step": 24910
    },
    {
      "entropy": 5.65964641571045,
      "epoch": 2.472707423580786,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.0524,
      "mean_token_accuracy": 0.17815046459436418,
      "num_tokens": 54577619.0,
      "step": 24915
    },
    {
      "entropy": 5.606850242614746,
      "epoch": 2.4732036522429537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004391197743069398,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18258052319288254,
      "num_tokens": 54587514.0,
      "step": 24920
    },
    {
      "entropy": 5.6557382106781,
      "epoch": 2.473699880905121,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.0612,
      "mean_token_accuracy": 0.1794478103518486,
      "num_tokens": 54598822.0,
      "step": 24925
    },
    {
      "entropy": 5.5942400932312015,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043907129128774525,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18718625009059905,
      "num_tokens": 54608724.0,
      "step": 24930
    },
    {
      "entropy": 5.573284482955932,
      "epoch": 2.474692338229456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18194435983896257,
      "num_tokens": 54618241.0,
      "step": 24935
    },
    {
      "entropy": 5.6116808414459225,
      "epoch": 2.4751885668916236,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004390227919940928,
      "loss": 5.0259,
      "mean_token_accuracy": 0.182367005944252,
      "num_tokens": 54628998.0,
      "step": 24940
    },
    {
      "entropy": 5.674281787872315,
      "epoch": 2.475684795553791,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.1356,
      "mean_token_accuracy": 0.17257674485445024,
      "num_tokens": 54639198.0,
      "step": 24945
    },
    {
      "entropy": 5.652850866317749,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043897427643079336,
      "loss": 5.0708,
      "mean_token_accuracy": 0.18327566981315613,
      "num_tokens": 54650308.0,
      "step": 24950
    },
    {
      "entropy": 5.689308023452758,
      "epoch": 2.476677252878126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.1664,
      "mean_token_accuracy": 0.17066766768693925,
      "num_tokens": 54662566.0,
      "step": 24955
    },
    {
      "entropy": 5.649431467056274,
      "epoch": 2.477173481540294,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000438925744602659,
      "loss": 5.0929,
      "mean_token_accuracy": 0.17782352417707442,
      "num_tokens": 54672624.0,
      "step": 24960
    },
    {
      "entropy": 5.620578813552856,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.0,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.0624,
      "mean_token_accuracy": 0.18287138044834136,
      "num_tokens": 54684506.0,
      "step": 24965
    },
    {
      "entropy": 5.69527850151062,
      "epoch": 2.478165938864629,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.0923,
      "mean_token_accuracy": 0.1710777387022972,
      "num_tokens": 54694923.0,
      "step": 24970
    },
    {
      "entropy": 5.65666389465332,
      "epoch": 2.4786621675267964,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.0902,
      "mean_token_accuracy": 0.17775825411081314,
      "num_tokens": 54706032.0,
      "step": 24975
    },
    {
      "entropy": 5.617386102676392,
      "epoch": 2.479158396188964,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18532098978757858,
      "num_tokens": 54717236.0,
      "step": 24980
    },
    {
      "entropy": 5.6473359107971195,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.0732,
      "mean_token_accuracy": 0.17353260666131973,
      "num_tokens": 54727178.0,
      "step": 24985
    },
    {
      "entropy": 5.628179597854614,
      "epoch": 2.480150853513299,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043878005157739486,
      "loss": 5.0421,
      "mean_token_accuracy": 0.1751515284180641,
      "num_tokens": 54738060.0,
      "step": 24990
    },
    {
      "entropy": 5.583707809448242,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004387557551881308,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18402256518602372,
      "num_tokens": 54749487.0,
      "step": 24995
    },
    {
      "entropy": 5.636624956130982,
      "epoch": 2.481143310837634,
      "grad_norm": 1.046875,
      "learning_rate": 0.000438731454738077,
      "loss": 5.0441,
      "mean_token_accuracy": 0.180065256357193,
      "num_tokens": 54760681.0,
      "step": 25000
    },
    {
      "entropy": 5.61650824546814,
      "epoch": 2.4816395394998016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043870715022783575,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18273367285728453,
      "num_tokens": 54774064.0,
      "step": 25005
    },
    {
      "entropy": 5.611126565933228,
      "epoch": 2.482135768161969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.068,
      "mean_token_accuracy": 0.17767339795827866,
      "num_tokens": 54786368.0,
      "step": 25010
    },
    {
      "entropy": 5.624009466171264,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004386585290292026,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17794309556484222,
      "num_tokens": 54796831.0,
      "step": 25015
    },
    {
      "entropy": 5.588738059997558,
      "epoch": 2.483128225486304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004386342123420161,
      "loss": 5.0895,
      "mean_token_accuracy": 0.1787877142429352,
      "num_tokens": 54806897.0,
      "step": 25020
    },
    {
      "entropy": 5.596393775939942,
      "epoch": 2.4836244541484715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043860989159705396,
      "loss": 5.0294,
      "mean_token_accuracy": 0.19057772755622865,
      "num_tokens": 54818877.0,
      "step": 25025
    },
    {
      "entropy": 5.6934051513671875,
      "epoch": 2.484120682810639,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.0806,
      "mean_token_accuracy": 0.18276965618133545,
      "num_tokens": 54830581.0,
      "step": 25030
    },
    {
      "entropy": 5.569712352752686,
      "epoch": 2.4846169114728065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043856123793621437,
      "loss": 4.9632,
      "mean_token_accuracy": 0.1831900507211685,
      "num_tokens": 54841344.0,
      "step": 25035
    },
    {
      "entropy": 5.549278211593628,
      "epoch": 2.485113140134974,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.0114,
      "mean_token_accuracy": 0.18186926245689392,
      "num_tokens": 54853333.0,
      "step": 25040
    },
    {
      "entropy": 5.63743839263916,
      "epoch": 2.485609368797142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.0396,
      "mean_token_accuracy": 0.183891598880291,
      "num_tokens": 54864461.0,
      "step": 25045
    },
    {
      "entropy": 5.5222009181976315,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004384882270267168,
      "loss": 4.8975,
      "mean_token_accuracy": 0.19392491579055787,
      "num_tokens": 54874766.0,
      "step": 25050
    },
    {
      "entropy": 5.521410226821899,
      "epoch": 2.486601826121477,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004384638819477681,
      "loss": 4.9966,
      "mean_token_accuracy": 0.18157795667648316,
      "num_tokens": 54885771.0,
      "step": 25055
    },
    {
      "entropy": 5.589873743057251,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043843953281526727,
      "loss": 5.0204,
      "mean_token_accuracy": 0.18167145699262618,
      "num_tokens": 54897233.0,
      "step": 25060
    },
    {
      "entropy": 5.628025674819947,
      "epoch": 2.487594283445812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043841517962981823,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18270446211099625,
      "num_tokens": 54907767.0,
      "step": 25065
    },
    {
      "entropy": 5.59253454208374,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043839082239202496,
      "loss": 5.035,
      "mean_token_accuracy": 0.18316146582365037,
      "num_tokens": 54918279.0,
      "step": 25070
    },
    {
      "entropy": 5.566329479217529,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.0352,
      "mean_token_accuracy": 0.1808595225214958,
      "num_tokens": 54928762.0,
      "step": 25075
    },
    {
      "entropy": 5.559020376205444,
      "epoch": 2.489082969432314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.0103,
      "mean_token_accuracy": 0.1858081504702568,
      "num_tokens": 54940471.0,
      "step": 25080
    },
    {
      "entropy": 5.625743103027344,
      "epoch": 2.489579198094482,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043831772637061944,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17795301526784896,
      "num_tokens": 54951359.0,
      "step": 25085
    },
    {
      "entropy": 5.7131908416748045,
      "epoch": 2.4900754267566496,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.1034,
      "mean_token_accuracy": 0.18122949302196503,
      "num_tokens": 54963062.0,
      "step": 25090
    },
    {
      "entropy": 5.678009510040283,
      "epoch": 2.490571655418817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043826897543903696,
      "loss": 5.1048,
      "mean_token_accuracy": 0.1843019649386406,
      "num_tokens": 54974074.0,
      "step": 25095
    },
    {
      "entropy": 5.602053833007813,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.0835,
      "mean_token_accuracy": 0.18227111250162126,
      "num_tokens": 54985543.0,
      "step": 25100
    },
    {
      "entropy": 5.639444971084595,
      "epoch": 2.491564112743152,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004382202083125795,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17217667400836945,
      "num_tokens": 54995939.0,
      "step": 25105
    },
    {
      "entropy": 5.622899103164673,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004381958186777842,
      "loss": 5.0828,
      "mean_token_accuracy": 0.18029522746801377,
      "num_tokens": 55006627.0,
      "step": 25110
    },
    {
      "entropy": 5.644660568237304,
      "epoch": 2.492556570067487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.0335,
      "mean_token_accuracy": 0.1844552531838417,
      "num_tokens": 55016821.0,
      "step": 25115
    },
    {
      "entropy": 5.569949817657471,
      "epoch": 2.4930527987296545,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004381470272680846,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1764412999153137,
      "num_tokens": 55026938.0,
      "step": 25120
    },
    {
      "entropy": 5.605702495574951,
      "epoch": 2.493549027391822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004381226254943902,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18070409595966339,
      "num_tokens": 55039257.0,
      "step": 25125
    },
    {
      "entropy": 5.687619256973266,
      "epoch": 2.49404525605399,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004380982196756062,
      "loss": 5.1151,
      "mean_token_accuracy": 0.17839727997779847,
      "num_tokens": 55049199.0,
      "step": 25130
    },
    {
      "entropy": 5.661785507202149,
      "epoch": 2.4945414847161573,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.074,
      "mean_token_accuracy": 0.17773790806531906,
      "num_tokens": 55060471.0,
      "step": 25135
    },
    {
      "entropy": 5.52242603302002,
      "epoch": 2.495037713378325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004380493959051903,
      "loss": 4.981,
      "mean_token_accuracy": 0.18619795888662338,
      "num_tokens": 55071332.0,
      "step": 25140
    },
    {
      "entropy": 5.627095508575439,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17947760671377183,
      "num_tokens": 55082823.0,
      "step": 25145
    },
    {
      "entropy": 5.629365539550781,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.058,
      "mean_token_accuracy": 0.17603259831666945,
      "num_tokens": 55094130.0,
      "step": 25150
    },
    {
      "entropy": 5.644581031799317,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.0447,
      "mean_token_accuracy": 0.18124939799308776,
      "num_tokens": 55104984.0,
      "step": 25155
    },
    {
      "entropy": 5.652818727493286,
      "epoch": 2.4970226280269947,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.0629,
      "mean_token_accuracy": 0.17909119874238968,
      "num_tokens": 55116691.0,
      "step": 25160
    },
    {
      "entropy": 5.614362192153931,
      "epoch": 2.497518856689162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004379272657324594,
      "loss": 4.9934,
      "mean_token_accuracy": 0.19258183985948563,
      "num_tokens": 55126605.0,
      "step": 25165
    },
    {
      "entropy": 5.617374563217163,
      "epoch": 2.49801508535133,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.045,
      "mean_token_accuracy": 0.18376689851284028,
      "num_tokens": 55137110.0,
      "step": 25170
    },
    {
      "entropy": 5.604290103912353,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043787838537741054,
      "loss": 5.0319,
      "mean_token_accuracy": 0.18091560304164886,
      "num_tokens": 55147416.0,
      "step": 25175
    },
    {
      "entropy": 5.602365303039551,
      "epoch": 2.499007542675665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004378539391410336,
      "loss": 4.9738,
      "mean_token_accuracy": 0.1851536825299263,
      "num_tokens": 55159643.0,
      "step": 25180
    },
    {
      "entropy": 5.55890383720398,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043782948886622987,
      "loss": 4.987,
      "mean_token_accuracy": 0.18475730419158937,
      "num_tokens": 55170196.0,
      "step": 25185
    },
    {
      "entropy": 5.574669599533081,
      "epoch": 2.5,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.0319,
      "mean_token_accuracy": 0.18446224629879,
      "num_tokens": 55181272.0,
      "step": 25190
    },
    {
      "entropy": 5.627013206481934,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.0515,
      "mean_token_accuracy": 0.18095765262842178,
      "num_tokens": 55191126.0,
      "step": 25195
    },
    {
      "entropy": 5.572001886367798,
      "epoch": 2.500992457324335,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043775611381732164,
      "loss": 4.952,
      "mean_token_accuracy": 0.18831441700458526,
      "num_tokens": 55201000.0,
      "step": 25200
    },
    {
      "entropy": 5.654511260986328,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.1875,
      "learning_rate": 0.000437731647394875,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17820547372102738,
      "num_tokens": 55211479.0,
      "step": 25205
    },
    {
      "entropy": 5.642922830581665,
      "epoch": 2.50198491464867,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004377071769370341,
      "loss": 5.006,
      "mean_token_accuracy": 0.1900687888264656,
      "num_tokens": 55220824.0,
      "step": 25210
    },
    {
      "entropy": 5.58218297958374,
      "epoch": 2.5024811433108374,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043768270244440587,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18347466886043548,
      "num_tokens": 55230857.0,
      "step": 25215
    },
    {
      "entropy": 5.564118766784668,
      "epoch": 2.5029773719730053,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043765822391759714,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1789253681898117,
      "num_tokens": 55241611.0,
      "step": 25220
    },
    {
      "entropy": 5.661553525924683,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.0847,
      "mean_token_accuracy": 0.17866115570068358,
      "num_tokens": 55252410.0,
      "step": 25225
    },
    {
      "entropy": 5.636365842819214,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.0614,
      "mean_token_accuracy": 0.1835451066493988,
      "num_tokens": 55262426.0,
      "step": 25230
    },
    {
      "entropy": 5.657071590423584,
      "epoch": 2.5044660579595077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004375847641381587,
      "loss": 5.1072,
      "mean_token_accuracy": 0.18098415434360504,
      "num_tokens": 55273240.0,
      "step": 25235
    },
    {
      "entropy": 5.624046373367309,
      "epoch": 2.504962286621675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.0509,
      "mean_token_accuracy": 0.18579548001289367,
      "num_tokens": 55284210.0,
      "step": 25240
    },
    {
      "entropy": 5.602657270431519,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.125,
      "learning_rate": 0.00043753577079209537,
      "loss": 4.9503,
      "mean_token_accuracy": 0.1903657540678978,
      "num_tokens": 55293752.0,
      "step": 25245
    },
    {
      "entropy": 5.608678817749023,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.125,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.1206,
      "mean_token_accuracy": 0.18305899798870087,
      "num_tokens": 55305260.0,
      "step": 25250
    },
    {
      "entropy": 5.7120805263519285,
      "epoch": 2.506450972608178,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.1158,
      "mean_token_accuracy": 0.1817322239279747,
      "num_tokens": 55316283.0,
      "step": 25255
    },
    {
      "entropy": 5.6573721408844,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.1508,
      "mean_token_accuracy": 0.16805029660463333,
      "num_tokens": 55327817.0,
      "step": 25260
    },
    {
      "entropy": 5.6244768619537355,
      "epoch": 2.507443429932513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043743773573838785,
      "loss": 5.1004,
      "mean_token_accuracy": 0.18015411347150803,
      "num_tokens": 55338406.0,
      "step": 25265
    },
    {
      "entropy": 5.649890041351318,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043741321690317673,
      "loss": 5.1052,
      "mean_token_accuracy": 0.1775095745921135,
      "num_tokens": 55349297.0,
      "step": 25270
    },
    {
      "entropy": 5.601337432861328,
      "epoch": 2.508435887256848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004373886940404677,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17846252024173737,
      "num_tokens": 55360229.0,
      "step": 25275
    },
    {
      "entropy": 5.555666494369507,
      "epoch": 2.5089321159190154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18491538912057875,
      "num_tokens": 55371570.0,
      "step": 25280
    },
    {
      "entropy": 5.561863422393799,
      "epoch": 2.509428344581183,
      "grad_norm": 1.0625,
      "learning_rate": 0.000437339636234989,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18419816792011262,
      "num_tokens": 55382781.0,
      "step": 25285
    },
    {
      "entropy": 5.596903371810913,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.0112,
      "mean_token_accuracy": 0.17911483347415924,
      "num_tokens": 55394631.0,
      "step": 25290
    },
    {
      "entropy": 5.632686758041382,
      "epoch": 2.510420801905518,
      "grad_norm": 1.125,
      "learning_rate": 0.00043729056232681766,
      "loss": 4.9732,
      "mean_token_accuracy": 0.19067534804344177,
      "num_tokens": 55406057.0,
      "step": 25295
    },
    {
      "entropy": 5.579542112350464,
      "epoch": 2.5109170305676853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.0039,
      "mean_token_accuracy": 0.18230041712522507,
      "num_tokens": 55416096.0,
      "step": 25300
    },
    {
      "entropy": 5.604883050918579,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004372414723208215,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18240909427404403,
      "num_tokens": 55427522.0,
      "step": 25305
    },
    {
      "entropy": 5.548048210144043,
      "epoch": 2.5119094878920207,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004372169212826606,
      "loss": 4.9901,
      "mean_token_accuracy": 0.18088106960058212,
      "num_tokens": 55438762.0,
      "step": 25310
    },
    {
      "entropy": 5.608601331710815,
      "epoch": 2.512405716554188,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.029,
      "mean_token_accuracy": 0.18277895003557204,
      "num_tokens": 55449552.0,
      "step": 25315
    },
    {
      "entropy": 5.622366285324096,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004371678071390576,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17786678224802016,
      "num_tokens": 55459610.0,
      "step": 25320
    },
    {
      "entropy": 5.578630876541138,
      "epoch": 2.513398173878523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043714324403483333,
      "loss": 4.9997,
      "mean_token_accuracy": 0.17966512590646744,
      "num_tokens": 55470212.0,
      "step": 25325
    },
    {
      "entropy": 5.537644290924073,
      "epoch": 2.5138944025406906,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043711867690980605,
      "loss": 4.9392,
      "mean_token_accuracy": 0.18994284123182298,
      "num_tokens": 55482639.0,
      "step": 25330
    },
    {
      "entropy": 5.549612951278687,
      "epoch": 2.5143906312028586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043709410576458477,
      "loss": 4.9389,
      "mean_token_accuracy": 0.19081737399101256,
      "num_tokens": 55493088.0,
      "step": 25335
    },
    {
      "entropy": 5.625964975357055,
      "epoch": 2.514886859865026,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.0558,
      "mean_token_accuracy": 0.1852238357067108,
      "num_tokens": 55503641.0,
      "step": 25340
    },
    {
      "entropy": 5.519804048538208,
      "epoch": 2.5153830885271935,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043704495141599816,
      "loss": 4.8893,
      "mean_token_accuracy": 0.18409974575042726,
      "num_tokens": 55514433.0,
      "step": 25345
    },
    {
      "entropy": 5.517437696456909,
      "epoch": 2.515879317189361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004370203682138517,
      "loss": 4.9406,
      "mean_token_accuracy": 0.182160584628582,
      "num_tokens": 55524646.0,
      "step": 25350
    },
    {
      "entropy": 5.58930287361145,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004369957809939492,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18230174481868744,
      "num_tokens": 55535587.0,
      "step": 25355
    },
    {
      "entropy": 5.6490990161895756,
      "epoch": 2.516871774513696,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004369711897569004,
      "loss": 5.0169,
      "mean_token_accuracy": 0.18055853992700577,
      "num_tokens": 55545861.0,
      "step": 25360
    },
    {
      "entropy": 5.655454826354981,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004369465945033151,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1778221383690834,
      "num_tokens": 55556930.0,
      "step": 25365
    },
    {
      "entropy": 5.64206280708313,
      "epoch": 2.517864231838031,
      "grad_norm": 1.0,
      "learning_rate": 0.00043692199523380325,
      "loss": 5.0858,
      "mean_token_accuracy": 0.18191395998001098,
      "num_tokens": 55570783.0,
      "step": 25370
    },
    {
      "entropy": 5.545043134689331,
      "epoch": 2.5183604605001984,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.0093,
      "mean_token_accuracy": 0.1841705173254013,
      "num_tokens": 55582123.0,
      "step": 25375
    },
    {
      "entropy": 5.601904964447021,
      "epoch": 2.518856689162366,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043687278464943986,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18321537375450134,
      "num_tokens": 55592977.0,
      "step": 25380
    },
    {
      "entropy": 5.5997992038726805,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004368481733358088,
      "loss": 4.9835,
      "mean_token_accuracy": 0.1918578863143921,
      "num_tokens": 55603179.0,
      "step": 25385
    },
    {
      "entropy": 5.601596212387085,
      "epoch": 2.519849146486701,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.0509,
      "mean_token_accuracy": 0.1793058469891548,
      "num_tokens": 55616144.0,
      "step": 25390
    },
    {
      "entropy": 5.650794792175293,
      "epoch": 2.5203453751488687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043679893866869904,
      "loss": 5.1227,
      "mean_token_accuracy": 0.1746693253517151,
      "num_tokens": 55626862.0,
      "step": 25395
    },
    {
      "entropy": 5.60001311302185,
      "epoch": 2.520841603811036,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004367743153164414,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18049713373184204,
      "num_tokens": 55637564.0,
      "step": 25400
    },
    {
      "entropy": 5.570561408996582,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043674968795252926,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18687838912010193,
      "num_tokens": 55647642.0,
      "step": 25405
    },
    {
      "entropy": 5.620639753341675,
      "epoch": 2.521834061135371,
      "grad_norm": 1.0,
      "learning_rate": 0.0004367250565775734,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1793530911207199,
      "num_tokens": 55659126.0,
      "step": 25410
    },
    {
      "entropy": 5.641758966445923,
      "epoch": 2.5223302897975386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004367004211921847,
      "loss": 4.999,
      "mean_token_accuracy": 0.18496960252523423,
      "num_tokens": 55669270.0,
      "step": 25415
    },
    {
      "entropy": 5.558000707626343,
      "epoch": 2.522826518459706,
      "grad_norm": 1.125,
      "learning_rate": 0.00043667578179697387,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18535540401935577,
      "num_tokens": 55680478.0,
      "step": 25420
    },
    {
      "entropy": 5.633058834075928,
      "epoch": 2.523322747121874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004366511383925521,
      "loss": 5.0412,
      "mean_token_accuracy": 0.17948123514652253,
      "num_tokens": 55691228.0,
      "step": 25425
    },
    {
      "entropy": 5.701124811172486,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.0,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.1057,
      "mean_token_accuracy": 0.18236868679523469,
      "num_tokens": 55702676.0,
      "step": 25430
    },
    {
      "entropy": 5.647402715682984,
      "epoch": 2.524315204446209,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004366018395585201,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18624897599220275,
      "num_tokens": 55711643.0,
      "step": 25435
    },
    {
      "entropy": 5.610743474960327,
      "epoch": 2.5248114331083764,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1796767845749855,
      "num_tokens": 55722478.0,
      "step": 25440
    },
    {
      "entropy": 5.611198282241821,
      "epoch": 2.525307661770544,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004365525246949786,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18192244023084642,
      "num_tokens": 55733489.0,
      "step": 25445
    },
    {
      "entropy": 5.602549076080322,
      "epoch": 2.5258038904327114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043652786125367035,
      "loss": 5.0095,
      "mean_token_accuracy": 0.1846819281578064,
      "num_tokens": 55745824.0,
      "step": 25450
    },
    {
      "entropy": 5.598931217193604,
      "epoch": 2.526300119094879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18556966930627822,
      "num_tokens": 55757024.0,
      "step": 25455
    },
    {
      "entropy": 5.647254228591919,
      "epoch": 2.5267963477570463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1746780291199684,
      "num_tokens": 55768674.0,
      "step": 25460
    },
    {
      "entropy": 5.614190530776978,
      "epoch": 2.527292576419214,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1820111334323883,
      "num_tokens": 55779226.0,
      "step": 25465
    },
    {
      "entropy": 5.681118774414062,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.1336,
      "mean_token_accuracy": 0.17475561052560806,
      "num_tokens": 55790470.0,
      "step": 25470
    },
    {
      "entropy": 5.730115270614624,
      "epoch": 2.528285033743549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000436404483976221,
      "loss": 5.1165,
      "mean_token_accuracy": 0.1789113014936447,
      "num_tokens": 55801215.0,
      "step": 25475
    },
    {
      "entropy": 5.66264796257019,
      "epoch": 2.5287812624057167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.1148,
      "mean_token_accuracy": 0.17947767823934554,
      "num_tokens": 55812663.0,
      "step": 25480
    },
    {
      "entropy": 5.5682474136352536,
      "epoch": 2.529277491067884,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043635510504357343,
      "loss": 4.9805,
      "mean_token_accuracy": 0.19083887040615083,
      "num_tokens": 55822337.0,
      "step": 25485
    },
    {
      "entropy": 5.624017763137817,
      "epoch": 2.5297737197300516,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043633040957505515,
      "loss": 5.111,
      "mean_token_accuracy": 0.17563399374485017,
      "num_tokens": 55833578.0,
      "step": 25490
    },
    {
      "entropy": 5.654432153701782,
      "epoch": 2.530269948392219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004363057101058903,
      "loss": 4.981,
      "mean_token_accuracy": 0.18899648487567902,
      "num_tokens": 55843768.0,
      "step": 25495
    },
    {
      "entropy": 5.657373714447021,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004362810066366915,
      "loss": 5.0331,
      "mean_token_accuracy": 0.17501474618911744,
      "num_tokens": 55853879.0,
      "step": 25500
    },
    {
      "entropy": 5.534429264068604,
      "epoch": 2.531262405716554,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1806861087679863,
      "num_tokens": 55864606.0,
      "step": 25505
    },
    {
      "entropy": 5.586514711380005,
      "epoch": 2.531758634378722,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043623158770064214,
      "loss": 4.9352,
      "mean_token_accuracy": 0.18963756710290908,
      "num_tokens": 55875036.0,
      "step": 25510
    },
    {
      "entropy": 5.599148797988891,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.0118,
      "mean_token_accuracy": 0.18061643540859224,
      "num_tokens": 55884924.0,
      "step": 25515
    },
    {
      "entropy": 5.593543958663941,
      "epoch": 2.532751091703057,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043618215277180917,
      "loss": 5.0481,
      "mean_token_accuracy": 0.18118804395198823,
      "num_tokens": 55894975.0,
      "step": 25520
    },
    {
      "entropy": 5.589440250396729,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.125,
      "learning_rate": 0.00043615742931163106,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18768583834171296,
      "num_tokens": 55905325.0,
      "step": 25525
    },
    {
      "entropy": 5.643001079559326,
      "epoch": 2.533743549027392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000436132701855096,
      "loss": 4.9967,
      "mean_token_accuracy": 0.18038791567087173,
      "num_tokens": 55916816.0,
      "step": 25530
    },
    {
      "entropy": 5.552133703231812,
      "epoch": 2.5342397776895593,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043610797040281705,
      "loss": 4.9786,
      "mean_token_accuracy": 0.18297821581363677,
      "num_tokens": 55928725.0,
      "step": 25535
    },
    {
      "entropy": 5.637594604492188,
      "epoch": 2.534736006351727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.0757,
      "mean_token_accuracy": 0.1878890261054039,
      "num_tokens": 55939555.0,
      "step": 25540
    },
    {
      "entropy": 5.595016956329346,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004360584955134811,
      "loss": 4.9216,
      "mean_token_accuracy": 0.18885112106800078,
      "num_tokens": 55949944.0,
      "step": 25545
    },
    {
      "entropy": 5.638816881179809,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.0189,
      "mean_token_accuracy": 0.18589355647563935,
      "num_tokens": 55959549.0,
      "step": 25550
    },
    {
      "entropy": 5.57937912940979,
      "epoch": 2.5362246923382292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.0212,
      "mean_token_accuracy": 0.18339101523160933,
      "num_tokens": 55970088.0,
      "step": 25555
    },
    {
      "entropy": 5.665594053268433,
      "epoch": 2.536720921000397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.0406,
      "mean_token_accuracy": 0.18636011928319932,
      "num_tokens": 55980485.0,
      "step": 25560
    },
    {
      "entropy": 5.634902572631836,
      "epoch": 2.5372171496625646,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18415289968252183,
      "num_tokens": 55991577.0,
      "step": 25565
    },
    {
      "entropy": 5.5908369541168215,
      "epoch": 2.537713378324732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043593473840756715,
      "loss": 4.9954,
      "mean_token_accuracy": 0.1853636160492897,
      "num_tokens": 56002813.0,
      "step": 25570
    },
    {
      "entropy": 5.713173294067383,
      "epoch": 2.5382096069868996,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17827597856521607,
      "num_tokens": 56013646.0,
      "step": 25575
    },
    {
      "entropy": 5.659455347061157,
      "epoch": 2.538705835649067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004358852076250613,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17671919614076614,
      "num_tokens": 56024777.0,
      "step": 25580
    },
    {
      "entropy": 5.651859617233276,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.078,
      "mean_token_accuracy": 0.17977611422538758,
      "num_tokens": 56035550.0,
      "step": 25585
    },
    {
      "entropy": 5.5611481189727785,
      "epoch": 2.539698292973402,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004358356608841296,
      "loss": 4.997,
      "mean_token_accuracy": 0.1834553897380829,
      "num_tokens": 56046835.0,
      "step": 25590
    },
    {
      "entropy": 5.540869235992432,
      "epoch": 2.54019452163557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043581088153078984,
      "loss": 5.0002,
      "mean_token_accuracy": 0.1852294236421585,
      "num_tokens": 56056404.0,
      "step": 25595
    },
    {
      "entropy": 5.64159197807312,
      "epoch": 2.5406907502977374,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.0565,
      "mean_token_accuracy": 0.17630048245191574,
      "num_tokens": 56066690.0,
      "step": 25600
    },
    {
      "entropy": 5.714221239089966,
      "epoch": 2.541186978959905,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.1201,
      "mean_token_accuracy": 0.17965082079172134,
      "num_tokens": 56076051.0,
      "step": 25605
    },
    {
      "entropy": 5.607185316085816,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043573651954664897,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18361276686191558,
      "num_tokens": 56086601.0,
      "step": 25610
    },
    {
      "entropy": 5.560170316696167,
      "epoch": 2.54217943628424,
      "grad_norm": 1.125,
      "learning_rate": 0.00043571172424594363,
      "loss": 5.0034,
      "mean_token_accuracy": 0.17875568866729735,
      "num_tokens": 56097954.0,
      "step": 25615
    },
    {
      "entropy": 5.526289367675782,
      "epoch": 2.5426756649464073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004356869249599339,
      "loss": 5.0094,
      "mean_token_accuracy": 0.1831750050187111,
      "num_tokens": 56109729.0,
      "step": 25620
    },
    {
      "entropy": 5.64527063369751,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18346219658851623,
      "num_tokens": 56120231.0,
      "step": 25625
    },
    {
      "entropy": 5.629915618896485,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043563731443446095,
      "loss": 5.0463,
      "mean_token_accuracy": 0.17191164791584015,
      "num_tokens": 56130558.0,
      "step": 25630
    },
    {
      "entropy": 5.714439916610718,
      "epoch": 2.5441643509329097,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.1428,
      "mean_token_accuracy": 0.17372167408466338,
      "num_tokens": 56142540.0,
      "step": 25635
    },
    {
      "entropy": 5.645591974258423,
      "epoch": 2.544660579595077,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.05,
      "mean_token_accuracy": 0.17706508785486222,
      "num_tokens": 56153802.0,
      "step": 25640
    },
    {
      "entropy": 5.682729959487915,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.0597,
      "mean_token_accuracy": 0.18631885647773744,
      "num_tokens": 56164990.0,
      "step": 25645
    },
    {
      "entropy": 5.613832473754883,
      "epoch": 2.5456530369194126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17728298306465148,
      "num_tokens": 56176472.0,
      "step": 25650
    },
    {
      "entropy": 5.556533527374268,
      "epoch": 2.54614926558158,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18052376806735992,
      "num_tokens": 56187504.0,
      "step": 25655
    },
    {
      "entropy": 5.591233110427856,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1753765895962715,
      "num_tokens": 56198214.0,
      "step": 25660
    },
    {
      "entropy": 5.669971656799317,
      "epoch": 2.547141722905915,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1851876676082611,
      "num_tokens": 56209885.0,
      "step": 25665
    },
    {
      "entropy": 5.591374635696411,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004354387130434332,
      "loss": 5.0295,
      "mean_token_accuracy": 0.17903601825237275,
      "num_tokens": 56221281.0,
      "step": 25670
    },
    {
      "entropy": 5.579173517227173,
      "epoch": 2.54813418023025,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18553335815668107,
      "num_tokens": 56231914.0,
      "step": 25675
    },
    {
      "entropy": 5.587691783905029,
      "epoch": 2.548630408892418,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004353890228980167,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18191301226615905,
      "num_tokens": 56242308.0,
      "step": 25680
    },
    {
      "entropy": 5.589905643463135,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004353641718590471,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18033185452222825,
      "num_tokens": 56252570.0,
      "step": 25685
    },
    {
      "entropy": 5.645611000061035,
      "epoch": 2.549622866216753,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.0411,
      "mean_token_accuracy": 0.18267372250556946,
      "num_tokens": 56263669.0,
      "step": 25690
    },
    {
      "entropy": 5.606331396102905,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004353144578516665,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17686076015233992,
      "num_tokens": 56273869.0,
      "step": 25695
    },
    {
      "entropy": 5.606118297576904,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043528959488448835,
      "loss": 5.0226,
      "mean_token_accuracy": 0.1863754063844681,
      "num_tokens": 56285670.0,
      "step": 25700
    },
    {
      "entropy": 5.70606107711792,
      "epoch": 2.5511115522032553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043526472794247365,
      "loss": 5.0645,
      "mean_token_accuracy": 0.1813206523656845,
      "num_tokens": 56296579.0,
      "step": 25705
    },
    {
      "entropy": 5.6660237312316895,
      "epoch": 2.5516077808654227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.0714,
      "mean_token_accuracy": 0.1807694524526596,
      "num_tokens": 56307879.0,
      "step": 25710
    },
    {
      "entropy": 5.622069644927978,
      "epoch": 2.55210400952759,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043521498213640113,
      "loss": 5.0312,
      "mean_token_accuracy": 0.17902485728263856,
      "num_tokens": 56318848.0,
      "step": 25715
    },
    {
      "entropy": 5.645676803588867,
      "epoch": 2.5526002381897577,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043519010327357704,
      "loss": 5.1023,
      "mean_token_accuracy": 0.18042174428701402,
      "num_tokens": 56332002.0,
      "step": 25720
    },
    {
      "entropy": 5.610456943511963,
      "epoch": 2.553096466851925,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.0964,
      "mean_token_accuracy": 0.18153021484613419,
      "num_tokens": 56342323.0,
      "step": 25725
    },
    {
      "entropy": 5.65478515625,
      "epoch": 2.5535926955140926,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004351403336314377,
      "loss": 5.0497,
      "mean_token_accuracy": 0.18038704991340637,
      "num_tokens": 56354925.0,
      "step": 25730
    },
    {
      "entropy": 5.605524492263794,
      "epoch": 2.5540889241762605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004351154428533566,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18093952238559724,
      "num_tokens": 56366987.0,
      "step": 25735
    },
    {
      "entropy": 5.554042148590088,
      "epoch": 2.554585152838428,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1852987140417099,
      "num_tokens": 56379496.0,
      "step": 25740
    },
    {
      "entropy": 5.551240396499634,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004350656493862579,
      "loss": 4.8863,
      "mean_token_accuracy": 0.19125194549560548,
      "num_tokens": 56391132.0,
      "step": 25745
    },
    {
      "entropy": 5.703465938568115,
      "epoch": 2.555577610162763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043504074669847517,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17080993056297303,
      "num_tokens": 56402687.0,
      "step": 25750
    },
    {
      "entropy": 5.556756591796875,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004350158400420266,
      "loss": 4.9632,
      "mean_token_accuracy": 0.18062463998794556,
      "num_tokens": 56414202.0,
      "step": 25755
    },
    {
      "entropy": 5.628629684448242,
      "epoch": 2.556570067487098,
      "grad_norm": 1.125,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.0573,
      "mean_token_accuracy": 0.18498581796884536,
      "num_tokens": 56424949.0,
      "step": 25760
    },
    {
      "entropy": 5.6714709281921385,
      "epoch": 2.557066296149266,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17234171628952027,
      "num_tokens": 56437185.0,
      "step": 25765
    },
    {
      "entropy": 5.651510238647461,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043494109626686394,
      "loss": 5.0341,
      "mean_token_accuracy": 0.17610188871622084,
      "num_tokens": 56447646.0,
      "step": 25770
    },
    {
      "entropy": 5.605867147445679,
      "epoch": 2.558058753473601,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.0648,
      "mean_token_accuracy": 0.18159355372190475,
      "num_tokens": 56458245.0,
      "step": 25775
    },
    {
      "entropy": 5.5957824230194095,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004348912472514197,
      "loss": 4.9807,
      "mean_token_accuracy": 0.18247201293706894,
      "num_tokens": 56470186.0,
      "step": 25780
    },
    {
      "entropy": 5.605624198913574,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043486631679595097,
      "loss": 5.082,
      "mean_token_accuracy": 0.18482763022184373,
      "num_tokens": 56481332.0,
      "step": 25785
    },
    {
      "entropy": 5.56342830657959,
      "epoch": 2.559547439460103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004348413823761421,
      "loss": 4.9653,
      "mean_token_accuracy": 0.18816055953502656,
      "num_tokens": 56491700.0,
      "step": 25790
    },
    {
      "entropy": 5.642283630371094,
      "epoch": 2.5600436681222707,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004348164439926112,
      "loss": 5.0445,
      "mean_token_accuracy": 0.18545080870389938,
      "num_tokens": 56502780.0,
      "step": 25795
    },
    {
      "entropy": 5.584514904022217,
      "epoch": 2.560539896784438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043479150164597695,
      "loss": 5.06,
      "mean_token_accuracy": 0.1804214298725128,
      "num_tokens": 56515039.0,
      "step": 25800
    },
    {
      "entropy": 5.612771654129029,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.1007,
      "mean_token_accuracy": 0.1764246106147766,
      "num_tokens": 56527705.0,
      "step": 25805
    },
    {
      "entropy": 5.581104898452759,
      "epoch": 2.561532354108773,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043474160506587234,
      "loss": 4.9782,
      "mean_token_accuracy": 0.1866317242383957,
      "num_tokens": 56537748.0,
      "step": 25810
    },
    {
      "entropy": 5.608246278762818,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004347166508336393,
      "loss": 4.9702,
      "mean_token_accuracy": 0.18130860924720765,
      "num_tokens": 56548804.0,
      "step": 25815
    },
    {
      "entropy": 5.622158098220825,
      "epoch": 2.5625248114331085,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043469169264077754,
      "loss": 5.037,
      "mean_token_accuracy": 0.17957046777009963,
      "num_tokens": 56561245.0,
      "step": 25820
    },
    {
      "entropy": 5.585957431793213,
      "epoch": 2.563021040095276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004346667304879058,
      "loss": 4.989,
      "mean_token_accuracy": 0.19081514179706574,
      "num_tokens": 56570823.0,
      "step": 25825
    },
    {
      "entropy": 5.57470555305481,
      "epoch": 2.5635172687574435,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.0101,
      "mean_token_accuracy": 0.18385200947523117,
      "num_tokens": 56581439.0,
      "step": 25830
    },
    {
      "entropy": 5.624047565460205,
      "epoch": 2.564013497419611,
      "grad_norm": 1.0,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18076393604278565,
      "num_tokens": 56593016.0,
      "step": 25835
    },
    {
      "entropy": 5.519311094284058,
      "epoch": 2.5645097260817784,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000434591820275422,
      "loss": 4.9438,
      "mean_token_accuracy": 0.18885384649038314,
      "num_tokens": 56603701.0,
      "step": 25840
    },
    {
      "entropy": 5.57692699432373,
      "epoch": 2.565005954743946,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.0075,
      "mean_token_accuracy": 0.1772146224975586,
      "num_tokens": 56614781.0,
      "step": 25845
    },
    {
      "entropy": 5.548316955566406,
      "epoch": 2.565502183406114,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004345418603450678,
      "loss": 4.9337,
      "mean_token_accuracy": 0.1856052905321121,
      "num_tokens": 56625477.0,
      "step": 25850
    },
    {
      "entropy": 5.551494455337524,
      "epoch": 2.5659984120682813,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345168744451395,
      "loss": 4.9861,
      "mean_token_accuracy": 0.1875208079814911,
      "num_tokens": 56635230.0,
      "step": 25855
    },
    {
      "entropy": 5.607312154769898,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004344918845895364,
      "loss": 4.9666,
      "mean_token_accuracy": 0.187286613881588,
      "num_tokens": 56645777.0,
      "step": 25860
    },
    {
      "entropy": 5.608931922912598,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18203712701797486,
      "num_tokens": 56656790.0,
      "step": 25865
    },
    {
      "entropy": 5.59069037437439,
      "epoch": 2.5674870980547837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004344418930137848,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18380700498819352,
      "num_tokens": 56668771.0,
      "step": 25870
    },
    {
      "entropy": 5.604913187026978,
      "epoch": 2.567983326716951,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.0325,
      "mean_token_accuracy": 0.17798234522342682,
      "num_tokens": 56680437.0,
      "step": 25875
    },
    {
      "entropy": 5.6567786693572994,
      "epoch": 2.5684795553791187,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.0334,
      "mean_token_accuracy": 0.18245895504951476,
      "num_tokens": 56691152.0,
      "step": 25880
    },
    {
      "entropy": 5.624453592300415,
      "epoch": 2.568975784041286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043436687599809226,
      "loss": 4.9999,
      "mean_token_accuracy": 0.1802006706595421,
      "num_tokens": 56701677.0,
      "step": 25885
    },
    {
      "entropy": 5.567034339904785,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043434186242145763,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19498291462659836,
      "num_tokens": 56712166.0,
      "step": 25890
    },
    {
      "entropy": 5.627406024932862,
      "epoch": 2.569968241365621,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004343168448934881,
      "loss": 5.02,
      "mean_token_accuracy": 0.18496530055999755,
      "num_tokens": 56722010.0,
      "step": 25895
    },
    {
      "entropy": 5.555558872222901,
      "epoch": 2.5704644700277886,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043429182341480423,
      "loss": 4.9928,
      "mean_token_accuracy": 0.18366943299770355,
      "num_tokens": 56732368.0,
      "step": 25900
    },
    {
      "entropy": 5.566744041442871,
      "epoch": 2.5709606986899565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004342667979860263,
      "loss": 5.101,
      "mean_token_accuracy": 0.17804741710424424,
      "num_tokens": 56745084.0,
      "step": 25905
    },
    {
      "entropy": 5.619599342346191,
      "epoch": 2.571456927352124,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043424176860777495,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17857166677713393,
      "num_tokens": 56756184.0,
      "step": 25910
    },
    {
      "entropy": 5.597119474411011,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043421673528067095,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18878730833530427,
      "num_tokens": 56767505.0,
      "step": 25915
    },
    {
      "entropy": 5.573809099197388,
      "epoch": 2.572449384676459,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.0068,
      "mean_token_accuracy": 0.17803461402654647,
      "num_tokens": 56778660.0,
      "step": 25920
    },
    {
      "entropy": 5.556943702697754,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.0413,
      "mean_token_accuracy": 0.1809852108359337,
      "num_tokens": 56787892.0,
      "step": 25925
    },
    {
      "entropy": 5.610515689849853,
      "epoch": 2.573441842000794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.08,
      "mean_token_accuracy": 0.17819587737321854,
      "num_tokens": 56797910.0,
      "step": 25930
    },
    {
      "entropy": 5.628737878799439,
      "epoch": 2.5739380706629618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17833257764577864,
      "num_tokens": 56807975.0,
      "step": 25935
    },
    {
      "entropy": 5.605467796325684,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043409150943409035,
      "loss": 4.9748,
      "mean_token_accuracy": 0.18037139624357224,
      "num_tokens": 56817540.0,
      "step": 25940
    },
    {
      "entropy": 5.6450193405151365,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043406645242690967,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18671874701976776,
      "num_tokens": 56828917.0,
      "step": 25945
    },
    {
      "entropy": 5.556448841094971,
      "epoch": 2.575426756649464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004340413914752237,
      "loss": 4.9467,
      "mean_token_accuracy": 0.18598034828901291,
      "num_tokens": 56839888.0,
      "step": 25950
    },
    {
      "entropy": 5.522010326385498,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004340163265796539,
      "loss": 4.9698,
      "mean_token_accuracy": 0.1780501574277878,
      "num_tokens": 56850191.0,
      "step": 25955
    },
    {
      "entropy": 5.560699558258056,
      "epoch": 2.576419213973799,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004339912577408218,
      "loss": 5.0094,
      "mean_token_accuracy": 0.18558891415596007,
      "num_tokens": 56861020.0,
      "step": 25960
    },
    {
      "entropy": 5.550577592849732,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004339661849593492,
      "loss": 4.9596,
      "mean_token_accuracy": 0.18667820394039153,
      "num_tokens": 56871184.0,
      "step": 25965
    },
    {
      "entropy": 5.671886920928955,
      "epoch": 2.577411671298134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.1385,
      "mean_token_accuracy": 0.1791170582175255,
      "num_tokens": 56882081.0,
      "step": 25970
    },
    {
      "entropy": 5.721893930435181,
      "epoch": 2.5779078999603016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17666661590337754,
      "num_tokens": 56891594.0,
      "step": 25975
    },
    {
      "entropy": 5.656576299667359,
      "epoch": 2.578404128622469,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.0915,
      "mean_token_accuracy": 0.1855011686682701,
      "num_tokens": 56903636.0,
      "step": 25980
    },
    {
      "entropy": 5.527581024169922,
      "epoch": 2.5789003572846365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.0568,
      "mean_token_accuracy": 0.18167801648378373,
      "num_tokens": 56915776.0,
      "step": 25985
    },
    {
      "entropy": 5.635126066207886,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17771140038967131,
      "num_tokens": 56926851.0,
      "step": 25990
    },
    {
      "entropy": 5.641993236541748,
      "epoch": 2.579892814608972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.0626,
      "mean_token_accuracy": 0.1788672611117363,
      "num_tokens": 56938373.0,
      "step": 25995
    },
    {
      "entropy": 5.658073043823242,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043379056514734334,
      "loss": 5.0426,
      "mean_token_accuracy": 0.1778601199388504,
      "num_tokens": 56948891.0,
      "step": 26000
    },
    {
      "entropy": 5.573230886459351,
      "epoch": 2.580885271933307,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004337654608471372,
      "loss": 4.9798,
      "mean_token_accuracy": 0.17995180785655976,
      "num_tokens": 56961091.0,
      "step": 26005
    },
    {
      "entropy": 5.6079388618469235,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043374035260988964,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17510178536176682,
      "num_tokens": 56972073.0,
      "step": 26010
    },
    {
      "entropy": 5.516574954986572,
      "epoch": 2.581877729257642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043371524043622344,
      "loss": 4.9393,
      "mean_token_accuracy": 0.1882518410682678,
      "num_tokens": 56981591.0,
      "step": 26015
    },
    {
      "entropy": 5.593730497360229,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004336901243267614,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18579913228750228,
      "num_tokens": 56992045.0,
      "step": 26020
    },
    {
      "entropy": 5.623245668411255,
      "epoch": 2.582870186581977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000433665004282126,
      "loss": 5.0391,
      "mean_token_accuracy": 0.18117234259843826,
      "num_tokens": 57003108.0,
      "step": 26025
    },
    {
      "entropy": 5.572370767593384,
      "epoch": 2.5833664152441447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043363988030294055,
      "loss": 4.9502,
      "mean_token_accuracy": 0.18478886485099794,
      "num_tokens": 57014446.0,
      "step": 26030
    },
    {
      "entropy": 5.64778413772583,
      "epoch": 2.583862643906312,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004336147523898279,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18364280164241792,
      "num_tokens": 57025758.0,
      "step": 26035
    },
    {
      "entropy": 5.574322938919067,
      "epoch": 2.5843588725684796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004335896205434111,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18470312356948854,
      "num_tokens": 57036060.0,
      "step": 26040
    },
    {
      "entropy": 5.577623462677002,
      "epoch": 2.584855101230647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043356448476431354,
      "loss": 5.0163,
      "mean_token_accuracy": 0.1747617170214653,
      "num_tokens": 57045567.0,
      "step": 26045
    },
    {
      "entropy": 5.512379789352417,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335393450531583,
      "loss": 4.9656,
      "mean_token_accuracy": 0.19710001051425935,
      "num_tokens": 57056942.0,
      "step": 26050
    },
    {
      "entropy": 5.565903282165527,
      "epoch": 2.585847558554982,
      "grad_norm": 1.0625,
      "learning_rate": 0.000433514201410569,
      "loss": 4.9804,
      "mean_token_accuracy": 0.18670771569013594,
      "num_tokens": 57068651.0,
      "step": 26055
    },
    {
      "entropy": 5.708491230010987,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043348905383716905,
      "loss": 5.1338,
      "mean_token_accuracy": 0.17601816058158876,
      "num_tokens": 57080503.0,
      "step": 26060
    },
    {
      "entropy": 5.601133537292481,
      "epoch": 2.586840015879317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043346390233358213,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18528538197278976,
      "num_tokens": 57090807.0,
      "step": 26065
    },
    {
      "entropy": 5.578626585006714,
      "epoch": 2.5873362445414845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.0434,
      "mean_token_accuracy": 0.17817013710737228,
      "num_tokens": 57102636.0,
      "step": 26070
    },
    {
      "entropy": 5.639294958114624,
      "epoch": 2.5878324732036524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18356675058603286,
      "num_tokens": 57113124.0,
      "step": 26075
    },
    {
      "entropy": 5.594115781784057,
      "epoch": 2.58832870186582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17817768901586534,
      "num_tokens": 57124286.0,
      "step": 26080
    },
    {
      "entropy": 5.560608816146851,
      "epoch": 2.5888249305279873,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043336325702983966,
      "loss": 5.0206,
      "mean_token_accuracy": 0.1791694462299347,
      "num_tokens": 57137024.0,
      "step": 26085
    },
    {
      "entropy": 5.6324492454528805,
      "epoch": 2.589321159190155,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043333808588467504,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18269590735435487,
      "num_tokens": 57147720.0,
      "step": 26090
    },
    {
      "entropy": 5.55548357963562,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004333129108130671,
      "loss": 4.9519,
      "mean_token_accuracy": 0.19013818651437758,
      "num_tokens": 57159156.0,
      "step": 26095
    },
    {
      "entropy": 5.6417499542236325,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004332877318156402,
      "loss": 5.0906,
      "mean_token_accuracy": 0.18248683959245682,
      "num_tokens": 57170737.0,
      "step": 26100
    },
    {
      "entropy": 5.578612804412842,
      "epoch": 2.5908098451766572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.0089,
      "mean_token_accuracy": 0.18346001505851744,
      "num_tokens": 57181548.0,
      "step": 26105
    },
    {
      "entropy": 5.644182348251343,
      "epoch": 2.591306073838825,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17601045817136765,
      "num_tokens": 57192661.0,
      "step": 26110
    },
    {
      "entropy": 5.647821855545044,
      "epoch": 2.5918023025009926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18457082509994507,
      "num_tokens": 57205108.0,
      "step": 26115
    },
    {
      "entropy": 5.6872203826904295,
      "epoch": 2.59229853116316,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043318697658023097,
      "loss": 5.1013,
      "mean_token_accuracy": 0.17669563740491867,
      "num_tokens": 57215995.0,
      "step": 26120
    },
    {
      "entropy": 5.589784479141235,
      "epoch": 2.5927947598253276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004331617779630765,
      "loss": 4.9742,
      "mean_token_accuracy": 0.18809205144643784,
      "num_tokens": 57227008.0,
      "step": 26125
    },
    {
      "entropy": 5.591264486312866,
      "epoch": 2.593290988487495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331365754238507,
      "loss": 4.9766,
      "mean_token_accuracy": 0.17976797819137574,
      "num_tokens": 57238279.0,
      "step": 26130
    },
    {
      "entropy": 5.594855546951294,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043311136896317855,
      "loss": 4.998,
      "mean_token_accuracy": 0.1806861564517021,
      "num_tokens": 57247811.0,
      "step": 26135
    },
    {
      "entropy": 5.641523218154907,
      "epoch": 2.59428344581183,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043308615858168517,
      "loss": 4.973,
      "mean_token_accuracy": 0.1856754392385483,
      "num_tokens": 57256850.0,
      "step": 26140
    },
    {
      "entropy": 5.5401393413543705,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.0065,
      "mean_token_accuracy": 0.18839261680841446,
      "num_tokens": 57267079.0,
      "step": 26145
    },
    {
      "entropy": 5.565443229675293,
      "epoch": 2.595275903136165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18190462291240692,
      "num_tokens": 57279122.0,
      "step": 26150
    },
    {
      "entropy": 5.6494897365570065,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004330105039185297,
      "loss": 5.0329,
      "mean_token_accuracy": 0.17748506367206573,
      "num_tokens": 57289601.0,
      "step": 26155
    },
    {
      "entropy": 5.669852781295776,
      "epoch": 2.5962683604605004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043298527786000393,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17895803302526475,
      "num_tokens": 57299942.0,
      "step": 26160
    },
    {
      "entropy": 5.67518572807312,
      "epoch": 2.596764589122668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.0766,
      "mean_token_accuracy": 0.18068482279777526,
      "num_tokens": 57310983.0,
      "step": 26165
    },
    {
      "entropy": 5.597938585281372,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043293481399049474,
      "loss": 4.9897,
      "mean_token_accuracy": 0.1786714404821396,
      "num_tokens": 57321724.0,
      "step": 26170
    },
    {
      "entropy": 5.613726663589477,
      "epoch": 2.597757046447003,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.023,
      "mean_token_accuracy": 0.17731043100357055,
      "num_tokens": 57332222.0,
      "step": 26175
    },
    {
      "entropy": 5.598468923568726,
      "epoch": 2.5982532751091703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328843344552133,
      "loss": 5.017,
      "mean_token_accuracy": 0.18567130267620086,
      "num_tokens": 57342956.0,
      "step": 26180
    },
    {
      "entropy": 5.5783158302307125,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328590888144726,
      "loss": 4.9977,
      "mean_token_accuracy": 0.1841711089015007,
      "num_tokens": 57355328.0,
      "step": 26185
    },
    {
      "entropy": 5.586469411849976,
      "epoch": 2.599245732433505,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.0257,
      "mean_token_accuracy": 0.18617238253355026,
      "num_tokens": 57366533.0,
      "step": 26190
    },
    {
      "entropy": 5.705792093276978,
      "epoch": 2.599741961095673,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043280858578992185,
      "loss": 5.1176,
      "mean_token_accuracy": 0.1728084772825241,
      "num_tokens": 57377150.0,
      "step": 26195
    },
    {
      "entropy": 5.645489168167114,
      "epoch": 2.6002381897578406,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.0127,
      "mean_token_accuracy": 0.18223857879638672,
      "num_tokens": 57388372.0,
      "step": 26200
    },
    {
      "entropy": 5.620415782928466,
      "epoch": 2.600734418420008,
      "grad_norm": 1.125,
      "learning_rate": 0.00043275806711211957,
      "loss": 5.0069,
      "mean_token_accuracy": 0.17691659033298493,
      "num_tokens": 57399613.0,
      "step": 26205
    },
    {
      "entropy": 5.566832399368286,
      "epoch": 2.6012306470821756,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043273280190481503,
      "loss": 5.0069,
      "mean_token_accuracy": 0.1855458065867424,
      "num_tokens": 57410117.0,
      "step": 26210
    },
    {
      "entropy": 5.613982248306274,
      "epoch": 2.601726875744343,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043270753278607687,
      "loss": 5.0,
      "mean_token_accuracy": 0.18384812474250795,
      "num_tokens": 57421511.0,
      "step": 26215
    },
    {
      "entropy": 5.649285888671875,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.1058,
      "mean_token_accuracy": 0.17595717012882234,
      "num_tokens": 57432785.0,
      "step": 26220
    },
    {
      "entropy": 5.620855665206909,
      "epoch": 2.602719333068678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.0687,
      "mean_token_accuracy": 0.17673179805278777,
      "num_tokens": 57443394.0,
      "step": 26225
    },
    {
      "entropy": 5.658952617645264,
      "epoch": 2.6032155617308455,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004326317019675275,
      "loss": 5.076,
      "mean_token_accuracy": 0.1755487307906151,
      "num_tokens": 57454245.0,
      "step": 26230
    },
    {
      "entropy": 5.600081443786621,
      "epoch": 2.603711790393013,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043260641720932214,
      "loss": 4.9728,
      "mean_token_accuracy": 0.1850302502512932,
      "num_tokens": 57465954.0,
      "step": 26235
    },
    {
      "entropy": 5.621303081512451,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043258112854281726,
      "loss": 5.0429,
      "mean_token_accuracy": 0.17870576530694962,
      "num_tokens": 57477599.0,
      "step": 26240
    },
    {
      "entropy": 5.6111713409423825,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004325558359686399,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18530425876379014,
      "num_tokens": 57489043.0,
      "step": 26245
    },
    {
      "entropy": 5.5582231998443605,
      "epoch": 2.605200476379516,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043253053948741734,
      "loss": 4.9694,
      "mean_token_accuracy": 0.18969885557889937,
      "num_tokens": 57499349.0,
      "step": 26250
    },
    {
      "entropy": 5.5517415523529055,
      "epoch": 2.6056967050416833,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.0381,
      "mean_token_accuracy": 0.18284438252449037,
      "num_tokens": 57510873.0,
      "step": 26255
    },
    {
      "entropy": 5.589234447479248,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.0097,
      "mean_token_accuracy": 0.17909045964479448,
      "num_tokens": 57522282.0,
      "step": 26260
    },
    {
      "entropy": 5.639321231842041,
      "epoch": 2.606689162366018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004324546266077518,
      "loss": 5.0057,
      "mean_token_accuracy": 0.17993039786815643,
      "num_tokens": 57532760.0,
      "step": 26265
    },
    {
      "entropy": 5.6247094631195065,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004324293145046222,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18536062836647033,
      "num_tokens": 57543373.0,
      "step": 26270
    },
    {
      "entropy": 5.56131911277771,
      "epoch": 2.607681619690353,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043240399849758484,
      "loss": 4.9702,
      "mean_token_accuracy": 0.18388832062482835,
      "num_tokens": 57553221.0,
      "step": 26275
    },
    {
      "entropy": 5.61955075263977,
      "epoch": 2.608177848352521,
      "grad_norm": 1.125,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.0237,
      "mean_token_accuracy": 0.18032329976558686,
      "num_tokens": 57563080.0,
      "step": 26280
    },
    {
      "entropy": 5.627870512008667,
      "epoch": 2.6086740770146886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.137,
      "mean_token_accuracy": 0.177892105281353,
      "num_tokens": 57573556.0,
      "step": 26285
    },
    {
      "entropy": 5.671854686737061,
      "epoch": 2.609170305676856,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.0897,
      "mean_token_accuracy": 0.17940495908260345,
      "num_tokens": 57584750.0,
      "step": 26290
    },
    {
      "entropy": 5.673301362991333,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043230269544291503,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18537402302026748,
      "num_tokens": 57594975.0,
      "step": 26295
    },
    {
      "entropy": 5.645266723632813,
      "epoch": 2.610162763001191,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004322773599257575,
      "loss": 5.09,
      "mean_token_accuracy": 0.1784910351037979,
      "num_tokens": 57606534.0,
      "step": 26300
    },
    {
      "entropy": 5.575696659088135,
      "epoch": 2.6106589916633585,
      "grad_norm": 1.125,
      "learning_rate": 0.00043225202050846043,
      "loss": 4.9847,
      "mean_token_accuracy": 0.18786044418811798,
      "num_tokens": 57616715.0,
      "step": 26305
    },
    {
      "entropy": 5.640942001342774,
      "epoch": 2.611155220325526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432226677191652,
      "loss": 5.0057,
      "mean_token_accuracy": 0.19004617929458617,
      "num_tokens": 57626658.0,
      "step": 26310
    },
    {
      "entropy": 5.620627307891846,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043220132997596084,
      "loss": 5.037,
      "mean_token_accuracy": 0.17913638055324554,
      "num_tokens": 57635620.0,
      "step": 26315
    },
    {
      "entropy": 5.609261512756348,
      "epoch": 2.612147677649861,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18030306994915007,
      "num_tokens": 57646584.0,
      "step": 26320
    },
    {
      "entropy": 5.619351577758789,
      "epoch": 2.6126439063120284,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.0253,
      "mean_token_accuracy": 0.18125885725021362,
      "num_tokens": 57656560.0,
      "step": 26325
    },
    {
      "entropy": 5.606994533538819,
      "epoch": 2.613140134974196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043212526494187665,
      "loss": 5.0535,
      "mean_token_accuracy": 0.178919218480587,
      "num_tokens": 57667555.0,
      "step": 26330
    },
    {
      "entropy": 5.610515880584717,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.0,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.0138,
      "mean_token_accuracy": 0.1818953737616539,
      "num_tokens": 57678106.0,
      "step": 26335
    },
    {
      "entropy": 5.569538736343384,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043207453543626587,
      "loss": 4.9521,
      "mean_token_accuracy": 0.18798024952411652,
      "num_tokens": 57688429.0,
      "step": 26340
    },
    {
      "entropy": 5.600183153152466,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.0,
      "learning_rate": 0.00043204916484048094,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18202543258666992,
      "num_tokens": 57699179.0,
      "step": 26345
    },
    {
      "entropy": 5.552109956741333,
      "epoch": 2.615125049622866,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.0431,
      "mean_token_accuracy": 0.17939110696315766,
      "num_tokens": 57710472.0,
      "step": 26350
    },
    {
      "entropy": 5.613166952133179,
      "epoch": 2.6156212782850337,
      "grad_norm": 1.0,
      "learning_rate": 0.0004319984119660977,
      "loss": 5.0678,
      "mean_token_accuracy": 0.17839189171791076,
      "num_tokens": 57723063.0,
      "step": 26355
    },
    {
      "entropy": 5.693273210525513,
      "epoch": 2.616117506947201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043197302968875795,
      "loss": 5.1692,
      "mean_token_accuracy": 0.17989958226680755,
      "num_tokens": 57734664.0,
      "step": 26360
    },
    {
      "entropy": 5.678853607177734,
      "epoch": 2.616613735609369,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.0535,
      "mean_token_accuracy": 0.1841815561056137,
      "num_tokens": 57745081.0,
      "step": 26365
    },
    {
      "entropy": 5.587251949310303,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043192225345692933,
      "loss": 5.0379,
      "mean_token_accuracy": 0.18054172098636628,
      "num_tokens": 57755074.0,
      "step": 26370
    },
    {
      "entropy": 5.6145562648773195,
      "epoch": 2.617606192933704,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043189685950369963,
      "loss": 5.0902,
      "mean_token_accuracy": 0.17682381868362426,
      "num_tokens": 57765503.0,
      "step": 26375
    },
    {
      "entropy": 5.686745357513428,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17767210453748702,
      "num_tokens": 57776905.0,
      "step": 26380
    },
    {
      "entropy": 5.663685512542725,
      "epoch": 2.618598650258039,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004318460599257578,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1845008462667465,
      "num_tokens": 57786720.0,
      "step": 26385
    },
    {
      "entropy": 5.635724449157715,
      "epoch": 2.6190948789202064,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043182065430230545,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1855747938156128,
      "num_tokens": 57797171.0,
      "step": 26390
    },
    {
      "entropy": 5.620061302185059,
      "epoch": 2.619591107582374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004317952447900388,
      "loss": 4.9783,
      "mean_token_accuracy": 0.18650651574134827,
      "num_tokens": 57807543.0,
      "step": 26395
    },
    {
      "entropy": 5.625080966949463,
      "epoch": 2.6200873362445414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000431769831389588,
      "loss": 5.0591,
      "mean_token_accuracy": 0.1769583284854889,
      "num_tokens": 57818136.0,
      "step": 26400
    },
    {
      "entropy": 5.580299663543701,
      "epoch": 2.620583564906709,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004317444141015831,
      "loss": 4.9783,
      "mean_token_accuracy": 0.1800295129418373,
      "num_tokens": 57830159.0,
      "step": 26405
    },
    {
      "entropy": 5.652906990051269,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.0987,
      "mean_token_accuracy": 0.1782780334353447,
      "num_tokens": 57841867.0,
      "step": 26410
    },
    {
      "entropy": 5.626113080978394,
      "epoch": 2.621576022231044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004316935678654327,
      "loss": 4.9839,
      "mean_token_accuracy": 0.18490560948848725,
      "num_tokens": 57853982.0,
      "step": 26415
    },
    {
      "entropy": 5.579687452316284,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043166813891854796,
      "loss": 4.9551,
      "mean_token_accuracy": 0.1868773877620697,
      "num_tokens": 57866005.0,
      "step": 26420
    },
    {
      "entropy": 5.561060571670533,
      "epoch": 2.622568479555379,
      "grad_norm": 1.015625,
      "learning_rate": 0.000431642706086631,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18355527371168137,
      "num_tokens": 57876732.0,
      "step": 26425
    },
    {
      "entropy": 5.595856809616089,
      "epoch": 2.6230647082175467,
      "grad_norm": 1.0,
      "learning_rate": 0.00043161726937031243,
      "loss": 4.9722,
      "mean_token_accuracy": 0.18919639140367508,
      "num_tokens": 57887178.0,
      "step": 26430
    },
    {
      "entropy": 5.6082500457763675,
      "epoch": 2.623560936879714,
      "grad_norm": 1.0625,
      "learning_rate": 0.000431591828770223,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1795947328209877,
      "num_tokens": 57898383.0,
      "step": 26435
    },
    {
      "entropy": 5.623133897781372,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17612971663475036,
      "num_tokens": 57909009.0,
      "step": 26440
    },
    {
      "entropy": 5.653026866912842,
      "epoch": 2.624553394204049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.1368,
      "mean_token_accuracy": 0.18122730404138565,
      "num_tokens": 57920814.0,
      "step": 26445
    },
    {
      "entropy": 5.644754791259766,
      "epoch": 2.625049622866217,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043151548367363943,
      "loss": 4.9872,
      "mean_token_accuracy": 0.18364538103342057,
      "num_tokens": 57932207.0,
      "step": 26450
    },
    {
      "entropy": 5.588458490371704,
      "epoch": 2.6255458515283845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043149002754477655,
      "loss": 5.0173,
      "mean_token_accuracy": 0.18231064677238465,
      "num_tokens": 57944663.0,
      "step": 26455
    },
    {
      "entropy": 5.647845363616943,
      "epoch": 2.626042080190552,
      "grad_norm": 1.0,
      "learning_rate": 0.0004314645675352982,
      "loss": 5.0884,
      "mean_token_accuracy": 0.17776259481906892,
      "num_tokens": 57956213.0,
      "step": 26460
    },
    {
      "entropy": 5.712221002578735,
      "epoch": 2.6265383088527194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.1545,
      "mean_token_accuracy": 0.17851427048444748,
      "num_tokens": 57967149.0,
      "step": 26465
    },
    {
      "entropy": 5.678065204620362,
      "epoch": 2.627034537514887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043141363587702054,
      "loss": 5.0362,
      "mean_token_accuracy": 0.1785648509860039,
      "num_tokens": 57978358.0,
      "step": 26470
    },
    {
      "entropy": 5.56087646484375,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004313881642294843,
      "loss": 4.965,
      "mean_token_accuracy": 0.1889139384031296,
      "num_tokens": 57990001.0,
      "step": 26475
    },
    {
      "entropy": 5.6044820785522464,
      "epoch": 2.628026994839222,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004313626887038585,
      "loss": 5.087,
      "mean_token_accuracy": 0.18412183076143265,
      "num_tokens": 58003140.0,
      "step": 26480
    },
    {
      "entropy": 5.627261638641357,
      "epoch": 2.6285232235013893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000431337209300775,
      "loss": 5.0568,
      "mean_token_accuracy": 0.18452318757772446,
      "num_tokens": 58014303.0,
      "step": 26485
    },
    {
      "entropy": 5.653543186187744,
      "epoch": 2.629019452163557,
      "grad_norm": 1.0,
      "learning_rate": 0.0004313117260208656,
      "loss": 4.99,
      "mean_token_accuracy": 0.19125299900770187,
      "num_tokens": 58024704.0,
      "step": 26490
    },
    {
      "entropy": 5.5969812870025635,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004312862388647622,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18354056477546693,
      "num_tokens": 58036587.0,
      "step": 26495
    },
    {
      "entropy": 5.541950654983521,
      "epoch": 2.6300119094878918,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043126074783309687,
      "loss": 4.9341,
      "mean_token_accuracy": 0.19045152366161347,
      "num_tokens": 58047106.0,
      "step": 26500
    },
    {
      "entropy": 5.549757099151611,
      "epoch": 2.6305081381500597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.0185,
      "mean_token_accuracy": 0.18193631917238234,
      "num_tokens": 58058227.0,
      "step": 26505
    },
    {
      "entropy": 5.685854959487915,
      "epoch": 2.631004366812227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.0721,
      "mean_token_accuracy": 0.1767584949731827,
      "num_tokens": 58069443.0,
      "step": 26510
    },
    {
      "entropy": 5.652746343612671,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.02,
      "mean_token_accuracy": 0.18524955660104753,
      "num_tokens": 58079925.0,
      "step": 26515
    },
    {
      "entropy": 5.60659646987915,
      "epoch": 2.631996824136562,
      "grad_norm": 1.0,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.0275,
      "mean_token_accuracy": 0.18399675637483598,
      "num_tokens": 58090974.0,
      "step": 26520
    },
    {
      "entropy": 5.6155946254730225,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1822713077068329,
      "num_tokens": 58103548.0,
      "step": 26525
    },
    {
      "entropy": 5.636002874374389,
      "epoch": 2.632989281460897,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043110772029170916,
      "loss": 5.0038,
      "mean_token_accuracy": 0.1851767420768738,
      "num_tokens": 58113777.0,
      "step": 26530
    },
    {
      "entropy": 5.601940298080445,
      "epoch": 2.633485510123065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043108220214881494,
      "loss": 4.9679,
      "mean_token_accuracy": 0.17949339598417283,
      "num_tokens": 58125230.0,
      "step": 26535
    },
    {
      "entropy": 5.611791467666626,
      "epoch": 2.6339817387852325,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1804782286286354,
      "num_tokens": 58137572.0,
      "step": 26540
    },
    {
      "entropy": 5.670458364486694,
      "epoch": 2.6344779674474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043103115425215246,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17577518671751022,
      "num_tokens": 58148207.0,
      "step": 26545
    },
    {
      "entropy": 5.659106731414795,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043100562449965016,
      "loss": 5.0645,
      "mean_token_accuracy": 0.18066994845867157,
      "num_tokens": 58159390.0,
      "step": 26550
    },
    {
      "entropy": 5.61831750869751,
      "epoch": 2.635470424771735,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18051901161670686,
      "num_tokens": 58170842.0,
      "step": 26555
    },
    {
      "entropy": 5.636032152175903,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17912397682666778,
      "num_tokens": 58181507.0,
      "step": 26560
    },
    {
      "entropy": 5.6037122249603275,
      "epoch": 2.63646288209607,
      "grad_norm": 1.046875,
      "learning_rate": 0.000430929012033056,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18574946969747544,
      "num_tokens": 58192515.0,
      "step": 26565
    },
    {
      "entropy": 5.641292572021484,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043090346680993983,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18481610864400863,
      "num_tokens": 58202661.0,
      "step": 26570
    },
    {
      "entropy": 5.703846836090088,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.1537,
      "mean_token_accuracy": 0.1794850394129753,
      "num_tokens": 58213982.0,
      "step": 26575
    },
    {
      "entropy": 5.723297882080078,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004308523647661309,
      "loss": 5.0723,
      "mean_token_accuracy": 0.18446995168924332,
      "num_tokens": 58225386.0,
      "step": 26580
    },
    {
      "entropy": 5.620122718811035,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043082680794670535,
      "loss": 5.0453,
      "mean_token_accuracy": 0.18207414597272872,
      "num_tokens": 58236866.0,
      "step": 26585
    },
    {
      "entropy": 5.564018583297729,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043080124726311084,
      "loss": 4.9954,
      "mean_token_accuracy": 0.18615613728761674,
      "num_tokens": 58248054.0,
      "step": 26590
    },
    {
      "entropy": 5.605708646774292,
      "epoch": 2.639440254069075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.0669,
      "mean_token_accuracy": 0.1825596421957016,
      "num_tokens": 58259495.0,
      "step": 26595
    },
    {
      "entropy": 5.534410762786865,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.0254,
      "mean_token_accuracy": 0.18302974104881287,
      "num_tokens": 58270766.0,
      "step": 26600
    },
    {
      "entropy": 5.692231512069702,
      "epoch": 2.64043271139341,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043072454203365175,
      "loss": 5.1355,
      "mean_token_accuracy": 0.18668349981307983,
      "num_tokens": 58281736.0,
      "step": 26605
    },
    {
      "entropy": 5.649362373352051,
      "epoch": 2.6409289400555775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004306989658997204,
      "loss": 5.0343,
      "mean_token_accuracy": 0.1891157314181328,
      "num_tokens": 58292663.0,
      "step": 26610
    },
    {
      "entropy": 5.649743747711182,
      "epoch": 2.641425168717745,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.0239,
      "mean_token_accuracy": 0.19017452895641326,
      "num_tokens": 58303327.0,
      "step": 26615
    },
    {
      "entropy": 5.6494996547698975,
      "epoch": 2.641921397379913,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.0563,
      "mean_token_accuracy": 0.18400470465421676,
      "num_tokens": 58313109.0,
      "step": 26620
    },
    {
      "entropy": 5.675230169296265,
      "epoch": 2.6424176260420804,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004306222143344706,
      "loss": 5.037,
      "mean_token_accuracy": 0.18764886558055877,
      "num_tokens": 58323587.0,
      "step": 26625
    },
    {
      "entropy": 5.599175691604614,
      "epoch": 2.642913854704248,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004305966227603503,
      "loss": 4.9953,
      "mean_token_accuracy": 0.1847551643848419,
      "num_tokens": 58333590.0,
      "step": 26630
    },
    {
      "entropy": 5.689720296859742,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.0713,
      "mean_token_accuracy": 0.17929439544677733,
      "num_tokens": 58346127.0,
      "step": 26635
    },
    {
      "entropy": 5.669593143463135,
      "epoch": 2.643906312028583,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043054542803736126,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18141761869192125,
      "num_tokens": 58355769.0,
      "step": 26640
    },
    {
      "entropy": 5.723526859283448,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.1225,
      "mean_token_accuracy": 0.18403030037879944,
      "num_tokens": 58366932.0,
      "step": 26645
    },
    {
      "entropy": 5.619926643371582,
      "epoch": 2.644898769352918,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004304942178856066,
      "loss": 5.0097,
      "mean_token_accuracy": 0.17940415889024736,
      "num_tokens": 58378198.0,
      "step": 26650
    },
    {
      "entropy": 5.626012897491455,
      "epoch": 2.6453949980150853,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043046860702552944,
      "loss": 4.9132,
      "mean_token_accuracy": 0.19626829028129578,
      "num_tokens": 58388627.0,
      "step": 26655
    },
    {
      "entropy": 5.609719324111938,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004304429923101658,
      "loss": 4.9927,
      "mean_token_accuracy": 0.19044185131788255,
      "num_tokens": 58400481.0,
      "step": 26660
    },
    {
      "entropy": 5.6430117130279545,
      "epoch": 2.64638745533942,
      "grad_norm": 1.65625,
      "learning_rate": 0.000430417373740151,
      "loss": 5.063,
      "mean_token_accuracy": 0.17881338000297547,
      "num_tokens": 58410827.0,
      "step": 26665
    },
    {
      "entropy": 5.62346510887146,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004303917513161201,
      "loss": 5.1022,
      "mean_token_accuracy": 0.17461609542369844,
      "num_tokens": 58422632.0,
      "step": 26670
    },
    {
      "entropy": 5.650367593765258,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.0957,
      "mean_token_accuracy": 0.1788036659359932,
      "num_tokens": 58433770.0,
      "step": 26675
    },
    {
      "entropy": 5.605927467346191,
      "epoch": 2.647876141325923,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004303404949085522,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18748779445886612,
      "num_tokens": 58442999.0,
      "step": 26680
    },
    {
      "entropy": 5.638716316223144,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004303148609262861,
      "loss": 5.0873,
      "mean_token_accuracy": 0.17708034068346024,
      "num_tokens": 58454633.0,
      "step": 26685
    },
    {
      "entropy": 5.633494281768799,
      "epoch": 2.648868598650258,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1749859184026718,
      "num_tokens": 58464949.0,
      "step": 26690
    },
    {
      "entropy": 5.68959608078003,
      "epoch": 2.6493648273124255,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1831283912062645,
      "num_tokens": 58475915.0,
      "step": 26695
    },
    {
      "entropy": 5.643813610076904,
      "epoch": 2.649861055974593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004302379358731878,
      "loss": 5.0212,
      "mean_token_accuracy": 0.18441510647535325,
      "num_tokens": 58486582.0,
      "step": 26700
    },
    {
      "entropy": 5.6022093296051025,
      "epoch": 2.6503572846367605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004302122864888411,
      "loss": 4.9888,
      "mean_token_accuracy": 0.1860027328133583,
      "num_tokens": 58498571.0,
      "step": 26705
    },
    {
      "entropy": 5.612578821182251,
      "epoch": 2.6508535132989284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.0743,
      "mean_token_accuracy": 0.18189477771520615,
      "num_tokens": 58510445.0,
      "step": 26710
    },
    {
      "entropy": 5.614954996109009,
      "epoch": 2.651349741961096,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043016097617399327,
      "loss": 5.003,
      "mean_token_accuracy": 0.1807946652173996,
      "num_tokens": 58520248.0,
      "step": 26715
    },
    {
      "entropy": 5.670732164382935,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.0868,
      "mean_token_accuracy": 0.182686747610569,
      "num_tokens": 58531993.0,
      "step": 26720
    },
    {
      "entropy": 5.587133264541626,
      "epoch": 2.652342199285431,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043010965046851404,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18435751646757126,
      "num_tokens": 58543126.0,
      "step": 26725
    },
    {
      "entropy": 5.579779958724975,
      "epoch": 2.6528384279475983,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.0383,
      "mean_token_accuracy": 0.17794621586799622,
      "num_tokens": 58554627.0,
      "step": 26730
    },
    {
      "entropy": 5.656536340713501,
      "epoch": 2.6533346566097658,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043005830937749447,
      "loss": 4.9929,
      "mean_token_accuracy": 0.1837254896759987,
      "num_tokens": 58565177.0,
      "step": 26735
    },
    {
      "entropy": 5.634506320953369,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004300326330639984,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18443531841039656,
      "num_tokens": 58576596.0,
      "step": 26740
    },
    {
      "entropy": 5.6427237033844,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17389679849147796,
      "num_tokens": 58586250.0,
      "step": 26745
    },
    {
      "entropy": 5.642675447463989,
      "epoch": 2.654823342596268,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.0358,
      "mean_token_accuracy": 0.17754562348127365,
      "num_tokens": 58597284.0,
      "step": 26750
    },
    {
      "entropy": 5.654104328155517,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.0148,
      "mean_token_accuracy": 0.1779759481549263,
      "num_tokens": 58608888.0,
      "step": 26755
    },
    {
      "entropy": 5.53582272529602,
      "epoch": 2.6558157999206036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042992988937163066,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18523982912302017,
      "num_tokens": 58620809.0,
      "step": 26760
    },
    {
      "entropy": 5.5461509227752686,
      "epoch": 2.656312028582771,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004299041938421275,
      "loss": 4.9445,
      "mean_token_accuracy": 0.1895992025732994,
      "num_tokens": 58630981.0,
      "step": 26765
    },
    {
      "entropy": 5.6690895557403564,
      "epoch": 2.6568082572449385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042987849447133423,
      "loss": 5.1017,
      "mean_token_accuracy": 0.1748144194483757,
      "num_tokens": 58642110.0,
      "step": 26770
    },
    {
      "entropy": 5.622683095932007,
      "epoch": 2.657304485907106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18913438767194748,
      "num_tokens": 58654680.0,
      "step": 26775
    },
    {
      "entropy": 5.5992847919464115,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004298270842084261,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18475309014320374,
      "num_tokens": 58666005.0,
      "step": 26780
    },
    {
      "entropy": 5.623987007141113,
      "epoch": 2.658296943231441,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17479742020368577,
      "num_tokens": 58677245.0,
      "step": 26785
    },
    {
      "entropy": 5.61262731552124,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042977565858800593,
      "loss": 4.9975,
      "mean_token_accuracy": 0.1820876643061638,
      "num_tokens": 58688470.0,
      "step": 26790
    },
    {
      "entropy": 5.678678750991821,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.125,
      "learning_rate": 0.00042974994002032275,
      "loss": 5.0826,
      "mean_token_accuracy": 0.1794186130166054,
      "num_tokens": 58698681.0,
      "step": 26795
    },
    {
      "entropy": 5.596742916107178,
      "epoch": 2.659785629217944,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.0765,
      "mean_token_accuracy": 0.17990992069244385,
      "num_tokens": 58710113.0,
      "step": 26800
    },
    {
      "entropy": 5.650368642807007,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004296984913731991,
      "loss": 5.0511,
      "mean_token_accuracy": 0.18222518712282182,
      "num_tokens": 58720744.0,
      "step": 26805
    },
    {
      "entropy": 5.57680606842041,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004296727612950344,
      "loss": 4.962,
      "mean_token_accuracy": 0.187581966817379,
      "num_tokens": 58730955.0,
      "step": 26810
    },
    {
      "entropy": 5.681655216217041,
      "epoch": 2.6612743152044462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.1126,
      "mean_token_accuracy": 0.17940174043178558,
      "num_tokens": 58743423.0,
      "step": 26815
    },
    {
      "entropy": 5.668560934066773,
      "epoch": 2.6617705438666137,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.0271,
      "mean_token_accuracy": 0.18698217123746871,
      "num_tokens": 58754076.0,
      "step": 26820
    },
    {
      "entropy": 5.61155686378479,
      "epoch": 2.662266772528781,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042959554804978596,
      "loss": 4.9756,
      "mean_token_accuracy": 0.1807797446846962,
      "num_tokens": 58765064.0,
      "step": 26825
    },
    {
      "entropy": 5.563446712493897,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042956980263324576,
      "loss": 4.9902,
      "mean_token_accuracy": 0.18699615746736525,
      "num_tokens": 58776585.0,
      "step": 26830
    },
    {
      "entropy": 5.596900224685669,
      "epoch": 2.663259229853116,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004295440533837075,
      "loss": 5.0107,
      "mean_token_accuracy": 0.1851891905069351,
      "num_tokens": 58787533.0,
      "step": 26835
    },
    {
      "entropy": 5.581999254226685,
      "epoch": 2.6637554585152836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042951830030180983,
      "loss": 4.9317,
      "mean_token_accuracy": 0.18404416590929032,
      "num_tokens": 58797543.0,
      "step": 26840
    },
    {
      "entropy": 5.530750036239624,
      "epoch": 2.6642516871774515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18082430362701415,
      "num_tokens": 58808582.0,
      "step": 26845
    },
    {
      "entropy": 5.552934885025024,
      "epoch": 2.664747915839619,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042946678264349036,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17616550624370575,
      "num_tokens": 58819570.0,
      "step": 26850
    },
    {
      "entropy": 5.658061265945435,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.1076,
      "mean_token_accuracy": 0.17909002751111985,
      "num_tokens": 58830063.0,
      "step": 26855
    },
    {
      "entropy": 5.592504358291626,
      "epoch": 2.665740373163954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004294152496633977,
      "loss": 4.988,
      "mean_token_accuracy": 0.1980138063430786,
      "num_tokens": 58842239.0,
      "step": 26860
    },
    {
      "entropy": 5.647050428390503,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042938947742928357,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17989234924316405,
      "num_tokens": 58852570.0,
      "step": 26865
    },
    {
      "entropy": 5.622722625732422,
      "epoch": 2.666732830488289,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1813586801290512,
      "num_tokens": 58862759.0,
      "step": 26870
    },
    {
      "entropy": 5.5485881805419925,
      "epoch": 2.6672290591504564,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004293379214761156,
      "loss": 4.9831,
      "mean_token_accuracy": 0.18371664136648178,
      "num_tokens": 58873976.0,
      "step": 26875
    },
    {
      "entropy": 5.56177396774292,
      "epoch": 2.6677252878126243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.014,
      "mean_token_accuracy": 0.1857266902923584,
      "num_tokens": 58885440.0,
      "step": 26880
    },
    {
      "entropy": 5.558112716674804,
      "epoch": 2.6682215164747918,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.0397,
      "mean_token_accuracy": 0.1815976992249489,
      "num_tokens": 58896301.0,
      "step": 26885
    },
    {
      "entropy": 5.614476728439331,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004292605588436031,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18396574556827544,
      "num_tokens": 58906888.0,
      "step": 26890
    },
    {
      "entropy": 5.655603647232056,
      "epoch": 2.6692139737991267,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042923476364792063,
      "loss": 5.0911,
      "mean_token_accuracy": 0.1782705679535866,
      "num_tokens": 58918025.0,
      "step": 26895
    },
    {
      "entropy": 5.70089635848999,
      "epoch": 2.669710202461294,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1735721081495285,
      "num_tokens": 58929012.0,
      "step": 26900
    },
    {
      "entropy": 5.691879558563232,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.0676,
      "mean_token_accuracy": 0.1763743504881859,
      "num_tokens": 58939133.0,
      "step": 26905
    },
    {
      "entropy": 5.6629432201385494,
      "epoch": 2.670702659785629,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.0465,
      "mean_token_accuracy": 0.17223544269800187,
      "num_tokens": 58950227.0,
      "step": 26910
    },
    {
      "entropy": 5.577391195297241,
      "epoch": 2.6711988884477966,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004291315446246909,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18822240084409714,
      "num_tokens": 58961614.0,
      "step": 26915
    },
    {
      "entropy": 5.597414922714234,
      "epoch": 2.671695117109964,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.0571,
      "mean_token_accuracy": 0.181965109705925,
      "num_tokens": 58973252.0,
      "step": 26920
    },
    {
      "entropy": 5.620335435867309,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.036,
      "mean_token_accuracy": 0.1809214323759079,
      "num_tokens": 58985223.0,
      "step": 26925
    },
    {
      "entropy": 5.558753967285156,
      "epoch": 2.672687574434299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004290540902226624,
      "loss": 4.9533,
      "mean_token_accuracy": 0.18620461523532866,
      "num_tokens": 58995651.0,
      "step": 26930
    },
    {
      "entropy": 5.608013582229614,
      "epoch": 2.673183803096467,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18261505365371705,
      "num_tokens": 59006171.0,
      "step": 26935
    },
    {
      "entropy": 5.586583662033081,
      "epoch": 2.6736800317586344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004290024348525289,
      "loss": 5.03,
      "mean_token_accuracy": 0.1851085364818573,
      "num_tokens": 59017595.0,
      "step": 26940
    },
    {
      "entropy": 5.626079797744751,
      "epoch": 2.674176260420802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004289766014387492,
      "loss": 5.0941,
      "mean_token_accuracy": 0.175155346095562,
      "num_tokens": 59028896.0,
      "step": 26945
    },
    {
      "entropy": 5.601846075057983,
      "epoch": 2.6746724890829694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004289507642066813,
      "loss": 4.9365,
      "mean_token_accuracy": 0.18305384516716003,
      "num_tokens": 59040239.0,
      "step": 26950
    },
    {
      "entropy": 5.642567873001099,
      "epoch": 2.675168717745137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004289249231569664,
      "loss": 5.1166,
      "mean_token_accuracy": 0.17672800421714782,
      "num_tokens": 59050742.0,
      "step": 26955
    },
    {
      "entropy": 5.737293148040772,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042889907829024506,
      "loss": 5.1716,
      "mean_token_accuracy": 0.17849161773920058,
      "num_tokens": 59062985.0,
      "step": 26960
    },
    {
      "entropy": 5.713742399215699,
      "epoch": 2.6761611750694723,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.0785,
      "mean_token_accuracy": 0.18263549208641053,
      "num_tokens": 59074384.0,
      "step": 26965
    },
    {
      "entropy": 5.649041366577149,
      "epoch": 2.6766574037316397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004288473771083468,
      "loss": 5.1664,
      "mean_token_accuracy": 0.17673540860414505,
      "num_tokens": 59086332.0,
      "step": 26970
    },
    {
      "entropy": 5.54486608505249,
      "epoch": 2.677153632393807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004288215207944519,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18606456965208054,
      "num_tokens": 59098300.0,
      "step": 26975
    },
    {
      "entropy": 5.674498605728149,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.1299,
      "mean_token_accuracy": 0.18352297097444534,
      "num_tokens": 59109596.0,
      "step": 26980
    },
    {
      "entropy": 5.68747353553772,
      "epoch": 2.678146089718142,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18558502942323685,
      "num_tokens": 59119590.0,
      "step": 26985
    },
    {
      "entropy": 5.617304992675781,
      "epoch": 2.6786423183803096,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000428743928968679,
      "loss": 5.0738,
      "mean_token_accuracy": 0.18066728711128235,
      "num_tokens": 59132219.0,
      "step": 26990
    },
    {
      "entropy": 5.635842275619507,
      "epoch": 2.679138547042477,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042871805740086317,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18674781173467636,
      "num_tokens": 59142752.0,
      "step": 26995
    },
    {
      "entropy": 5.59537034034729,
      "epoch": 2.6796347757046446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004286921820211707,
      "loss": 4.9796,
      "mean_token_accuracy": 0.1840292602777481,
      "num_tokens": 59153458.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.362994519286044,
      "eval_loss": 5.1428022384643555,
      "eval_mean_token_accuracy": 0.18292152676461398,
      "eval_num_tokens": 59153458.0,
      "eval_runtime": 20.2067,
      "eval_samples_per_second": 1666.23,
      "eval_steps_per_second": 208.297,
      "step": 27000
    },
    {
      "entropy": 5.610469484329224,
      "epoch": 2.680131004366812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042866630283024327,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18691931217908858,
      "num_tokens": 59164755.0,
      "step": 27005
    },
    {
      "entropy": 5.547487545013428,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.9488,
      "mean_token_accuracy": 0.17940691113471985,
      "num_tokens": 59174470.0,
      "step": 27010
    },
    {
      "entropy": 5.63067512512207,
      "epoch": 2.681123461691147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004286145330172508,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18745761811733247,
      "num_tokens": 59184839.0,
      "step": 27015
    },
    {
      "entropy": 5.5811717987060545,
      "epoch": 2.681619690353315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004285886423964695,
      "loss": 4.9677,
      "mean_token_accuracy": 0.18680686205625535,
      "num_tokens": 59194380.0,
      "step": 27020
    },
    {
      "entropy": 5.623590326309204,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.0765,
      "mean_token_accuracy": 0.18077082186937332,
      "num_tokens": 59205509.0,
      "step": 27025
    },
    {
      "entropy": 5.615319490432739,
      "epoch": 2.68261214767765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004285368497295468,
      "loss": 4.9492,
      "mean_token_accuracy": 0.1915307030081749,
      "num_tokens": 59216342.0,
      "step": 27030
    },
    {
      "entropy": 5.628066110610962,
      "epoch": 2.6831083763398174,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004285109476846898,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18301252871751786,
      "num_tokens": 59227217.0,
      "step": 27035
    },
    {
      "entropy": 5.67046275138855,
      "epoch": 2.683604605001985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004284850418330921,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17255839407444,
      "num_tokens": 59239441.0,
      "step": 27040
    },
    {
      "entropy": 5.660286808013916,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1800799697637558,
      "num_tokens": 59250015.0,
      "step": 27045
    },
    {
      "entropy": 5.59221076965332,
      "epoch": 2.6845970623263202,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042843321871224426,
      "loss": 4.9222,
      "mean_token_accuracy": 0.19030543714761733,
      "num_tokens": 59259673.0,
      "step": 27050
    },
    {
      "entropy": 5.6301239967346195,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.125,
      "learning_rate": 0.00042840730144427927,
      "loss": 5.0995,
      "mean_token_accuracy": 0.18026266098022461,
      "num_tokens": 59270927.0,
      "step": 27055
    },
    {
      "entropy": 5.560189247131348,
      "epoch": 2.685589519650655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042838138037214375,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18111251443624496,
      "num_tokens": 59280831.0,
      "step": 27060
    },
    {
      "entropy": 5.694497013092041,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.125,
      "learning_rate": 0.00042835545549648054,
      "loss": 5.067,
      "mean_token_accuracy": 0.17390521317720414,
      "num_tokens": 59291728.0,
      "step": 27065
    },
    {
      "entropy": 5.59930305480957,
      "epoch": 2.68658197697499,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042832952681793235,
      "loss": 4.9518,
      "mean_token_accuracy": 0.18847705721855162,
      "num_tokens": 59302094.0,
      "step": 27070
    },
    {
      "entropy": 5.580347633361816,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042830359433714243,
      "loss": 5.0228,
      "mean_token_accuracy": 0.18443397730588912,
      "num_tokens": 59312951.0,
      "step": 27075
    },
    {
      "entropy": 5.627540493011475,
      "epoch": 2.687574434299325,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042827765805475367,
      "loss": 5.053,
      "mean_token_accuracy": 0.1745489090681076,
      "num_tokens": 59322858.0,
      "step": 27080
    },
    {
      "entropy": 5.6244548797607425,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042825171797140926,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18401222079992294,
      "num_tokens": 59332440.0,
      "step": 27085
    },
    {
      "entropy": 5.5580246925354,
      "epoch": 2.68856689162366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004282257740877525,
      "loss": 4.9589,
      "mean_token_accuracy": 0.187635238468647,
      "num_tokens": 59343265.0,
      "step": 27090
    },
    {
      "entropy": 5.535512161254883,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004281998264044267,
      "loss": 4.9711,
      "mean_token_accuracy": 0.18208514600992204,
      "num_tokens": 59354212.0,
      "step": 27095
    },
    {
      "entropy": 5.634253597259521,
      "epoch": 2.689559348947995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004281738749220753,
      "loss": 4.9832,
      "mean_token_accuracy": 0.18813330829143524,
      "num_tokens": 59365674.0,
      "step": 27100
    },
    {
      "entropy": 5.619606924057007,
      "epoch": 2.690055577610163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042814791964134184,
      "loss": 4.9732,
      "mean_token_accuracy": 0.19271332323551177,
      "num_tokens": 59376329.0,
      "step": 27105
    },
    {
      "entropy": 5.630713701248169,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042812196056287,
      "loss": 5.0566,
      "mean_token_accuracy": 0.18373579829931258,
      "num_tokens": 59387689.0,
      "step": 27110
    },
    {
      "entropy": 5.635879564285278,
      "epoch": 2.691048034934498,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.0958,
      "mean_token_accuracy": 0.18195635080337524,
      "num_tokens": 59397285.0,
      "step": 27115
    },
    {
      "entropy": 5.564650917053223,
      "epoch": 2.6915442635966653,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004280700310152861,
      "loss": 5.0016,
      "mean_token_accuracy": 0.18132225126028062,
      "num_tokens": 59409284.0,
      "step": 27120
    },
    {
      "entropy": 5.608187818527222,
      "epoch": 2.692040492258833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004280440605474619,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18444342017173768,
      "num_tokens": 59421086.0,
      "step": 27125
    },
    {
      "entropy": 5.68573751449585,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17841769009828568,
      "num_tokens": 59433226.0,
      "step": 27130
    },
    {
      "entropy": 5.604090833663941,
      "epoch": 2.693032949583168,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042799210822696865,
      "loss": 4.9705,
      "mean_token_accuracy": 0.19219637513160706,
      "num_tokens": 59442613.0,
      "step": 27135
    },
    {
      "entropy": 5.64068832397461,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000427966126375588,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18101863116025924,
      "num_tokens": 59453473.0,
      "step": 27140
    },
    {
      "entropy": 5.606676483154297,
      "epoch": 2.694025406907503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004279401407309771,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1900673732161522,
      "num_tokens": 59464495.0,
      "step": 27145
    },
    {
      "entropy": 5.599293899536133,
      "epoch": 2.6945216355696706,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.0535,
      "mean_token_accuracy": 0.17849814891815186,
      "num_tokens": 59475575.0,
      "step": 27150
    },
    {
      "entropy": 5.5497406959533695,
      "epoch": 2.695017864231838,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004278881580646418,
      "loss": 4.9802,
      "mean_token_accuracy": 0.18267161250114441,
      "num_tokens": 59487432.0,
      "step": 27155
    },
    {
      "entropy": 5.625388717651367,
      "epoch": 2.6955140928940056,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004278621610442066,
      "loss": 4.9882,
      "mean_token_accuracy": 0.18459143489599228,
      "num_tokens": 59498694.0,
      "step": 27160
    },
    {
      "entropy": 5.698960208892823,
      "epoch": 2.696010321556173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004278361602331192,
      "loss": 5.1069,
      "mean_token_accuracy": 0.18502866327762604,
      "num_tokens": 59509303.0,
      "step": 27165
    },
    {
      "entropy": 5.586273908615112,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042781015563202434,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18225349634885787,
      "num_tokens": 59519924.0,
      "step": 27170
    },
    {
      "entropy": 5.61257586479187,
      "epoch": 2.697002778880508,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004277841472415669,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18672785460948943,
      "num_tokens": 59530880.0,
      "step": 27175
    },
    {
      "entropy": 5.639212369918823,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17589643299579621,
      "num_tokens": 59542314.0,
      "step": 27180
    },
    {
      "entropy": 5.639403390884399,
      "epoch": 2.697995236204843,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004277321190951441,
      "loss": 5.0227,
      "mean_token_accuracy": 0.17470503896474837,
      "num_tokens": 59552776.0,
      "step": 27185
    },
    {
      "entropy": 5.6525147438049315,
      "epoch": 2.698491464867011,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042770609934046904,
      "loss": 5.0125,
      "mean_token_accuracy": 0.18526321947574614,
      "num_tokens": 59563302.0,
      "step": 27190
    },
    {
      "entropy": 5.566699123382568,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042768007579901176,
      "loss": 4.999,
      "mean_token_accuracy": 0.17922999560832978,
      "num_tokens": 59574151.0,
      "step": 27195
    },
    {
      "entropy": 5.6173083782196045,
      "epoch": 2.699483922191346,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.0511,
      "mean_token_accuracy": 0.1806842178106308,
      "num_tokens": 59585093.0,
      "step": 27200
    },
    {
      "entropy": 5.614673280715943,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004276280173583319,
      "loss": 4.9483,
      "mean_token_accuracy": 0.19660182297229767,
      "num_tokens": 59594908.0,
      "step": 27205
    },
    {
      "entropy": 5.56649341583252,
      "epoch": 2.7004763795156808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042760198246040025,
      "loss": 4.9764,
      "mean_token_accuracy": 0.1939328208565712,
      "num_tokens": 59607015.0,
      "step": 27210
    },
    {
      "entropy": 5.667684841156006,
      "epoch": 2.7009726081778482,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.118,
      "mean_token_accuracy": 0.17699771523475646,
      "num_tokens": 59617597.0,
      "step": 27215
    },
    {
      "entropy": 5.610971689224243,
      "epoch": 2.701468836840016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042754990131258175,
      "loss": 4.9626,
      "mean_token_accuracy": 0.18390912115573882,
      "num_tokens": 59627934.0,
      "step": 27220
    },
    {
      "entropy": 5.61381721496582,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.0701,
      "mean_token_accuracy": 0.17441867887973786,
      "num_tokens": 59638759.0,
      "step": 27225
    },
    {
      "entropy": 5.623507165908814,
      "epoch": 2.702461294164351,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000427497805033128,
      "loss": 5.0539,
      "mean_token_accuracy": 0.17944359332323073,
      "num_tokens": 59650199.0,
      "step": 27230
    },
    {
      "entropy": 5.614063501358032,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004274717512206527,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1727106511592865,
      "num_tokens": 59661480.0,
      "step": 27235
    },
    {
      "entropy": 5.672021913528442,
      "epoch": 2.703453751488686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004274456936272065,
      "loss": 5.1024,
      "mean_token_accuracy": 0.18157360851764678,
      "num_tokens": 59673048.0,
      "step": 27240
    },
    {
      "entropy": 5.687104797363281,
      "epoch": 2.7039499801508535,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042741963225343554,
      "loss": 5.1362,
      "mean_token_accuracy": 0.17049107998609542,
      "num_tokens": 59686218.0,
      "step": 27245
    },
    {
      "entropy": 5.608295917510986,
      "epoch": 2.704446208813021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.0152,
      "mean_token_accuracy": 0.18225342482328416,
      "num_tokens": 59695870.0,
      "step": 27250
    },
    {
      "entropy": 5.57289228439331,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042736749816750476,
      "loss": 4.9392,
      "mean_token_accuracy": 0.18435097634792327,
      "num_tokens": 59706479.0,
      "step": 27255
    },
    {
      "entropy": 5.63434910774231,
      "epoch": 2.705438666137356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004273414254566376,
      "loss": 5.1005,
      "mean_token_accuracy": 0.18087097257375717,
      "num_tokens": 59717411.0,
      "step": 27260
    },
    {
      "entropy": 5.639881944656372,
      "epoch": 2.7059348947995234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.079,
      "mean_token_accuracy": 0.1774018093943596,
      "num_tokens": 59728363.0,
      "step": 27265
    },
    {
      "entropy": 5.595034217834472,
      "epoch": 2.706431123461691,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042728926870233287,
      "loss": 4.9802,
      "mean_token_accuracy": 0.18331198394298553,
      "num_tokens": 59738714.0,
      "step": 27270
    },
    {
      "entropy": 5.627304172515869,
      "epoch": 2.706927352123859,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18690527230501175,
      "num_tokens": 59749292.0,
      "step": 27275
    },
    {
      "entropy": 5.643618965148926,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.0,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17658531218767165,
      "num_tokens": 59760207.0,
      "step": 27280
    },
    {
      "entropy": 5.578765630722046,
      "epoch": 2.7079198094481938,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042721100524915017,
      "loss": 4.9358,
      "mean_token_accuracy": 0.18938130438327788,
      "num_tokens": 59771780.0,
      "step": 27285
    },
    {
      "entropy": 5.687202072143554,
      "epoch": 2.7084160381103612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042718490988155014,
      "loss": 5.05,
      "mean_token_accuracy": 0.18195660263299943,
      "num_tokens": 59782298.0,
      "step": 27290
    },
    {
      "entropy": 5.640627098083496,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17971403151750565,
      "num_tokens": 59794097.0,
      "step": 27295
    },
    {
      "entropy": 5.631379461288452,
      "epoch": 2.709408495434696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042713270782542364,
      "loss": 4.986,
      "mean_token_accuracy": 0.18911930322647094,
      "num_tokens": 59804373.0,
      "step": 27300
    },
    {
      "entropy": 5.591650724411011,
      "epoch": 2.7099047240968637,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004271066011381917,
      "loss": 5.038,
      "mean_token_accuracy": 0.1816621631383896,
      "num_tokens": 59817422.0,
      "step": 27305
    },
    {
      "entropy": 5.593688488006592,
      "epoch": 2.7104009527590316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004270804906790438,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17794821113348008,
      "num_tokens": 59828531.0,
      "step": 27310
    },
    {
      "entropy": 5.668269252777099,
      "epoch": 2.710897181421199,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004270543764486275,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17984321415424348,
      "num_tokens": 59839817.0,
      "step": 27315
    },
    {
      "entropy": 5.642149353027344,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.0276,
      "mean_token_accuracy": 0.1874216988682747,
      "num_tokens": 59850697.0,
      "step": 27320
    },
    {
      "entropy": 5.568265056610107,
      "epoch": 2.711889638745534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18770492523908616,
      "num_tokens": 59861255.0,
      "step": 27325
    },
    {
      "entropy": 5.5994030952453615,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.078125,
      "learning_rate": 0.000426976011136244,
      "loss": 5.0081,
      "mean_token_accuracy": 0.1858513042330742,
      "num_tokens": 59872225.0,
      "step": 27330
    },
    {
      "entropy": 5.683399438858032,
      "epoch": 2.712882096069869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.0626,
      "mean_token_accuracy": 0.18225167095661163,
      "num_tokens": 59882588.0,
      "step": 27335
    },
    {
      "entropy": 5.643529796600342,
      "epoch": 2.7133783247320364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.1562,
      "mean_token_accuracy": 0.17775670289993287,
      "num_tokens": 59895023.0,
      "step": 27340
    },
    {
      "entropy": 5.6315428733825685,
      "epoch": 2.713874553394204,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042689761190576263,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18100069165229798,
      "num_tokens": 59906497.0,
      "step": 27345
    },
    {
      "entropy": 5.65015606880188,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042687147129460447,
      "loss": 4.9832,
      "mean_token_accuracy": 0.18016820698976516,
      "num_tokens": 59917261.0,
      "step": 27350
    },
    {
      "entropy": 5.568005657196045,
      "epoch": 2.714867010718539,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18184488266706467,
      "num_tokens": 59928398.0,
      "step": 27355
    },
    {
      "entropy": 5.565901231765747,
      "epoch": 2.715363239380707,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042681917877468076,
      "loss": 4.9469,
      "mean_token_accuracy": 0.18393584042787553,
      "num_tokens": 59939259.0,
      "step": 27360
    },
    {
      "entropy": 5.624455404281616,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042679302686721187,
      "loss": 4.9794,
      "mean_token_accuracy": 0.18358264863491058,
      "num_tokens": 59950495.0,
      "step": 27365
    },
    {
      "entropy": 5.619688701629639,
      "epoch": 2.7163556967050417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004267668711956031,
      "loss": 5.052,
      "mean_token_accuracy": 0.18143973648548126,
      "num_tokens": 59963038.0,
      "step": 27370
    },
    {
      "entropy": 5.592446327209473,
      "epoch": 2.716851925367209,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042674071176050296,
      "loss": 5.0581,
      "mean_token_accuracy": 0.18214502632617952,
      "num_tokens": 59973698.0,
      "step": 27375
    },
    {
      "entropy": 5.635765743255615,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042671454856256,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18157849758863448,
      "num_tokens": 59984098.0,
      "step": 27380
    },
    {
      "entropy": 5.628144788742065,
      "epoch": 2.717844382691544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004266883816024233,
      "loss": 4.944,
      "mean_token_accuracy": 0.18435648828744888,
      "num_tokens": 59995222.0,
      "step": 27385
    },
    {
      "entropy": 5.6253662109375,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004266622108807415,
      "loss": 5.0843,
      "mean_token_accuracy": 0.18819310069084166,
      "num_tokens": 60006886.0,
      "step": 27390
    },
    {
      "entropy": 5.616521549224854,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.0651,
      "mean_token_accuracy": 0.18369756489992142,
      "num_tokens": 60018408.0,
      "step": 27395
    },
    {
      "entropy": 5.668343639373779,
      "epoch": 2.719333068678047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042660985815533905,
      "loss": 5.0679,
      "mean_token_accuracy": 0.1725982114672661,
      "num_tokens": 60029927.0,
      "step": 27400
    },
    {
      "entropy": 5.655353546142578,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.125,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18041362315416337,
      "num_tokens": 60040521.0,
      "step": 27405
    },
    {
      "entropy": 5.560429620742798,
      "epoch": 2.720325526002382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004265574903915454,
      "loss": 4.9764,
      "mean_token_accuracy": 0.1866554692387581,
      "num_tokens": 60051709.0,
      "step": 27410
    },
    {
      "entropy": 5.570276927947998,
      "epoch": 2.7208217546645495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004265313008718751,
      "loss": 4.9895,
      "mean_token_accuracy": 0.1822122037410736,
      "num_tokens": 60062478.0,
      "step": 27415
    },
    {
      "entropy": 5.657826900482178,
      "epoch": 2.721317983326717,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042650510759455516,
      "loss": 5.0656,
      "mean_token_accuracy": 0.18130711019039153,
      "num_tokens": 60072957.0,
      "step": 27420
    },
    {
      "entropy": 5.67852463722229,
      "epoch": 2.7218142119888844,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042647891056023495,
      "loss": 5.0356,
      "mean_token_accuracy": 0.1855493262410164,
      "num_tokens": 60084174.0,
      "step": 27425
    },
    {
      "entropy": 5.601739311218262,
      "epoch": 2.722310440651052,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042645270976956426,
      "loss": 4.9504,
      "mean_token_accuracy": 0.19351554214954375,
      "num_tokens": 60095647.0,
      "step": 27430
    },
    {
      "entropy": 5.605778074264526,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.0405,
      "mean_token_accuracy": 0.17666068971157073,
      "num_tokens": 60106355.0,
      "step": 27435
    },
    {
      "entropy": 5.574138021469116,
      "epoch": 2.723302897975387,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004264002969217699,
      "loss": 4.9322,
      "mean_token_accuracy": 0.18869348764419555,
      "num_tokens": 60116616.0,
      "step": 27440
    },
    {
      "entropy": 5.657160472869873,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.0579,
      "mean_token_accuracy": 0.18423389196395873,
      "num_tokens": 60127168.0,
      "step": 27445
    },
    {
      "entropy": 5.615782022476196,
      "epoch": 2.724295355299722,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042634786905637123,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18738060891628266,
      "num_tokens": 60137953.0,
      "step": 27450
    },
    {
      "entropy": 5.645580720901489,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004263216494936953,
      "loss": 5.0359,
      "mean_token_accuracy": 0.18219154626131057,
      "num_tokens": 60149513.0,
      "step": 27455
    },
    {
      "entropy": 5.613449144363403,
      "epoch": 2.725287812624057,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042629542617856855,
      "loss": 4.9871,
      "mean_token_accuracy": 0.18548113405704497,
      "num_tokens": 60160977.0,
      "step": 27460
    },
    {
      "entropy": 5.600784778594971,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004262691991116412,
      "loss": 5.006,
      "mean_token_accuracy": 0.18338669538497926,
      "num_tokens": 60171607.0,
      "step": 27465
    },
    {
      "entropy": 5.620282173156738,
      "epoch": 2.726280269948392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004262429682935637,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17497291415929794,
      "num_tokens": 60183386.0,
      "step": 27470
    },
    {
      "entropy": 5.624786710739135,
      "epoch": 2.7267764986105596,
      "grad_norm": 1.125,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.0167,
      "mean_token_accuracy": 0.1791770026087761,
      "num_tokens": 60194494.0,
      "step": 27475
    },
    {
      "entropy": 5.546691417694092,
      "epoch": 2.7272727272727275,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004261904954065601,
      "loss": 4.9339,
      "mean_token_accuracy": 0.1923592671751976,
      "num_tokens": 60206053.0,
      "step": 27480
    },
    {
      "entropy": 5.600972604751587,
      "epoch": 2.727768955934895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004261642533389352,
      "loss": 5.0262,
      "mean_token_accuracy": 0.1836771249771118,
      "num_tokens": 60218564.0,
      "step": 27485
    },
    {
      "entropy": 5.6251123428344725,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042613800752276264,
      "loss": 4.9956,
      "mean_token_accuracy": 0.193145090341568,
      "num_tokens": 60228979.0,
      "step": 27490
    },
    {
      "entropy": 5.6415948390960695,
      "epoch": 2.72876141325923,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17410266995429993,
      "num_tokens": 60239143.0,
      "step": 27495
    },
    {
      "entropy": 5.660939645767212,
      "epoch": 2.7292576419213974,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004260855046473775,
      "loss": 5.1243,
      "mean_token_accuracy": 0.17569600641727448,
      "num_tokens": 60249815.0,
      "step": 27500
    },
    {
      "entropy": 5.615096378326416,
      "epoch": 2.729753870583565,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042605924758946706,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18661460727453233,
      "num_tokens": 60259909.0,
      "step": 27505
    },
    {
      "entropy": 5.668798923492432,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042603298678561277,
      "loss": 5.0672,
      "mean_token_accuracy": 0.18343012779951096,
      "num_tokens": 60271020.0,
      "step": 27510
    },
    {
      "entropy": 5.638863658905029,
      "epoch": 2.7307463279079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004260067222364659,
      "loss": 4.947,
      "mean_token_accuracy": 0.18489052206277848,
      "num_tokens": 60281986.0,
      "step": 27515
    },
    {
      "entropy": 5.613531637191772,
      "epoch": 2.7312425565700673,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042598045394267757,
      "loss": 5.062,
      "mean_token_accuracy": 0.1804831638932228,
      "num_tokens": 60292435.0,
      "step": 27520
    },
    {
      "entropy": 5.630309820175171,
      "epoch": 2.731738785232235,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042595418190489944,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18255549371242524,
      "num_tokens": 60303783.0,
      "step": 27525
    },
    {
      "entropy": 5.605478572845459,
      "epoch": 2.7322350138944023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1823076128959656,
      "num_tokens": 60313525.0,
      "step": 27530
    },
    {
      "entropy": 5.5632617473602295,
      "epoch": 2.73273124255657,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042590162659997933,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18820941150188447,
      "num_tokens": 60323788.0,
      "step": 27535
    },
    {
      "entropy": 5.6485004901885985,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.1591,
      "mean_token_accuracy": 0.18091691434383392,
      "num_tokens": 60334006.0,
      "step": 27540
    },
    {
      "entropy": 5.66347484588623,
      "epoch": 2.733723699880905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1842745453119278,
      "num_tokens": 60345339.0,
      "step": 27545
    },
    {
      "entropy": 5.621183013916015,
      "epoch": 2.7342199285430726,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004258227655789651,
      "loss": 5.0534,
      "mean_token_accuracy": 0.18076043725013732,
      "num_tokens": 60356642.0,
      "step": 27550
    },
    {
      "entropy": 5.576416301727295,
      "epoch": 2.73471615720524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004257964710909318,
      "loss": 4.9669,
      "mean_token_accuracy": 0.19468714892864228,
      "num_tokens": 60367244.0,
      "step": 27555
    },
    {
      "entropy": 5.563063716888427,
      "epoch": 2.7352123858674076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042577017286347116,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18722303509712218,
      "num_tokens": 60377992.0,
      "step": 27560
    },
    {
      "entropy": 5.570778512954712,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.25,
      "learning_rate": 0.00042574387089723504,
      "loss": 4.9492,
      "mean_token_accuracy": 0.182950758934021,
      "num_tokens": 60389007.0,
      "step": 27565
    },
    {
      "entropy": 5.638987827301025,
      "epoch": 2.736204843191743,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042571756519287567,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1773553803563118,
      "num_tokens": 60399274.0,
      "step": 27570
    },
    {
      "entropy": 5.65325961112976,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042569125575104557,
      "loss": 5.056,
      "mean_token_accuracy": 0.17999257892370224,
      "num_tokens": 60411158.0,
      "step": 27575
    },
    {
      "entropy": 5.524842691421509,
      "epoch": 2.737197300516078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042566494257239695,
      "loss": 4.9818,
      "mean_token_accuracy": 0.19014695882797242,
      "num_tokens": 60422790.0,
      "step": 27580
    },
    {
      "entropy": 5.6130811214447025,
      "epoch": 2.7376935291782454,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.1578,
      "mean_token_accuracy": 0.1709316298365593,
      "num_tokens": 60434353.0,
      "step": 27585
    },
    {
      "entropy": 5.573091316223144,
      "epoch": 2.738189757840413,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004256123050072546,
      "loss": 4.9394,
      "mean_token_accuracy": 0.18900058269500733,
      "num_tokens": 60445497.0,
      "step": 27590
    },
    {
      "entropy": 5.63444447517395,
      "epoch": 2.7386859865025803,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042558598062206616,
      "loss": 4.9417,
      "mean_token_accuracy": 0.18754749298095702,
      "num_tokens": 60456358.0,
      "step": 27595
    },
    {
      "entropy": 5.626675939559936,
      "epoch": 2.739182215164748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042555965250266984,
      "loss": 5.0735,
      "mean_token_accuracy": 0.18262193500995635,
      "num_tokens": 60467613.0,
      "step": 27600
    },
    {
      "entropy": 5.603899431228638,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1772838532924652,
      "num_tokens": 60479366.0,
      "step": 27605
    },
    {
      "entropy": 5.677643251419068,
      "epoch": 2.7401746724890828,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042550698506386536,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17912378609180452,
      "num_tokens": 60489199.0,
      "step": 27610
    },
    {
      "entropy": 5.615841484069824,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042548064574576323,
      "loss": 5.1101,
      "mean_token_accuracy": 0.17853691279888154,
      "num_tokens": 60500051.0,
      "step": 27615
    },
    {
      "entropy": 5.623627138137818,
      "epoch": 2.741167129813418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004254543026960654,
      "loss": 5.088,
      "mean_token_accuracy": 0.17212245166301726,
      "num_tokens": 60510699.0,
      "step": 27620
    },
    {
      "entropy": 5.6296416282653805,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004254279559154249,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18348565697669983,
      "num_tokens": 60520224.0,
      "step": 27625
    },
    {
      "entropy": 5.645521593093872,
      "epoch": 2.742159587137753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1839382380247116,
      "num_tokens": 60531829.0,
      "step": 27630
    },
    {
      "entropy": 5.555017805099487,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004253752511639301,
      "loss": 4.991,
      "mean_token_accuracy": 0.17737457752227784,
      "num_tokens": 60543278.0,
      "step": 27635
    },
    {
      "entropy": 5.672971296310425,
      "epoch": 2.743152044462088,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004253488931943826,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18135726153850557,
      "num_tokens": 60554158.0,
      "step": 27640
    },
    {
      "entropy": 5.623119926452636,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042532253149650653,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18573970198631287,
      "num_tokens": 60563981.0,
      "step": 27645
    },
    {
      "entropy": 5.627855539321899,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18251785039901733,
      "num_tokens": 60574648.0,
      "step": 27650
    },
    {
      "entropy": 5.632614469528198,
      "epoch": 2.744640730448591,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042526979691838353,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18453242480754853,
      "num_tokens": 60585715.0,
      "step": 27655
    },
    {
      "entropy": 5.671186017990112,
      "epoch": 2.7451369591107584,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.1462,
      "mean_token_accuracy": 0.17998500913381577,
      "num_tokens": 60597523.0,
      "step": 27660
    },
    {
      "entropy": 5.697353982925415,
      "epoch": 2.745633187772926,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.0934,
      "mean_token_accuracy": 0.17595824152231215,
      "num_tokens": 60607887.0,
      "step": 27665
    },
    {
      "entropy": 5.648639726638794,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.125,
      "learning_rate": 0.00042519066710508064,
      "loss": 5.0971,
      "mean_token_accuracy": 0.1769426241517067,
      "num_tokens": 60619007.0,
      "step": 27670
    },
    {
      "entropy": 5.610061311721802,
      "epoch": 2.746625645097261,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042516428305096435,
      "loss": 4.976,
      "mean_token_accuracy": 0.19190390706062316,
      "num_tokens": 60629351.0,
      "step": 27675
    },
    {
      "entropy": 5.666254997253418,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004251378952730973,
      "loss": 5.0406,
      "mean_token_accuracy": 0.18697345852851868,
      "num_tokens": 60640145.0,
      "step": 27680
    },
    {
      "entropy": 5.637033557891845,
      "epoch": 2.7476181024215958,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004251115037721341,
      "loss": 5.0191,
      "mean_token_accuracy": 0.19072166532278062,
      "num_tokens": 60649981.0,
      "step": 27685
    },
    {
      "entropy": 5.634880495071411,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.0163,
      "mean_token_accuracy": 0.1808687075972557,
      "num_tokens": 60660572.0,
      "step": 27690
    },
    {
      "entropy": 5.602802181243897,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.003,
      "mean_token_accuracy": 0.18138339072465898,
      "num_tokens": 60673136.0,
      "step": 27695
    },
    {
      "entropy": 5.666414976119995,
      "epoch": 2.749106788408098,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.0744,
      "mean_token_accuracy": 0.18024762868881225,
      "num_tokens": 60683220.0,
      "step": 27700
    },
    {
      "entropy": 5.702374982833862,
      "epoch": 2.749603017070266,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.0877,
      "mean_token_accuracy": 0.17729443609714507,
      "num_tokens": 60695030.0,
      "step": 27705
    },
    {
      "entropy": 5.614454936981201,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.0167,
      "mean_token_accuracy": 0.17777205258607864,
      "num_tokens": 60705268.0,
      "step": 27710
    },
    {
      "entropy": 5.586823034286499,
      "epoch": 2.750595474394601,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042495307661798837,
      "loss": 4.987,
      "mean_token_accuracy": 0.184844271838665,
      "num_tokens": 60716091.0,
      "step": 27715
    },
    {
      "entropy": 5.659141206741333,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.0727,
      "mean_token_accuracy": 0.1811782419681549,
      "num_tokens": 60727255.0,
      "step": 27720
    },
    {
      "entropy": 5.642379999160767,
      "epoch": 2.751587931718936,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042490023781151546,
      "loss": 5.0268,
      "mean_token_accuracy": 0.18253542184829713,
      "num_tokens": 60738014.0,
      "step": 27725
    },
    {
      "entropy": 5.6611504554748535,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.125,
      "learning_rate": 0.00042487381283214706,
      "loss": 5.1453,
      "mean_token_accuracy": 0.17326537221670152,
      "num_tokens": 60748935.0,
      "step": 27730
    },
    {
      "entropy": 5.6171351909637455,
      "epoch": 2.7525803890432714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18802643865346907,
      "num_tokens": 60760046.0,
      "step": 27735
    },
    {
      "entropy": 5.6595869064331055,
      "epoch": 2.753076617705439,
      "grad_norm": 1.0,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.0596,
      "mean_token_accuracy": 0.17857325673103333,
      "num_tokens": 60771198.0,
      "step": 27740
    },
    {
      "entropy": 5.643102216720581,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004247945155973778,
      "loss": 5.0907,
      "mean_token_accuracy": 0.1820570409297943,
      "num_tokens": 60782018.0,
      "step": 27745
    },
    {
      "entropy": 5.6719728946685795,
      "epoch": 2.754069075029774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.0965,
      "mean_token_accuracy": 0.1824449822306633,
      "num_tokens": 60794108.0,
      "step": 27750
    },
    {
      "entropy": 5.636381721496582,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042474163220019987,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18201542198657988,
      "num_tokens": 60803941.0,
      "step": 27755
    },
    {
      "entropy": 5.622543621063232,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.0271,
      "mean_token_accuracy": 0.1840425431728363,
      "num_tokens": 60814772.0,
      "step": 27760
    },
    {
      "entropy": 5.692751216888428,
      "epoch": 2.7555577610162763,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004246887339499427,
      "loss": 5.1229,
      "mean_token_accuracy": 0.1732679858803749,
      "num_tokens": 60825289.0,
      "step": 27765
    },
    {
      "entropy": 5.636129331588745,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042466227925654917,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17570931911468507,
      "num_tokens": 60836440.0,
      "step": 27770
    },
    {
      "entropy": 5.613993883132935,
      "epoch": 2.756550218340611,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004246358208518535,
      "loss": 5.0428,
      "mean_token_accuracy": 0.18410896211862565,
      "num_tokens": 60847002.0,
      "step": 27775
    },
    {
      "entropy": 5.648705339431762,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042460935873651196,
      "loss": 5.0519,
      "mean_token_accuracy": 0.18234162777662277,
      "num_tokens": 60858240.0,
      "step": 27780
    },
    {
      "entropy": 5.597128820419312,
      "epoch": 2.757542675664946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1803037405014038,
      "num_tokens": 60870018.0,
      "step": 27785
    },
    {
      "entropy": 5.61336498260498,
      "epoch": 2.758038904327114,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042455642337651595,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18599406033754348,
      "num_tokens": 60881143.0,
      "step": 27790
    },
    {
      "entropy": 5.622269487380981,
      "epoch": 2.7585351329892815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004245299501331743,
      "loss": 5.0793,
      "mean_token_accuracy": 0.1823672503232956,
      "num_tokens": 60892347.0,
      "step": 27795
    },
    {
      "entropy": 5.683091449737549,
      "epoch": 2.759031361651449,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.0919,
      "mean_token_accuracy": 0.18166449815034866,
      "num_tokens": 60903420.0,
      "step": 27800
    },
    {
      "entropy": 5.6001993179321286,
      "epoch": 2.7595275903136165,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042447699252308566,
      "loss": 4.9295,
      "mean_token_accuracy": 0.18959758132696153,
      "num_tokens": 60913213.0,
      "step": 27805
    },
    {
      "entropy": 5.6577005863189695,
      "epoch": 2.760023818975784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004244505081576521,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1789952978491783,
      "num_tokens": 60924505.0,
      "step": 27810
    },
    {
      "entropy": 5.6088262557983395,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.0348,
      "mean_token_accuracy": 0.18614365309476852,
      "num_tokens": 60934583.0,
      "step": 27815
    },
    {
      "entropy": 5.613558387756347,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042439752830929,
      "loss": 5.0309,
      "mean_token_accuracy": 0.179447890073061,
      "num_tokens": 60944471.0,
      "step": 27820
    },
    {
      "entropy": 5.644979953765869,
      "epoch": 2.761512504962287,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18740174025297165,
      "num_tokens": 60954298.0,
      "step": 27825
    },
    {
      "entropy": 5.63624792098999,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.03125,
      "learning_rate": 0.000424344533641981,
      "loss": 5.0236,
      "mean_token_accuracy": 0.1833224803209305,
      "num_tokens": 60965885.0,
      "step": 27830
    },
    {
      "entropy": 5.597893762588501,
      "epoch": 2.762504962286622,
      "grad_norm": 1.03125,
      "learning_rate": 0.000424318030752864,
      "loss": 5.0063,
      "mean_token_accuracy": 0.17906152307987214,
      "num_tokens": 60977177.0,
      "step": 27835
    },
    {
      "entropy": 5.637393522262573,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.125,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17727639079093932,
      "num_tokens": 60988530.0,
      "step": 27840
    },
    {
      "entropy": 5.630018424987793,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.09375,
      "learning_rate": 0.000424265013866991,
      "loss": 4.9484,
      "mean_token_accuracy": 0.18830775022506713,
      "num_tokens": 60999298.0,
      "step": 27845
    },
    {
      "entropy": 5.605065965652466,
      "epoch": 2.763993648273124,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.0152,
      "mean_token_accuracy": 0.19081733673810958,
      "num_tokens": 61010641.0,
      "step": 27850
    },
    {
      "entropy": 5.644233083724975,
      "epoch": 2.7644898769352917,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004242119821753153,
      "loss": 5.1137,
      "mean_token_accuracy": 0.17999153435230256,
      "num_tokens": 61021597.0,
      "step": 27855
    },
    {
      "entropy": 5.652552700042724,
      "epoch": 2.764986105597459,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004241854607789452,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1829531192779541,
      "num_tokens": 61033507.0,
      "step": 27860
    },
    {
      "entropy": 5.605023527145386,
      "epoch": 2.7654823342596266,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042415893568309714,
      "loss": 4.9386,
      "mean_token_accuracy": 0.18356868028640747,
      "num_tokens": 61044302.0,
      "step": 27865
    },
    {
      "entropy": 5.552216577529907,
      "epoch": 2.765978562921794,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042413240688842884,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18364781588315965,
      "num_tokens": 61056262.0,
      "step": 27870
    },
    {
      "entropy": 5.6073596477508545,
      "epoch": 2.766474791583962,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004241058743955982,
      "loss": 5.0514,
      "mean_token_accuracy": 0.1848957434296608,
      "num_tokens": 61068689.0,
      "step": 27875
    },
    {
      "entropy": 5.648856592178345,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004240793382052633,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1849773868918419,
      "num_tokens": 61080151.0,
      "step": 27880
    },
    {
      "entropy": 5.576858043670654,
      "epoch": 2.767467248908297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000424052798318082,
      "loss": 4.9803,
      "mean_token_accuracy": 0.1823891818523407,
      "num_tokens": 61090488.0,
      "step": 27885
    },
    {
      "entropy": 5.555740690231323,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004240262547347124,
      "loss": 4.9245,
      "mean_token_accuracy": 0.1911711499094963,
      "num_tokens": 61101250.0,
      "step": 27890
    },
    {
      "entropy": 5.5901837825775145,
      "epoch": 2.768459706232632,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042399970745581283,
      "loss": 4.9747,
      "mean_token_accuracy": 0.19075821191072465,
      "num_tokens": 61112920.0,
      "step": 27895
    },
    {
      "entropy": 5.616354036331177,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042397315648204164,
      "loss": 4.9899,
      "mean_token_accuracy": 0.1920622929930687,
      "num_tokens": 61122716.0,
      "step": 27900
    },
    {
      "entropy": 5.609708023071289,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.032,
      "mean_token_accuracy": 0.18635383993387222,
      "num_tokens": 61133794.0,
      "step": 27905
    },
    {
      "entropy": 5.69371280670166,
      "epoch": 2.769948392219135,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.0596,
      "mean_token_accuracy": 0.17820429503917695,
      "num_tokens": 61145690.0,
      "step": 27910
    },
    {
      "entropy": 5.6164751052856445,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004238934813980824,
      "loss": 5.0189,
      "mean_token_accuracy": 0.1791920393705368,
      "num_tokens": 61157111.0,
      "step": 27915
    },
    {
      "entropy": 5.5859208583831785,
      "epoch": 2.7709408495434698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18621775954961778,
      "num_tokens": 61168380.0,
      "step": 27920
    },
    {
      "entropy": 5.677512311935425,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042384034621315773,
      "loss": 5.117,
      "mean_token_accuracy": 0.17591751366853714,
      "num_tokens": 61179318.0,
      "step": 27925
    },
    {
      "entropy": 5.634420394897461,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004238137730839862,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17499317079782487,
      "num_tokens": 61190448.0,
      "step": 27930
    },
    {
      "entropy": 5.579683256149292,
      "epoch": 2.772429535529972,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004237871962645538,
      "loss": 4.9467,
      "mean_token_accuracy": 0.1808662846684456,
      "num_tokens": 61200746.0,
      "step": 27935
    },
    {
      "entropy": 5.58314847946167,
      "epoch": 2.7729257641921397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042376061575551946,
      "loss": 4.9659,
      "mean_token_accuracy": 0.1876770794391632,
      "num_tokens": 61213093.0,
      "step": 27940
    },
    {
      "entropy": 5.573267793655395,
      "epoch": 2.773421992854307,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004237340315575425,
      "loss": 4.9576,
      "mean_token_accuracy": 0.18799831569194794,
      "num_tokens": 61223749.0,
      "step": 27945
    },
    {
      "entropy": 5.618696022033691,
      "epoch": 2.7739182215164746,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004237074436712819,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17017575055360795,
      "num_tokens": 61234755.0,
      "step": 27950
    },
    {
      "entropy": 5.705599117279053,
      "epoch": 2.774414450178642,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17549383342266084,
      "num_tokens": 61244669.0,
      "step": 27955
    },
    {
      "entropy": 5.626323509216308,
      "epoch": 2.77491067884081,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004236542568365478,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18738559782505035,
      "num_tokens": 61257240.0,
      "step": 27960
    },
    {
      "entropy": 5.584547519683838,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042362765788939307,
      "loss": 4.9151,
      "mean_token_accuracy": 0.1890839159488678,
      "num_tokens": 61267389.0,
      "step": 27965
    },
    {
      "entropy": 5.529129362106323,
      "epoch": 2.775903136165145,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.0028,
      "mean_token_accuracy": 0.19548858106136321,
      "num_tokens": 61277381.0,
      "step": 27970
    },
    {
      "entropy": 5.611777305603027,
      "epoch": 2.7763993648273124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042357444893880645,
      "loss": 5.0032,
      "mean_token_accuracy": 0.1814234271645546,
      "num_tokens": 61286931.0,
      "step": 27975
    },
    {
      "entropy": 5.636136484146118,
      "epoch": 2.77689559348948,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042354783893669385,
      "loss": 4.9944,
      "mean_token_accuracy": 0.18543974906206132,
      "num_tokens": 61297706.0,
      "step": 27980
    },
    {
      "entropy": 5.680049800872803,
      "epoch": 2.7773918221516474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.048,
      "mean_token_accuracy": 0.17736678272485734,
      "num_tokens": 61308516.0,
      "step": 27985
    },
    {
      "entropy": 5.634879875183105,
      "epoch": 2.777888050813815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.108,
      "mean_token_accuracy": 0.17719312608242035,
      "num_tokens": 61319454.0,
      "step": 27990
    },
    {
      "entropy": 5.696097087860108,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042346798683099844,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17677246481180192,
      "num_tokens": 61330970.0,
      "step": 27995
    },
    {
      "entropy": 5.627599096298217,
      "epoch": 2.7788805081381502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004234413620981809,
      "loss": 4.931,
      "mean_token_accuracy": 0.18937927186489106,
      "num_tokens": 61342466.0,
      "step": 28000
    },
    {
      "entropy": 5.5982283592224125,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042341473368433735,
      "loss": 4.9556,
      "mean_token_accuracy": 0.19137683063745498,
      "num_tokens": 61351988.0,
      "step": 28005
    },
    {
      "entropy": 5.583123826980591,
      "epoch": 2.779872965462485,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004233881015901282,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18045781701803207,
      "num_tokens": 61364270.0,
      "step": 28010
    },
    {
      "entropy": 5.638596391677856,
      "epoch": 2.7803691941246527,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004233614658162139,
      "loss": 4.9933,
      "mean_token_accuracy": 0.1873441681265831,
      "num_tokens": 61374996.0,
      "step": 28015
    },
    {
      "entropy": 5.559839105606079,
      "epoch": 2.78086542278682,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042333482636325504,
      "loss": 4.9749,
      "mean_token_accuracy": 0.18142868876457213,
      "num_tokens": 61385394.0,
      "step": 28020
    },
    {
      "entropy": 5.599838209152222,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000423308183231912,
      "loss": 5.0199,
      "mean_token_accuracy": 0.1812323585152626,
      "num_tokens": 61397110.0,
      "step": 28025
    },
    {
      "entropy": 5.584038019180298,
      "epoch": 2.781857880111155,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042328153642284567,
      "loss": 5.0469,
      "mean_token_accuracy": 0.18571022301912307,
      "num_tokens": 61408462.0,
      "step": 28030
    },
    {
      "entropy": 5.603032350540161,
      "epoch": 2.7823541087733226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.024,
      "mean_token_accuracy": 0.1844381347298622,
      "num_tokens": 61419543.0,
      "step": 28035
    },
    {
      "entropy": 5.588289785385132,
      "epoch": 2.78285033743549,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000423228231774186,
      "loss": 4.9984,
      "mean_token_accuracy": 0.17940342724323272,
      "num_tokens": 61430476.0,
      "step": 28040
    },
    {
      "entropy": 5.584598445892334,
      "epoch": 2.783346566097658,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004232015739359146,
      "loss": 5.0135,
      "mean_token_accuracy": 0.1883295774459839,
      "num_tokens": 61440384.0,
      "step": 28045
    },
    {
      "entropy": 5.59212737083435,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042317491242256346,
      "loss": 5.0623,
      "mean_token_accuracy": 0.17840179353952407,
      "num_tokens": 61450957.0,
      "step": 28050
    },
    {
      "entropy": 5.563445425033569,
      "epoch": 2.784339023421993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004231482472347938,
      "loss": 5.0483,
      "mean_token_accuracy": 0.1891816347837448,
      "num_tokens": 61462033.0,
      "step": 28055
    },
    {
      "entropy": 5.696949481964111,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.125,
      "learning_rate": 0.0004231215783732668,
      "loss": 5.0901,
      "mean_token_accuracy": 0.179105144739151,
      "num_tokens": 61472153.0,
      "step": 28060
    },
    {
      "entropy": 5.679319572448731,
      "epoch": 2.785331480746328,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042309490583864395,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17984455227851867,
      "num_tokens": 61483836.0,
      "step": 28065
    },
    {
      "entropy": 5.620220279693603,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18606883734464646,
      "num_tokens": 61494819.0,
      "step": 28070
    },
    {
      "entropy": 5.625227260589599,
      "epoch": 2.786323938070663,
      "grad_norm": 1.015625,
      "learning_rate": 0.000423041549752756,
      "loss": 5.0292,
      "mean_token_accuracy": 0.17825910896062852,
      "num_tokens": 61506385.0,
      "step": 28075
    },
    {
      "entropy": 5.614091682434082,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004230148662028141,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18073854446411133,
      "num_tokens": 61517679.0,
      "step": 28080
    },
    {
      "entropy": 5.6015537738800045,
      "epoch": 2.787316395394998,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042298817898242237,
      "loss": 5.0133,
      "mean_token_accuracy": 0.1951308473944664,
      "num_tokens": 61529548.0,
      "step": 28085
    },
    {
      "entropy": 5.616475677490234,
      "epoch": 2.7878126240571657,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1760414570569992,
      "num_tokens": 61540341.0,
      "step": 28090
    },
    {
      "entropy": 5.650433111190796,
      "epoch": 2.788308852719333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.0091,
      "mean_token_accuracy": 0.18754830956459045,
      "num_tokens": 61550692.0,
      "step": 28095
    },
    {
      "entropy": 5.604069948196411,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042290809530516724,
      "loss": 4.9634,
      "mean_token_accuracy": 0.18084899187088013,
      "num_tokens": 61562126.0,
      "step": 28100
    },
    {
      "entropy": 5.578427791595459,
      "epoch": 2.789301310043668,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.0015,
      "mean_token_accuracy": 0.17986824363470078,
      "num_tokens": 61573109.0,
      "step": 28105
    },
    {
      "entropy": 5.646801471710205,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042285468784688363,
      "loss": 5.0674,
      "mean_token_accuracy": 0.1783211499452591,
      "num_tokens": 61584389.0,
      "step": 28110
    },
    {
      "entropy": 5.649658250808716,
      "epoch": 2.790293767368003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.0207,
      "mean_token_accuracy": 0.1861119031906128,
      "num_tokens": 61595173.0,
      "step": 28115
    },
    {
      "entropy": 5.604185342788696,
      "epoch": 2.7907899960301705,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.0609,
      "mean_token_accuracy": 0.18213358521461487,
      "num_tokens": 61605077.0,
      "step": 28120
    },
    {
      "entropy": 5.686222743988037,
      "epoch": 2.791286224692338,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042277454916253186,
      "loss": 5.0815,
      "mean_token_accuracy": 0.18021010756492614,
      "num_tokens": 61617265.0,
      "step": 28125
    },
    {
      "entropy": 5.626056814193726,
      "epoch": 2.791782453354506,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004227478289378838,
      "loss": 5.0388,
      "mean_token_accuracy": 0.18116615563631058,
      "num_tokens": 61629441.0,
      "step": 28130
    },
    {
      "entropy": 5.593112373352051,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.0641,
      "mean_token_accuracy": 0.1777639627456665,
      "num_tokens": 61641032.0,
      "step": 28135
    },
    {
      "entropy": 5.646547603607178,
      "epoch": 2.792774910678841,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042269437749776703,
      "loss": 5.0344,
      "mean_token_accuracy": 0.17825360149145125,
      "num_tokens": 61652625.0,
      "step": 28140
    },
    {
      "entropy": 5.599216890335083,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004226676462836238,
      "loss": 4.9336,
      "mean_token_accuracy": 0.19351912885904313,
      "num_tokens": 61664386.0,
      "step": 28145
    },
    {
      "entropy": 5.6114202499389645,
      "epoch": 2.793767368003176,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004226409114076411,
      "loss": 4.9883,
      "mean_token_accuracy": 0.1854022577404976,
      "num_tokens": 61675040.0,
      "step": 28150
    },
    {
      "entropy": 5.555829954147339,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042261417287048196,
      "loss": 4.9485,
      "mean_token_accuracy": 0.18833045214414595,
      "num_tokens": 61684702.0,
      "step": 28155
    },
    {
      "entropy": 5.565008544921875,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18630063980817796,
      "num_tokens": 61695888.0,
      "step": 28160
    },
    {
      "entropy": 5.655192089080811,
      "epoch": 2.7952560539896787,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042256068481528666,
      "loss": 5.0517,
      "mean_token_accuracy": 0.1825764000415802,
      "num_tokens": 61705655.0,
      "step": 28165
    },
    {
      "entropy": 5.700200176239013,
      "epoch": 2.795752282651846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004225339352985768,
      "loss": 5.0314,
      "mean_token_accuracy": 0.17750136703252792,
      "num_tokens": 61715365.0,
      "step": 28170
    },
    {
      "entropy": 5.65279860496521,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042250718212334334,
      "loss": 5.0932,
      "mean_token_accuracy": 0.179241906106472,
      "num_tokens": 61727104.0,
      "step": 28175
    },
    {
      "entropy": 5.63136920928955,
      "epoch": 2.796744739976181,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042248042529024956,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17806874513626098,
      "num_tokens": 61737000.0,
      "step": 28180
    },
    {
      "entropy": 5.5494965553283695,
      "epoch": 2.7972409686383486,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042245366479995894,
      "loss": 4.9965,
      "mean_token_accuracy": 0.1838792681694031,
      "num_tokens": 61748801.0,
      "step": 28185
    },
    {
      "entropy": 5.593189764022827,
      "epoch": 2.797737197300516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.0185,
      "mean_token_accuracy": 0.18076723963022232,
      "num_tokens": 61760202.0,
      "step": 28190
    },
    {
      "entropy": 5.639090824127197,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042240013285044215,
      "loss": 5.074,
      "mean_token_accuracy": 0.1759035423398018,
      "num_tokens": 61770886.0,
      "step": 28195
    },
    {
      "entropy": 5.6679693222045895,
      "epoch": 2.798729654624851,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042237336139254336,
      "loss": 5.0788,
      "mean_token_accuracy": 0.18128773868083953,
      "num_tokens": 61781464.0,
      "step": 28200
    },
    {
      "entropy": 5.557321691513062,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042234658628010274,
      "loss": 4.9654,
      "mean_token_accuracy": 0.18514203429222106,
      "num_tokens": 61791183.0,
      "step": 28205
    },
    {
      "entropy": 5.59428448677063,
      "epoch": 2.799722111949186,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042231980751378435,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18410906046628953,
      "num_tokens": 61801113.0,
      "step": 28210
    },
    {
      "entropy": 5.663757371902466,
      "epoch": 2.8002183406113534,
      "grad_norm": 1.125,
      "learning_rate": 0.00042229302509425205,
      "loss": 5.1017,
      "mean_token_accuracy": 0.1788233757019043,
      "num_tokens": 61811806.0,
      "step": 28215
    },
    {
      "entropy": 5.700465297698974,
      "epoch": 2.8007145692735214,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.061,
      "mean_token_accuracy": 0.17932450622320176,
      "num_tokens": 61822279.0,
      "step": 28220
    },
    {
      "entropy": 5.587378454208374,
      "epoch": 2.801210797935689,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004222394492982029,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18305806517601014,
      "num_tokens": 61832303.0,
      "step": 28225
    },
    {
      "entropy": 5.6874199390411375,
      "epoch": 2.8017070265978563,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004222126559230146,
      "loss": 5.0629,
      "mean_token_accuracy": 0.17938068360090256,
      "num_tokens": 61844187.0,
      "step": 28230
    },
    {
      "entropy": 5.5321990013122555,
      "epoch": 2.802203255260024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004221858588972696,
      "loss": 4.9562,
      "mean_token_accuracy": 0.18768896758556367,
      "num_tokens": 61855375.0,
      "step": 28235
    },
    {
      "entropy": 5.593571901321411,
      "epoch": 2.8026994839221913,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042215905822163245,
      "loss": 5.0188,
      "mean_token_accuracy": 0.1780177116394043,
      "num_tokens": 61865810.0,
      "step": 28240
    },
    {
      "entropy": 5.59357180595398,
      "epoch": 2.8031957125843587,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.019,
      "mean_token_accuracy": 0.1861756071448326,
      "num_tokens": 61876641.0,
      "step": 28245
    },
    {
      "entropy": 5.64210205078125,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.0314,
      "mean_token_accuracy": 0.18982091844081878,
      "num_tokens": 61887854.0,
      "step": 28250
    },
    {
      "entropy": 5.691206169128418,
      "epoch": 2.804188169908694,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.0548,
      "mean_token_accuracy": 0.18069641292095184,
      "num_tokens": 61898879.0,
      "step": 28255
    },
    {
      "entropy": 5.638670921325684,
      "epoch": 2.8046843985708616,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042205181903345586,
      "loss": 5.0394,
      "mean_token_accuracy": 0.17883214801549913,
      "num_tokens": 61909425.0,
      "step": 28260
    },
    {
      "entropy": 5.604291582107544,
      "epoch": 2.805180627233029,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18831333369016648,
      "num_tokens": 61922097.0,
      "step": 28265
    },
    {
      "entropy": 5.62190260887146,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004219981775572986,
      "loss": 4.969,
      "mean_token_accuracy": 0.18266938477754593,
      "num_tokens": 61932457.0,
      "step": 28270
    },
    {
      "entropy": 5.591121053695678,
      "epoch": 2.806173084557364,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18949588835239412,
      "num_tokens": 61942354.0,
      "step": 28275
    },
    {
      "entropy": 5.604897451400757,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042194452150018917,
      "loss": 5.0393,
      "mean_token_accuracy": 0.18269131630659102,
      "num_tokens": 61952026.0,
      "step": 28280
    },
    {
      "entropy": 5.567713403701783,
      "epoch": 2.807165541881699,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042191768800544055,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18043863475322724,
      "num_tokens": 61961976.0,
      "step": 28285
    },
    {
      "entropy": 5.5813782691955565,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.125,
      "learning_rate": 0.00042189085086744986,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18743563443422318,
      "num_tokens": 61974529.0,
      "step": 28290
    },
    {
      "entropy": 5.618631172180176,
      "epoch": 2.808157999206034,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004218640100868826,
      "loss": 5.0806,
      "mean_token_accuracy": 0.18082202076911927,
      "num_tokens": 61985912.0,
      "step": 28295
    },
    {
      "entropy": 5.620187568664551,
      "epoch": 2.8086542278682014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17783131897449495,
      "num_tokens": 61997534.0,
      "step": 28300
    },
    {
      "entropy": 5.6106727600097654,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004218103176006808,
      "loss": 4.9365,
      "mean_token_accuracy": 0.1827211007475853,
      "num_tokens": 62008292.0,
      "step": 28305
    },
    {
      "entropy": 5.5646195888519285,
      "epoch": 2.809646685192537,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004217834658963777,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18329723924398422,
      "num_tokens": 62019005.0,
      "step": 28310
    },
    {
      "entropy": 5.595174503326416,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042175661055216093,
      "loss": 5.0798,
      "mean_token_accuracy": 0.17932990938425064,
      "num_tokens": 62030127.0,
      "step": 28315
    },
    {
      "entropy": 5.5903013229370115,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004217297515686965,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17378199845552444,
      "num_tokens": 62041080.0,
      "step": 28320
    },
    {
      "entropy": 5.5953977584838865,
      "epoch": 2.811135371179039,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004217028889466503,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18346406668424606,
      "num_tokens": 62051420.0,
      "step": 28325
    },
    {
      "entropy": 5.522453117370605,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004216760226866887,
      "loss": 4.9221,
      "mean_token_accuracy": 0.18328099697828293,
      "num_tokens": 62063303.0,
      "step": 28330
    },
    {
      "entropy": 5.647215032577515,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.0068,
      "mean_token_accuracy": 0.18482659757137299,
      "num_tokens": 62074654.0,
      "step": 28335
    },
    {
      "entropy": 5.592301845550537,
      "epoch": 2.812624057165542,
      "grad_norm": 1.125,
      "learning_rate": 0.0004216222792556838,
      "loss": 4.9134,
      "mean_token_accuracy": 0.19437462389469146,
      "num_tokens": 62085872.0,
      "step": 28340
    },
    {
      "entropy": 5.5622529029846195,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004215954020859732,
      "loss": 4.9741,
      "mean_token_accuracy": 0.1896725133061409,
      "num_tokens": 62095185.0,
      "step": 28345
    },
    {
      "entropy": 5.635363006591797,
      "epoch": 2.813616514489877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042156852128101264,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18732836842536926,
      "num_tokens": 62105643.0,
      "step": 28350
    },
    {
      "entropy": 5.576270294189453,
      "epoch": 2.8141127431520445,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042154163684146855,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18611236959695815,
      "num_tokens": 62115290.0,
      "step": 28355
    },
    {
      "entropy": 5.593039321899414,
      "epoch": 2.814608971814212,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004215147487680076,
      "loss": 4.9826,
      "mean_token_accuracy": 0.1892538160085678,
      "num_tokens": 62125493.0,
      "step": 28360
    },
    {
      "entropy": 5.629483842849732,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.0342,
      "mean_token_accuracy": 0.17796495854854583,
      "num_tokens": 62135813.0,
      "step": 28365
    },
    {
      "entropy": 5.560048961639405,
      "epoch": 2.815601429138547,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004214609617220024,
      "loss": 4.9662,
      "mean_token_accuracy": 0.18972467482089997,
      "num_tokens": 62146423.0,
      "step": 28370
    },
    {
      "entropy": 5.593180847167969,
      "epoch": 2.8160976578007144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042143406275079186,
      "loss": 5.0826,
      "mean_token_accuracy": 0.18352461904287337,
      "num_tokens": 62157601.0,
      "step": 28375
    },
    {
      "entropy": 5.633260822296142,
      "epoch": 2.816593886462882,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004214071601483322,
      "loss": 5.1026,
      "mean_token_accuracy": 0.1784863993525505,
      "num_tokens": 62167918.0,
      "step": 28380
    },
    {
      "entropy": 5.653325366973877,
      "epoch": 2.8170901151250494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004213802539152905,
      "loss": 5.0027,
      "mean_token_accuracy": 0.1918745756149292,
      "num_tokens": 62179176.0,
      "step": 28385
    },
    {
      "entropy": 5.637861537933349,
      "epoch": 2.8175863437872173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.015,
      "mean_token_accuracy": 0.1938460201025009,
      "num_tokens": 62190238.0,
      "step": 28390
    },
    {
      "entropy": 5.605133008956909,
      "epoch": 2.8180825724493848,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004213264305601296,
      "loss": 5.0026,
      "mean_token_accuracy": 0.1814677193760872,
      "num_tokens": 62202041.0,
      "step": 28395
    },
    {
      "entropy": 5.604564142227173,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042129951343934527,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18168165534734726,
      "num_tokens": 62213309.0,
      "step": 28400
    },
    {
      "entropy": 5.655052328109742,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.0,
      "learning_rate": 0.0004212725926906482,
      "loss": 5.0935,
      "mean_token_accuracy": 0.18357770442962645,
      "num_tokens": 62223936.0,
      "step": 28405
    },
    {
      "entropy": 5.60878005027771,
      "epoch": 2.819571258435887,
      "grad_norm": 1.125,
      "learning_rate": 0.00042124566831470604,
      "loss": 4.9896,
      "mean_token_accuracy": 0.18787377029657365,
      "num_tokens": 62233800.0,
      "step": 28410
    },
    {
      "entropy": 5.671301555633545,
      "epoch": 2.8200674870980547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004212187403121864,
      "loss": 5.031,
      "mean_token_accuracy": 0.18323247730731965,
      "num_tokens": 62245888.0,
      "step": 28415
    },
    {
      "entropy": 5.5703269958496096,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042119180868375707,
      "loss": 4.9724,
      "mean_token_accuracy": 0.18776115477085115,
      "num_tokens": 62255796.0,
      "step": 28420
    },
    {
      "entropy": 5.637016439437867,
      "epoch": 2.82105994442239,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004211648734300859,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17691047340631486,
      "num_tokens": 62266463.0,
      "step": 28425
    },
    {
      "entropy": 5.584423780441284,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042113793455184084,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18952811062335967,
      "num_tokens": 62278077.0,
      "step": 28430
    },
    {
      "entropy": 5.638034057617188,
      "epoch": 2.822052401746725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.0784,
      "mean_token_accuracy": 0.18724767863750458,
      "num_tokens": 62289489.0,
      "step": 28435
    },
    {
      "entropy": 5.545647668838501,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042108404592430126,
      "loss": 4.94,
      "mean_token_accuracy": 0.19896889477968216,
      "num_tokens": 62301402.0,
      "step": 28440
    },
    {
      "entropy": 5.567494916915893,
      "epoch": 2.82304485907106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000421057096176343,
      "loss": 4.9066,
      "mean_token_accuracy": 0.18680199533700942,
      "num_tokens": 62311253.0,
      "step": 28445
    },
    {
      "entropy": 5.603012990951538,
      "epoch": 2.8235410877332274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004210301428064835,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18341346830129623,
      "num_tokens": 62321312.0,
      "step": 28450
    },
    {
      "entropy": 5.615802383422851,
      "epoch": 2.824037316395395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.0402,
      "mean_token_accuracy": 0.17836439460515977,
      "num_tokens": 62332318.0,
      "step": 28455
    },
    {
      "entropy": 5.682572317123413,
      "epoch": 2.8245335450575624,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.0321,
      "mean_token_accuracy": 0.18632835596799852,
      "num_tokens": 62341945.0,
      "step": 28460
    },
    {
      "entropy": 5.613305139541626,
      "epoch": 2.82502977371973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18756728917360305,
      "num_tokens": 62353509.0,
      "step": 28465
    },
    {
      "entropy": 5.584838390350342,
      "epoch": 2.8255260023818973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004209222931214019,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18419019430875777,
      "num_tokens": 62364044.0,
      "step": 28470
    },
    {
      "entropy": 5.542319965362549,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042089532165206367,
      "loss": 4.9168,
      "mean_token_accuracy": 0.18852000832557678,
      "num_tokens": 62375113.0,
      "step": 28475
    },
    {
      "entropy": 5.585206842422485,
      "epoch": 2.8265184597062327,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004208683465648358,
      "loss": 4.9825,
      "mean_token_accuracy": 0.1883363351225853,
      "num_tokens": 62387077.0,
      "step": 28480
    },
    {
      "entropy": 5.651433658599854,
      "epoch": 2.8270146883684,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042084136786038737,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18614548891782762,
      "num_tokens": 62397320.0,
      "step": 28485
    },
    {
      "entropy": 5.606999826431275,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004208143855393872,
      "loss": 4.9718,
      "mean_token_accuracy": 0.1944418117403984,
      "num_tokens": 62407675.0,
      "step": 28490
    },
    {
      "entropy": 5.585662460327148,
      "epoch": 2.828007145692735,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042078739960250446,
      "loss": 4.9695,
      "mean_token_accuracy": 0.19465111047029496,
      "num_tokens": 62417969.0,
      "step": 28495
    },
    {
      "entropy": 5.62890887260437,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042076041005040835,
      "loss": 4.9653,
      "mean_token_accuracy": 0.19183972328901291,
      "num_tokens": 62428761.0,
      "step": 28500
    },
    {
      "entropy": 5.578619956970215,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.125,
      "learning_rate": 0.0004207334168837682,
      "loss": 4.9647,
      "mean_token_accuracy": 0.1802323803305626,
      "num_tokens": 62438425.0,
      "step": 28505
    },
    {
      "entropy": 5.563123512268066,
      "epoch": 2.829495831679238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004207064201032533,
      "loss": 4.9264,
      "mean_token_accuracy": 0.19015867859125138,
      "num_tokens": 62448819.0,
      "step": 28510
    },
    {
      "entropy": 5.64803524017334,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042067941970953316,
      "loss": 5.0512,
      "mean_token_accuracy": 0.18327915966510772,
      "num_tokens": 62458110.0,
      "step": 28515
    },
    {
      "entropy": 5.590819072723389,
      "epoch": 2.830488289003573,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18665423095226288,
      "num_tokens": 62468158.0,
      "step": 28520
    },
    {
      "entropy": 5.626048755645752,
      "epoch": 2.8309845176657404,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.1131,
      "mean_token_accuracy": 0.17949173003435134,
      "num_tokens": 62479926.0,
      "step": 28525
    },
    {
      "entropy": 5.683513355255127,
      "epoch": 2.831480746327908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042059839685583726,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1810275673866272,
      "num_tokens": 62490012.0,
      "step": 28530
    },
    {
      "entropy": 5.62437629699707,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18756619542837144,
      "num_tokens": 62501268.0,
      "step": 28535
    },
    {
      "entropy": 5.578715848922729,
      "epoch": 2.832473203652243,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042054436356629113,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18461152166128159,
      "num_tokens": 62513117.0,
      "step": 28540
    },
    {
      "entropy": 5.6718353748321535,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004205173415074032,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17637509554624559,
      "num_tokens": 62524673.0,
      "step": 28545
    },
    {
      "entropy": 5.589991855621338,
      "epoch": 2.833465660976578,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004204903158399989,
      "loss": 4.9676,
      "mean_token_accuracy": 0.18218406587839125,
      "num_tokens": 62536336.0,
      "step": 28550
    },
    {
      "entropy": 5.667876720428467,
      "epoch": 2.8339618896387453,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.1477,
      "mean_token_accuracy": 0.17582537233829498,
      "num_tokens": 62546976.0,
      "step": 28555
    },
    {
      "entropy": 5.58868727684021,
      "epoch": 2.834458118300913,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204362536823213,
      "loss": 4.9286,
      "mean_token_accuracy": 0.19463050067424775,
      "num_tokens": 62558081.0,
      "step": 28560
    },
    {
      "entropy": 5.639292526245117,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042040921719338877,
      "loss": 4.964,
      "mean_token_accuracy": 0.1850041463971138,
      "num_tokens": 62568909.0,
      "step": 28565
    },
    {
      "entropy": 5.614081335067749,
      "epoch": 2.835450575625248,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000420382177098621,
      "loss": 4.9752,
      "mean_token_accuracy": 0.19236454665660857,
      "num_tokens": 62579765.0,
      "step": 28570
    },
    {
      "entropy": 5.590819883346557,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18093298822641374,
      "num_tokens": 62591150.0,
      "step": 28575
    },
    {
      "entropy": 5.78311939239502,
      "epoch": 2.836443032949583,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.1291,
      "mean_token_accuracy": 0.17858947366476058,
      "num_tokens": 62601582.0,
      "step": 28580
    },
    {
      "entropy": 5.6497279644012455,
      "epoch": 2.8369392616117506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004203010351860119,
      "loss": 4.9229,
      "mean_token_accuracy": 0.18980394899845124,
      "num_tokens": 62612381.0,
      "step": 28585
    },
    {
      "entropy": 5.643128395080566,
      "epoch": 2.837435490273918,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17863920331001282,
      "num_tokens": 62623302.0,
      "step": 28590
    },
    {
      "entropy": 5.649317598342895,
      "epoch": 2.837931718936086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.0209,
      "mean_token_accuracy": 0.17231608778238297,
      "num_tokens": 62633850.0,
      "step": 28595
    },
    {
      "entropy": 5.672271680831909,
      "epoch": 2.8384279475982535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.0378,
      "mean_token_accuracy": 0.18198233097791672,
      "num_tokens": 62644869.0,
      "step": 28600
    },
    {
      "entropy": 5.620412731170655,
      "epoch": 2.838924176260421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004201927955281957,
      "loss": 5.0348,
      "mean_token_accuracy": 0.18181626200675965,
      "num_tokens": 62655553.0,
      "step": 28605
    },
    {
      "entropy": 5.59302716255188,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201657266108926,
      "loss": 4.9908,
      "mean_token_accuracy": 0.18730361610651017,
      "num_tokens": 62667043.0,
      "step": 28610
    },
    {
      "entropy": 5.642810678482055,
      "epoch": 2.839916633584756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042013865409379244,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18233148008584976,
      "num_tokens": 62678081.0,
      "step": 28615
    },
    {
      "entropy": 5.640724182128906,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.0347,
      "mean_token_accuracy": 0.18424116373062133,
      "num_tokens": 62689101.0,
      "step": 28620
    },
    {
      "entropy": 5.610619115829468,
      "epoch": 2.840909090909091,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004200844982628857,
      "loss": 4.9777,
      "mean_token_accuracy": 0.18921141624450682,
      "num_tokens": 62698584.0,
      "step": 28625
    },
    {
      "entropy": 5.595674610137939,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.0174,
      "mean_token_accuracy": 0.1861177161335945,
      "num_tokens": 62709936.0,
      "step": 28630
    },
    {
      "entropy": 5.603438901901245,
      "epoch": 2.841901548233426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004200303280408474,
      "loss": 4.9869,
      "mean_token_accuracy": 0.18629567921161652,
      "num_tokens": 62720926.0,
      "step": 28635
    },
    {
      "entropy": 5.609008646011352,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042000323753483296,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18533147722482682,
      "num_tokens": 62732221.0,
      "step": 28640
    },
    {
      "entropy": 5.599571609497071,
      "epoch": 2.842894005557761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041997614343305067,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18195525854825972,
      "num_tokens": 62743579.0,
      "step": 28645
    },
    {
      "entropy": 5.626251983642578,
      "epoch": 2.8433902342199286,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041994904573617247,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18612316995859146,
      "num_tokens": 62754246.0,
      "step": 28650
    },
    {
      "entropy": 5.650132703781128,
      "epoch": 2.843886462882096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041992194444487026,
      "loss": 5.129,
      "mean_token_accuracy": 0.18083446472883224,
      "num_tokens": 62765813.0,
      "step": 28655
    },
    {
      "entropy": 5.6629319190979,
      "epoch": 2.8443826915442636,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041989483955981607,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17962580174207687,
      "num_tokens": 62777208.0,
      "step": 28660
    },
    {
      "entropy": 5.626197719573975,
      "epoch": 2.844878920206431,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041986773108168216,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1779976010322571,
      "num_tokens": 62787549.0,
      "step": 28665
    },
    {
      "entropy": 5.589355230331421,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041984061901114057,
      "loss": 4.9551,
      "mean_token_accuracy": 0.1980505809187889,
      "num_tokens": 62798609.0,
      "step": 28670
    },
    {
      "entropy": 5.599371099472046,
      "epoch": 2.845871377530766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004198135033488638,
      "loss": 4.9792,
      "mean_token_accuracy": 0.184997995197773,
      "num_tokens": 62809705.0,
      "step": 28675
    },
    {
      "entropy": 5.675292205810547,
      "epoch": 2.846367606192934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041978638409552417,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1859294131398201,
      "num_tokens": 62820740.0,
      "step": 28680
    },
    {
      "entropy": 5.679944133758545,
      "epoch": 2.8468638348551014,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.0997,
      "mean_token_accuracy": 0.1828433394432068,
      "num_tokens": 62832294.0,
      "step": 28685
    },
    {
      "entropy": 5.556138229370117,
      "epoch": 2.847360063517269,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004197321348183465,
      "loss": 4.9269,
      "mean_token_accuracy": 0.1867649883031845,
      "num_tokens": 62843174.0,
      "step": 28690
    },
    {
      "entropy": 5.605006170272827,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.0324,
      "mean_token_accuracy": 0.18635590523481368,
      "num_tokens": 62852983.0,
      "step": 28695
    },
    {
      "entropy": 5.548183488845825,
      "epoch": 2.848352520841604,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041967787118498854,
      "loss": 4.9791,
      "mean_token_accuracy": 0.18597790598869324,
      "num_tokens": 62863587.0,
      "step": 28700
    },
    {
      "entropy": 5.645437097549438,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004196507339864241,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17715835571289062,
      "num_tokens": 62874938.0,
      "step": 28705
    },
    {
      "entropy": 5.607238101959228,
      "epoch": 2.849344978165939,
      "grad_norm": 1.046875,
      "learning_rate": 0.000419623593200833,
      "loss": 4.9516,
      "mean_token_accuracy": 0.1852348566055298,
      "num_tokens": 62885987.0,
      "step": 28710
    },
    {
      "entropy": 5.553633975982666,
      "epoch": 2.8498412068281063,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041959644882888844,
      "loss": 4.9351,
      "mean_token_accuracy": 0.18382916301488877,
      "num_tokens": 62896827.0,
      "step": 28715
    },
    {
      "entropy": 5.663754606246949,
      "epoch": 2.8503374354902737,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041956930087126357,
      "loss": 5.0188,
      "mean_token_accuracy": 0.1850905954837799,
      "num_tokens": 62906596.0,
      "step": 28720
    },
    {
      "entropy": 5.579660511016845,
      "epoch": 2.850833664152441,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041954214932863156,
      "loss": 4.9787,
      "mean_token_accuracy": 0.1862029567360878,
      "num_tokens": 62918773.0,
      "step": 28725
    },
    {
      "entropy": 5.643328905105591,
      "epoch": 2.851329892814609,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004195149942016657,
      "loss": 5.0021,
      "mean_token_accuracy": 0.18715063035488128,
      "num_tokens": 62929395.0,
      "step": 28730
    },
    {
      "entropy": 5.609548044204712,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004194878354910394,
      "loss": 4.9681,
      "mean_token_accuracy": 0.18341030031442643,
      "num_tokens": 62939905.0,
      "step": 28735
    },
    {
      "entropy": 5.6377194881439205,
      "epoch": 2.852322350138944,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.001,
      "mean_token_accuracy": 0.1888953909277916,
      "num_tokens": 62950684.0,
      "step": 28740
    },
    {
      "entropy": 5.566971254348755,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041943350732149935,
      "loss": 4.9642,
      "mean_token_accuracy": 0.18864668309688568,
      "num_tokens": 62960754.0,
      "step": 28745
    },
    {
      "entropy": 5.555545282363892,
      "epoch": 2.853314807463279,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.9575,
      "mean_token_accuracy": 0.1817317858338356,
      "num_tokens": 62970659.0,
      "step": 28750
    },
    {
      "entropy": 5.581815481185913,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041937916482540045,
      "loss": 4.9263,
      "mean_token_accuracy": 0.1906468838453293,
      "num_tokens": 62981135.0,
      "step": 28755
    },
    {
      "entropy": 5.623519325256348,
      "epoch": 2.854307264787614,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041935198820657575,
      "loss": 4.9851,
      "mean_token_accuracy": 0.18814168870449066,
      "num_tokens": 62991979.0,
      "step": 28760
    },
    {
      "entropy": 5.626046562194825,
      "epoch": 2.854803493449782,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004193248080081329,
      "loss": 5.0336,
      "mean_token_accuracy": 0.18453649431467056,
      "num_tokens": 63003180.0,
      "step": 28765
    },
    {
      "entropy": 5.6430517673492435,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004192976242307457,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18547267615795135,
      "num_tokens": 63014529.0,
      "step": 28770
    },
    {
      "entropy": 5.59978461265564,
      "epoch": 2.855795950774117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004192704368750884,
      "loss": 5.0364,
      "mean_token_accuracy": 0.17678449302911758,
      "num_tokens": 63025293.0,
      "step": 28775
    },
    {
      "entropy": 5.612225294113159,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.0358,
      "mean_token_accuracy": 0.18477218598127365,
      "num_tokens": 63036210.0,
      "step": 28780
    },
    {
      "entropy": 5.650445461273193,
      "epoch": 2.856788408098452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17990680187940597,
      "num_tokens": 63047212.0,
      "step": 28785
    },
    {
      "entropy": 5.625139760971069,
      "epoch": 2.8572846367606193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004191888533452379,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18436428904533386,
      "num_tokens": 63058398.0,
      "step": 28790
    },
    {
      "entropy": 5.645731639862061,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.0459,
      "mean_token_accuracy": 0.1827764555811882,
      "num_tokens": 63070099.0,
      "step": 28795
    },
    {
      "entropy": 5.588117933273315,
      "epoch": 2.8582770940849542,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004191344464463494,
      "loss": 4.9737,
      "mean_token_accuracy": 0.19065602272748947,
      "num_tokens": 63081743.0,
      "step": 28800
    },
    {
      "entropy": 5.629089641571045,
      "epoch": 2.8587733227471217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.0447,
      "mean_token_accuracy": 0.1819518968462944,
      "num_tokens": 63092548.0,
      "step": 28805
    },
    {
      "entropy": 5.627474784851074,
      "epoch": 2.859269551409289,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.0545,
      "mean_token_accuracy": 0.18376210778951646,
      "num_tokens": 63102765.0,
      "step": 28810
    },
    {
      "entropy": 5.5518045902252195,
      "epoch": 2.8597657800714567,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004190528092930259,
      "loss": 4.9423,
      "mean_token_accuracy": 0.18979331254959106,
      "num_tokens": 63113847.0,
      "step": 28815
    },
    {
      "entropy": 5.613998937606811,
      "epoch": 2.8602620087336246,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004190255897632863,
      "loss": 5.014,
      "mean_token_accuracy": 0.18149924725294114,
      "num_tokens": 63125288.0,
      "step": 28820
    },
    {
      "entropy": 5.654726362228393,
      "epoch": 2.860758237395792,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18603657633066178,
      "num_tokens": 63136590.0,
      "step": 28825
    },
    {
      "entropy": 5.658866262435913,
      "epoch": 2.8612544660579595,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004189711399899095,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17745965868234634,
      "num_tokens": 63146711.0,
      "step": 28830
    },
    {
      "entropy": 5.635267210006714,
      "epoch": 2.861750694720127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004189439097476224,
      "loss": 5.0626,
      "mean_token_accuracy": 0.18137509375810623,
      "num_tokens": 63157805.0,
      "step": 28835
    },
    {
      "entropy": 5.601698398590088,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004189166759358366,
      "loss": 4.9835,
      "mean_token_accuracy": 0.19347959458827974,
      "num_tokens": 63167891.0,
      "step": 28840
    },
    {
      "entropy": 5.5842784404754635,
      "epoch": 2.862743152044462,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041888943855522727,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19033675044775009,
      "num_tokens": 63179438.0,
      "step": 28845
    },
    {
      "entropy": 5.6138160705566404,
      "epoch": 2.86323938070663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041886219760646997,
      "loss": 5.0326,
      "mean_token_accuracy": 0.17887755781412124,
      "num_tokens": 63190234.0,
      "step": 28850
    },
    {
      "entropy": 5.597001218795777,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004188349530902402,
      "loss": 4.9941,
      "mean_token_accuracy": 0.17916476726531982,
      "num_tokens": 63201880.0,
      "step": 28855
    },
    {
      "entropy": 5.553423881530762,
      "epoch": 2.864231838030965,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041880770500721346,
      "loss": 4.8868,
      "mean_token_accuracy": 0.1969156265258789,
      "num_tokens": 63211906.0,
      "step": 28860
    },
    {
      "entropy": 5.613456106185913,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004187804533580655,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18456721007823945,
      "num_tokens": 63223365.0,
      "step": 28865
    },
    {
      "entropy": 5.6545624256134035,
      "epoch": 2.8652242953552998,
      "grad_norm": 1.125,
      "learning_rate": 0.0004187531981434721,
      "loss": 5.0975,
      "mean_token_accuracy": 0.17728629410266877,
      "num_tokens": 63233849.0,
      "step": 28870
    },
    {
      "entropy": 5.651711893081665,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.0672,
      "mean_token_accuracy": 0.17656503319740297,
      "num_tokens": 63245742.0,
      "step": 28875
    },
    {
      "entropy": 5.6336956977844235,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18608967810869217,
      "num_tokens": 63257284.0,
      "step": 28880
    },
    {
      "entropy": 5.6447642803192135,
      "epoch": 2.866712981341802,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004186714111137787,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17924389839172364,
      "num_tokens": 63268893.0,
      "step": 28885
    },
    {
      "entropy": 5.685910272598266,
      "epoch": 2.8672092100039697,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18164164274930955,
      "num_tokens": 63281763.0,
      "step": 28890
    },
    {
      "entropy": 5.639322662353516,
      "epoch": 2.867705438666137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18290508538484573,
      "num_tokens": 63292304.0,
      "step": 28895
    },
    {
      "entropy": 5.691169261932373,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.0,
      "learning_rate": 0.00041858959201941307,
      "loss": 5.1102,
      "mean_token_accuracy": 0.1814054161310196,
      "num_tokens": 63303539.0,
      "step": 28900
    },
    {
      "entropy": 5.674127578735352,
      "epoch": 2.8686978959904725,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.0652,
      "mean_token_accuracy": 0.17995190918445586,
      "num_tokens": 63313678.0,
      "step": 28905
    },
    {
      "entropy": 5.633136463165283,
      "epoch": 2.86919412465264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041853502815181306,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18277678191661834,
      "num_tokens": 63324988.0,
      "step": 28910
    },
    {
      "entropy": 5.702474546432495,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004185077408786357,
      "loss": 5.1147,
      "mean_token_accuracy": 0.18060988783836365,
      "num_tokens": 63335325.0,
      "step": 28915
    },
    {
      "entropy": 5.661550998687744,
      "epoch": 2.870186581976975,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004184804500467756,
      "loss": 5.0479,
      "mean_token_accuracy": 0.18008514791727065,
      "num_tokens": 63346838.0,
      "step": 28920
    },
    {
      "entropy": 5.665519332885742,
      "epoch": 2.8706828106391424,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041845315565690945,
      "loss": 5.0769,
      "mean_token_accuracy": 0.17956572622060776,
      "num_tokens": 63356752.0,
      "step": 28925
    },
    {
      "entropy": 5.695985651016235,
      "epoch": 2.87117903930131,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041842585770971417,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1818717509508133,
      "num_tokens": 63367464.0,
      "step": 28930
    },
    {
      "entropy": 5.57865629196167,
      "epoch": 2.871675267963478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004183985562058667,
      "loss": 5.014,
      "mean_token_accuracy": 0.18362112939357758,
      "num_tokens": 63377986.0,
      "step": 28935
    },
    {
      "entropy": 5.618691110610962,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.03125,
      "learning_rate": 0.000418371251146044,
      "loss": 5.0408,
      "mean_token_accuracy": 0.1838704153895378,
      "num_tokens": 63391026.0,
      "step": 28940
    },
    {
      "entropy": 5.632062387466431,
      "epoch": 2.872667725287813,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004183439425309232,
      "loss": 5.0648,
      "mean_token_accuracy": 0.18163959234952926,
      "num_tokens": 63401946.0,
      "step": 28945
    },
    {
      "entropy": 5.583399915695191,
      "epoch": 2.8731639539499803,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004183166303611816,
      "loss": 4.9602,
      "mean_token_accuracy": 0.1849286213517189,
      "num_tokens": 63412800.0,
      "step": 28950
    },
    {
      "entropy": 5.609708452224732,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004182893146374963,
      "loss": 4.9453,
      "mean_token_accuracy": 0.185700486600399,
      "num_tokens": 63423153.0,
      "step": 28955
    },
    {
      "entropy": 5.572841453552246,
      "epoch": 2.874156411274315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041826199536054476,
      "loss": 4.9133,
      "mean_token_accuracy": 0.1901962012052536,
      "num_tokens": 63433115.0,
      "step": 28960
    },
    {
      "entropy": 5.613576126098633,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.0503,
      "mean_token_accuracy": 0.1761757329106331,
      "num_tokens": 63443215.0,
      "step": 28965
    },
    {
      "entropy": 5.592602682113648,
      "epoch": 2.87514886859865,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041820734614955295,
      "loss": 4.9631,
      "mean_token_accuracy": 0.18091457039117814,
      "num_tokens": 63453619.0,
      "step": 28970
    },
    {
      "entropy": 5.631352233886719,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004181800162168678,
      "loss": 5.0919,
      "mean_token_accuracy": 0.18049390763044357,
      "num_tokens": 63464284.0,
      "step": 28975
    },
    {
      "entropy": 5.603013706207276,
      "epoch": 2.876141325922985,
      "grad_norm": 1.25,
      "learning_rate": 0.0004181526827336268,
      "loss": 4.9205,
      "mean_token_accuracy": 0.19253764897584916,
      "num_tokens": 63474318.0,
      "step": 28980
    },
    {
      "entropy": 5.583526754379273,
      "epoch": 2.8766375545851526,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004181253457005077,
      "loss": 4.9808,
      "mean_token_accuracy": 0.19531305134296417,
      "num_tokens": 63484179.0,
      "step": 28985
    },
    {
      "entropy": 5.6808678150177006,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041809800511818853,
      "loss": 5.0361,
      "mean_token_accuracy": 0.18322037905454636,
      "num_tokens": 63494876.0,
      "step": 28990
    },
    {
      "entropy": 5.74129695892334,
      "epoch": 2.877630011909488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.1602,
      "mean_token_accuracy": 0.18333101719617845,
      "num_tokens": 63505563.0,
      "step": 28995
    },
    {
      "entropy": 5.650716543197632,
      "epoch": 2.8781262405716554,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.1121,
      "mean_token_accuracy": 0.17352036982774735,
      "num_tokens": 63517493.0,
      "step": 29000
    },
    {
      "entropy": 5.702355813980103,
      "epoch": 2.878622469233823,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1844149872660637,
      "num_tokens": 63528187.0,
      "step": 29005
    },
    {
      "entropy": 5.710976552963257,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004179886073104709,
      "loss": 5.1489,
      "mean_token_accuracy": 0.17933813929557801,
      "num_tokens": 63539297.0,
      "step": 29010
    },
    {
      "entropy": 5.652044057846069,
      "epoch": 2.879614926558158,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004179612489923224,
      "loss": 5.072,
      "mean_token_accuracy": 0.17842003256082534,
      "num_tokens": 63551162.0,
      "step": 29015
    },
    {
      "entropy": 5.623454475402832,
      "epoch": 2.880111155220326,
      "grad_norm": 1.078125,
      "learning_rate": 0.000417933887129043,
      "loss": 5.0354,
      "mean_token_accuracy": 0.18050960451364517,
      "num_tokens": 63561472.0,
      "step": 29020
    },
    {
      "entropy": 5.6400000095367435,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004179065217213111,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1790015235543251,
      "num_tokens": 63571682.0,
      "step": 29025
    },
    {
      "entropy": 5.629941606521607,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18337429612874984,
      "num_tokens": 63583288.0,
      "step": 29030
    },
    {
      "entropy": 5.648037767410278,
      "epoch": 2.881599841206828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041785178027520486,
      "loss": 5.0413,
      "mean_token_accuracy": 0.17945483922958375,
      "num_tokens": 63594774.0,
      "step": 29035
    },
    {
      "entropy": 5.650640153884888,
      "epoch": 2.8820960698689957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.0466,
      "mean_token_accuracy": 0.18706305027008058,
      "num_tokens": 63605156.0,
      "step": 29040
    },
    {
      "entropy": 5.617401123046875,
      "epoch": 2.882592298531163,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004177970246594332,
      "loss": 4.955,
      "mean_token_accuracy": 0.18468340784311293,
      "num_tokens": 63615224.0,
      "step": 29045
    },
    {
      "entropy": 5.646003532409668,
      "epoch": 2.8830885271933306,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18342808037996292,
      "num_tokens": 63625566.0,
      "step": 29050
    },
    {
      "entropy": 5.693893671035767,
      "epoch": 2.883584755855498,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041774225487942765,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17596696615219115,
      "num_tokens": 63635697.0,
      "step": 29055
    },
    {
      "entropy": 5.6371434211730955,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.0153,
      "mean_token_accuracy": 0.18857091814279556,
      "num_tokens": 63646377.0,
      "step": 29060
    },
    {
      "entropy": 5.639778470993042,
      "epoch": 2.884577213179833,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.955,
      "mean_token_accuracy": 0.1885196104645729,
      "num_tokens": 63657356.0,
      "step": 29065
    },
    {
      "entropy": 5.6855333805084225,
      "epoch": 2.8850734418420005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.0759,
      "mean_token_accuracy": 0.1873919203877449,
      "num_tokens": 63667267.0,
      "step": 29070
    },
    {
      "entropy": 5.5532244682312015,
      "epoch": 2.8855696705041685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041763267284844675,
      "loss": 5.0152,
      "mean_token_accuracy": 0.1854139730334282,
      "num_tokens": 63678242.0,
      "step": 29075
    },
    {
      "entropy": 5.670841932296753,
      "epoch": 2.886065899166336,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004176052684965456,
      "loss": 5.0212,
      "mean_token_accuracy": 0.1847448691725731,
      "num_tokens": 63689757.0,
      "step": 29080
    },
    {
      "entropy": 5.636821603775024,
      "epoch": 2.8865621278285034,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004175778606083411,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1851016029715538,
      "num_tokens": 63701196.0,
      "step": 29085
    },
    {
      "entropy": 5.607085704803467,
      "epoch": 2.887058356490671,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1788066267967224,
      "num_tokens": 63712802.0,
      "step": 29090
    },
    {
      "entropy": 5.652802324295044,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18072767853736876,
      "num_tokens": 63723543.0,
      "step": 29095
    },
    {
      "entropy": 5.597383069992065,
      "epoch": 2.888050813815006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.0462,
      "mean_token_accuracy": 0.18113614916801452,
      "num_tokens": 63734954.0,
      "step": 29100
    },
    {
      "entropy": 5.664287185668945,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.0,
      "learning_rate": 0.0004174681937060837,
      "loss": 5.0634,
      "mean_token_accuracy": 0.18343387693166732,
      "num_tokens": 63746991.0,
      "step": 29105
    },
    {
      "entropy": 5.611398458480835,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.125,
      "learning_rate": 0.00041744076814655893,
      "loss": 4.9051,
      "mean_token_accuracy": 0.19188318848609925,
      "num_tokens": 63756047.0,
      "step": 29110
    },
    {
      "entropy": 5.640645503997803,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041741333905481,
      "loss": 5.0525,
      "mean_token_accuracy": 0.18327870666980745,
      "num_tokens": 63768006.0,
      "step": 29115
    },
    {
      "entropy": 5.610749530792236,
      "epoch": 2.890035728463676,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.0106,
      "mean_token_accuracy": 0.17859515845775603,
      "num_tokens": 63778765.0,
      "step": 29120
    },
    {
      "entropy": 5.6283860206604,
      "epoch": 2.8905319571258437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.0138,
      "mean_token_accuracy": 0.1907537192106247,
      "num_tokens": 63791305.0,
      "step": 29125
    },
    {
      "entropy": 5.625473546981811,
      "epoch": 2.891028185788011,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004173310305930209,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18990800380706788,
      "num_tokens": 63803378.0,
      "step": 29130
    },
    {
      "entropy": 5.597894811630249,
      "epoch": 2.8915244144501786,
      "grad_norm": 1.0,
      "learning_rate": 0.00041730358737917823,
      "loss": 4.9838,
      "mean_token_accuracy": 0.18871217519044875,
      "num_tokens": 63815621.0,
      "step": 29135
    },
    {
      "entropy": 5.618187522888183,
      "epoch": 2.892020643112346,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041727614063651313,
      "loss": 5.0477,
      "mean_token_accuracy": 0.18123273849487304,
      "num_tokens": 63828009.0,
      "step": 29140
    },
    {
      "entropy": 5.604408311843872,
      "epoch": 2.8925168717745136,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.0037,
      "mean_token_accuracy": 0.18375225514173507,
      "num_tokens": 63840506.0,
      "step": 29145
    },
    {
      "entropy": 5.638003206253051,
      "epoch": 2.893013100436681,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.0403,
      "mean_token_accuracy": 0.18737085461616515,
      "num_tokens": 63851894.0,
      "step": 29150
    },
    {
      "entropy": 5.688841962814331,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17477547973394394,
      "num_tokens": 63863815.0,
      "step": 29155
    },
    {
      "entropy": 5.673680019378662,
      "epoch": 2.8940055577610164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18246382027864455,
      "num_tokens": 63875115.0,
      "step": 29160
    },
    {
      "entropy": 5.6721510887146,
      "epoch": 2.894501786423184,
      "grad_norm": 1.21875,
      "learning_rate": 0.000417138854014678,
      "loss": 5.0449,
      "mean_token_accuracy": 0.18355215191841126,
      "num_tokens": 63885652.0,
      "step": 29165
    },
    {
      "entropy": 5.602971982955933,
      "epoch": 2.8949980150853514,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004171113861133751,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1828499898314476,
      "num_tokens": 63896861.0,
      "step": 29170
    },
    {
      "entropy": 5.605241966247559,
      "epoch": 2.895494243747519,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17967898398637772,
      "num_tokens": 63908736.0,
      "step": 29175
    },
    {
      "entropy": 5.600818252563476,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18447552472352982,
      "num_tokens": 63920841.0,
      "step": 29180
    },
    {
      "entropy": 5.658558082580567,
      "epoch": 2.896486701071854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004170289612678545,
      "loss": 5.0344,
      "mean_token_accuracy": 0.17961991280317308,
      "num_tokens": 63931421.0,
      "step": 29185
    },
    {
      "entropy": 5.666605043411255,
      "epoch": 2.8969829297340213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18217270076274872,
      "num_tokens": 63941548.0,
      "step": 29190
    },
    {
      "entropy": 5.657286596298218,
      "epoch": 2.897479158396189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.0516,
      "mean_token_accuracy": 0.18012888729572296,
      "num_tokens": 63953492.0,
      "step": 29195
    },
    {
      "entropy": 5.590347480773926,
      "epoch": 2.8979753870583567,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041694650472422467,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18304546177387238,
      "num_tokens": 63965601.0,
      "step": 29200
    },
    {
      "entropy": 5.661989355087281,
      "epoch": 2.898471615720524,
      "grad_norm": 1.125,
      "learning_rate": 0.00041691901216883763,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18492138087749482,
      "num_tokens": 63975857.0,
      "step": 29205
    },
    {
      "entropy": 5.674267959594727,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000416891516094165,
      "loss": 5.0897,
      "mean_token_accuracy": 0.18045836985111235,
      "num_tokens": 63987504.0,
      "step": 29210
    },
    {
      "entropy": 5.594697427749634,
      "epoch": 2.899464073044859,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004168640165008884,
      "loss": 4.9752,
      "mean_token_accuracy": 0.19127483516931534,
      "num_tokens": 63998362.0,
      "step": 29215
    },
    {
      "entropy": 5.6241774559021,
      "epoch": 2.8999603017070266,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041683651338968994,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18619669526815413,
      "num_tokens": 64008523.0,
      "step": 29220
    },
    {
      "entropy": 5.571866130828857,
      "epoch": 2.900456530369194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041680900676125147,
      "loss": 4.9814,
      "mean_token_accuracy": 0.18539904057979584,
      "num_tokens": 64019310.0,
      "step": 29225
    },
    {
      "entropy": 5.732583284378052,
      "epoch": 2.9009527590313615,
      "grad_norm": 1.125,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.1032,
      "mean_token_accuracy": 0.18077268451452255,
      "num_tokens": 64030625.0,
      "step": 29230
    },
    {
      "entropy": 5.54557204246521,
      "epoch": 2.901448987693529,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.9064,
      "mean_token_accuracy": 0.1982469767332077,
      "num_tokens": 64041947.0,
      "step": 29235
    },
    {
      "entropy": 5.5619142055511475,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041672646577931835,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18424175530672074,
      "num_tokens": 64052709.0,
      "step": 29240
    },
    {
      "entropy": 5.564585494995117,
      "epoch": 2.9024414450178644,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041669894508874225,
      "loss": 4.9552,
      "mean_token_accuracy": 0.19219148755073548,
      "num_tokens": 64064284.0,
      "step": 29245
    },
    {
      "entropy": 5.623712062835693,
      "epoch": 2.902937673680032,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041667142088433766,
      "loss": 5.0197,
      "mean_token_accuracy": 0.19259105920791625,
      "num_tokens": 64074086.0,
      "step": 29250
    },
    {
      "entropy": 5.5658097743988035,
      "epoch": 2.9034339023421993,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18521055430173874,
      "num_tokens": 64085958.0,
      "step": 29255
    },
    {
      "entropy": 5.578352546691894,
      "epoch": 2.903930131004367,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004166163619367733,
      "loss": 5.0092,
      "mean_token_accuracy": 0.1824960619211197,
      "num_tokens": 64098027.0,
      "step": 29260
    },
    {
      "entropy": 5.7089855670928955,
      "epoch": 2.9044263596665343,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00041658882719497876,
      "loss": 5.1133,
      "mean_token_accuracy": 0.18311674594879152,
      "num_tokens": 64109452.0,
      "step": 29265
    },
    {
      "entropy": 5.701556158065796,
      "epoch": 2.9049225883287018,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18517804890871048,
      "num_tokens": 64120803.0,
      "step": 29270
    },
    {
      "entropy": 5.6579454898834225,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004165337471787792,
      "loss": 5.0428,
      "mean_token_accuracy": 0.18345514982938765,
      "num_tokens": 64132594.0,
      "step": 29275
    },
    {
      "entropy": 5.60672435760498,
      "epoch": 2.905915045653037,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041650620190574004,
      "loss": 5.0221,
      "mean_token_accuracy": 0.1799276739358902,
      "num_tokens": 64143039.0,
      "step": 29280
    },
    {
      "entropy": 5.6222193241119385,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.0287,
      "mean_token_accuracy": 0.17873555570840835,
      "num_tokens": 64153171.0,
      "step": 29285
    },
    {
      "entropy": 5.453498697280883,
      "epoch": 2.906907502977372,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041645110083319805,
      "loss": 4.8225,
      "mean_token_accuracy": 0.20693499147891997,
      "num_tokens": 64163456.0,
      "step": 29290
    },
    {
      "entropy": 5.63717303276062,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041642354503506175,
      "loss": 5.0823,
      "mean_token_accuracy": 0.17891049683094024,
      "num_tokens": 64175728.0,
      "step": 29295
    },
    {
      "entropy": 5.71824369430542,
      "epoch": 2.907899960301707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041639598572992625,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1810062512755394,
      "num_tokens": 64186600.0,
      "step": 29300
    },
    {
      "entropy": 5.605459594726563,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.0318,
      "mean_token_accuracy": 0.1843825489282608,
      "num_tokens": 64198112.0,
      "step": 29305
    },
    {
      "entropy": 5.624286556243897,
      "epoch": 2.908892417626042,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041634085660139145,
      "loss": 4.9878,
      "mean_token_accuracy": 0.1896026462316513,
      "num_tokens": 64209561.0,
      "step": 29310
    },
    {
      "entropy": 5.633346939086914,
      "epoch": 2.9093886462882095,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004163132867793592,
      "loss": 5.0383,
      "mean_token_accuracy": 0.18183338344097139,
      "num_tokens": 64221100.0,
      "step": 29315
    },
    {
      "entropy": 5.5901648044586185,
      "epoch": 2.909884874950377,
      "grad_norm": 1.09375,
      "learning_rate": 0.000416285713453062,
      "loss": 4.9928,
      "mean_token_accuracy": 0.18911289423704147,
      "num_tokens": 64232262.0,
      "step": 29320
    },
    {
      "entropy": 5.627358055114746,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004162581366231835,
      "loss": 5.0021,
      "mean_token_accuracy": 0.18106300681829451,
      "num_tokens": 64242722.0,
      "step": 29325
    },
    {
      "entropy": 5.670232248306275,
      "epoch": 2.9108773322747123,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041623055629040764,
      "loss": 5.073,
      "mean_token_accuracy": 0.17760708034038544,
      "num_tokens": 64253286.0,
      "step": 29330
    },
    {
      "entropy": 5.682677125930786,
      "epoch": 2.91137356093688,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004162029724554183,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17873548120260238,
      "num_tokens": 64264944.0,
      "step": 29335
    },
    {
      "entropy": 5.586065196990967,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041617538511889946,
      "loss": 5.0379,
      "mean_token_accuracy": 0.1855611249804497,
      "num_tokens": 64276681.0,
      "step": 29340
    },
    {
      "entropy": 5.625227594375611,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004161477942815353,
      "loss": 4.9624,
      "mean_token_accuracy": 0.18881118148565293,
      "num_tokens": 64286586.0,
      "step": 29345
    },
    {
      "entropy": 5.66846718788147,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041612019994401,
      "loss": 5.0352,
      "mean_token_accuracy": 0.1799033373594284,
      "num_tokens": 64297231.0,
      "step": 29350
    },
    {
      "entropy": 5.619969034194947,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004160926021070078,
      "loss": 5.042,
      "mean_token_accuracy": 0.18226848542690277,
      "num_tokens": 64308405.0,
      "step": 29355
    },
    {
      "entropy": 5.67510232925415,
      "epoch": 2.913854704247717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.0121,
      "mean_token_accuracy": 0.1824918121099472,
      "num_tokens": 64318287.0,
      "step": 29360
    },
    {
      "entropy": 5.6454884052276615,
      "epoch": 2.914350932909885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.0578,
      "mean_token_accuracy": 0.18325304687023164,
      "num_tokens": 64330627.0,
      "step": 29365
    },
    {
      "entropy": 5.601852130889893,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004160097876059842,
      "loss": 4.9611,
      "mean_token_accuracy": 0.1830483078956604,
      "num_tokens": 64340967.0,
      "step": 29370
    },
    {
      "entropy": 5.626524019241333,
      "epoch": 2.91534339023422,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004159821757779191,
      "loss": 4.9902,
      "mean_token_accuracy": 0.1915719911456108,
      "num_tokens": 64351060.0,
      "step": 29375
    },
    {
      "entropy": 5.640109920501709,
      "epoch": 2.9158396188963875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041595456045379975,
      "loss": 4.939,
      "mean_token_accuracy": 0.19367250204086303,
      "num_tokens": 64362177.0,
      "step": 29380
    },
    {
      "entropy": 5.726779365539551,
      "epoch": 2.916335847558555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004159269416343111,
      "loss": 5.0689,
      "mean_token_accuracy": 0.18087469041347504,
      "num_tokens": 64374188.0,
      "step": 29385
    },
    {
      "entropy": 5.685316705703736,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000415899319320138,
      "loss": 5.0271,
      "mean_token_accuracy": 0.1810691848397255,
      "num_tokens": 64384727.0,
      "step": 29390
    },
    {
      "entropy": 5.657645320892334,
      "epoch": 2.91732830488289,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.0183,
      "mean_token_accuracy": 0.17914471179246902,
      "num_tokens": 64395305.0,
      "step": 29395
    },
    {
      "entropy": 5.622785663604736,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041584406421047825,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1833483099937439,
      "num_tokens": 64406746.0,
      "step": 29400
    },
    {
      "entropy": 5.560365009307861,
      "epoch": 2.918320762207225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041581643141636184,
      "loss": 4.9255,
      "mean_token_accuracy": 0.18711145222187042,
      "num_tokens": 64417044.0,
      "step": 29405
    },
    {
      "entropy": 5.595288944244385,
      "epoch": 2.9188169908693924,
      "grad_norm": 1.125,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.0035,
      "mean_token_accuracy": 0.1789948746562004,
      "num_tokens": 64427875.0,
      "step": 29410
    },
    {
      "entropy": 5.6152369499206545,
      "epoch": 2.91931321953156,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004157611553529822,
      "loss": 5.0077,
      "mean_token_accuracy": 0.19114532619714736,
      "num_tokens": 64440086.0,
      "step": 29415
    },
    {
      "entropy": 5.630182933807373,
      "epoch": 2.919809448193728,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.0247,
      "mean_token_accuracy": 0.18421404510736467,
      "num_tokens": 64450988.0,
      "step": 29420
    },
    {
      "entropy": 5.599168205261231,
      "epoch": 2.9203056768558953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041570586532730926,
      "loss": 5.0148,
      "mean_token_accuracy": 0.1816291406750679,
      "num_tokens": 64461666.0,
      "step": 29425
    },
    {
      "entropy": 5.609896516799926,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.0363,
      "mean_token_accuracy": 0.17943146377801894,
      "num_tokens": 64473111.0,
      "step": 29430
    },
    {
      "entropy": 5.610741662979126,
      "epoch": 2.92129813418023,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.0268,
      "mean_token_accuracy": 0.183017997443676,
      "num_tokens": 64484047.0,
      "step": 29435
    },
    {
      "entropy": 5.66738338470459,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.0714,
      "mean_token_accuracy": 0.18442994356155396,
      "num_tokens": 64496147.0,
      "step": 29440
    },
    {
      "entropy": 5.702509975433349,
      "epoch": 2.922290591504565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004155952434110221,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17642902582883835,
      "num_tokens": 64507255.0,
      "step": 29445
    },
    {
      "entropy": 5.588204431533813,
      "epoch": 2.922786820166733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004155675792140879,
      "loss": 4.9823,
      "mean_token_accuracy": 0.1837150499224663,
      "num_tokens": 64519261.0,
      "step": 29450
    },
    {
      "entropy": 5.5746471881866455,
      "epoch": 2.9232830488289006,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004155399115313806,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18056039661169052,
      "num_tokens": 64530618.0,
      "step": 29455
    },
    {
      "entropy": 5.641719341278076,
      "epoch": 2.923779277491068,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1816745400428772,
      "num_tokens": 64543399.0,
      "step": 29460
    },
    {
      "entropy": 5.597669696807861,
      "epoch": 2.9242755061532355,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041548456571139145,
      "loss": 4.9822,
      "mean_token_accuracy": 0.1865371972322464,
      "num_tokens": 64553673.0,
      "step": 29465
    },
    {
      "entropy": 5.71241569519043,
      "epoch": 2.924771734815403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.1165,
      "mean_token_accuracy": 0.18219785243272782,
      "num_tokens": 64563387.0,
      "step": 29470
    },
    {
      "entropy": 5.66056981086731,
      "epoch": 2.9252679634775705,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041542920595654423,
      "loss": 4.967,
      "mean_token_accuracy": 0.19155568331480027,
      "num_tokens": 64573877.0,
      "step": 29475
    },
    {
      "entropy": 5.712409019470215,
      "epoch": 2.925764192139738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041540152085526476,
      "loss": 5.0531,
      "mean_token_accuracy": 0.17788354605436324,
      "num_tokens": 64585222.0,
      "step": 29480
    },
    {
      "entropy": 5.6126793384552,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.0,
      "learning_rate": 0.00041537383227233023,
      "loss": 4.9931,
      "mean_token_accuracy": 0.18555389046669007,
      "num_tokens": 64597303.0,
      "step": 29485
    },
    {
      "entropy": 5.599435949325562,
      "epoch": 2.926756649464073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004153461402084271,
      "loss": 4.9663,
      "mean_token_accuracy": 0.1878674656152725,
      "num_tokens": 64608855.0,
      "step": 29490
    },
    {
      "entropy": 5.631173944473266,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004153184446642421,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18025742769241332,
      "num_tokens": 64617708.0,
      "step": 29495
    },
    {
      "entropy": 5.67069263458252,
      "epoch": 2.927749106788408,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041529074564046204,
      "loss": 5.0046,
      "mean_token_accuracy": 0.1843477800488472,
      "num_tokens": 64628494.0,
      "step": 29500
    },
    {
      "entropy": 5.665101623535156,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004152630431377738,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18277495950460435,
      "num_tokens": 64639840.0,
      "step": 29505
    },
    {
      "entropy": 5.664980506896972,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004152353371568643,
      "loss": 5.0169,
      "mean_token_accuracy": 0.18971168994903564,
      "num_tokens": 64650842.0,
      "step": 29510
    },
    {
      "entropy": 5.629262542724609,
      "epoch": 2.9292377927749107,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041520762769842065,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18437568843364716,
      "num_tokens": 64662383.0,
      "step": 29515
    },
    {
      "entropy": 5.617868566513062,
      "epoch": 2.929734021437078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041517991476312994,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18237659931182862,
      "num_tokens": 64672840.0,
      "step": 29520
    },
    {
      "entropy": 5.633835792541504,
      "epoch": 2.9302302500992456,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004151521983516794,
      "loss": 4.9957,
      "mean_token_accuracy": 0.18568356931209565,
      "num_tokens": 64684001.0,
      "step": 29525
    },
    {
      "entropy": 5.630213642120362,
      "epoch": 2.930726478761413,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.0885,
      "mean_token_accuracy": 0.18228597640991212,
      "num_tokens": 64695967.0,
      "step": 29530
    },
    {
      "entropy": 5.617193937301636,
      "epoch": 2.931222707423581,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004150967551030482,
      "loss": 4.9462,
      "mean_token_accuracy": 0.18213803917169571,
      "num_tokens": 64706335.0,
      "step": 29535
    },
    {
      "entropy": 5.594126129150391,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18093772530555724,
      "num_tokens": 64717833.0,
      "step": 29540
    },
    {
      "entropy": 5.55467004776001,
      "epoch": 2.932215164747916,
      "grad_norm": 1.125,
      "learning_rate": 0.0004150412979580264,
      "loss": 4.9512,
      "mean_token_accuracy": 0.1885417014360428,
      "num_tokens": 64727378.0,
      "step": 29545
    },
    {
      "entropy": 5.574867153167725,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.125,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.0249,
      "mean_token_accuracy": 0.18061089217662812,
      "num_tokens": 64738340.0,
      "step": 29550
    },
    {
      "entropy": 5.67502613067627,
      "epoch": 2.933207622072251,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041498582692211513,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18926794528961183,
      "num_tokens": 64748137.0,
      "step": 29555
    },
    {
      "entropy": 5.611539554595947,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041495808619679525,
      "loss": 4.9543,
      "mean_token_accuracy": 0.1869277685880661,
      "num_tokens": 64758829.0,
      "step": 29560
    },
    {
      "entropy": 5.62513952255249,
      "epoch": 2.934200079396586,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.0258,
      "mean_token_accuracy": 0.17968139201402664,
      "num_tokens": 64770981.0,
      "step": 29565
    },
    {
      "entropy": 5.601516962051392,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004149025943348666,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18441947549581528,
      "num_tokens": 64781475.0,
      "step": 29570
    },
    {
      "entropy": 5.685845136642456,
      "epoch": 2.935192536720921,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041487484319963395,
      "loss": 5.1324,
      "mean_token_accuracy": 0.18173361271619798,
      "num_tokens": 64793478.0,
      "step": 29575
    },
    {
      "entropy": 5.628015995025635,
      "epoch": 2.9356887653830883,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004148470885958066,
      "loss": 4.9635,
      "mean_token_accuracy": 0.1882938638329506,
      "num_tokens": 64803959.0,
      "step": 29580
    },
    {
      "entropy": 5.640724039077758,
      "epoch": 2.936184994045256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004148193305240727,
      "loss": 5.0428,
      "mean_token_accuracy": 0.17817279696464539,
      "num_tokens": 64814426.0,
      "step": 29585
    },
    {
      "entropy": 5.5592866897583,
      "epoch": 2.9366812227074237,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041479156898512076,
      "loss": 4.9379,
      "mean_token_accuracy": 0.19113812744617462,
      "num_tokens": 64826128.0,
      "step": 29590
    },
    {
      "entropy": 5.615678024291992,
      "epoch": 2.937177451369591,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004147638039796391,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18982513397932052,
      "num_tokens": 64836285.0,
      "step": 29595
    },
    {
      "entropy": 5.638897800445557,
      "epoch": 2.9376736800317587,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041473603550831636,
      "loss": 5.0448,
      "mean_token_accuracy": 0.180505234003067,
      "num_tokens": 64848342.0,
      "step": 29600
    },
    {
      "entropy": 5.642990684509277,
      "epoch": 2.938169908693926,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041470826357184096,
      "loss": 5.0418,
      "mean_token_accuracy": 0.18202642500400543,
      "num_tokens": 64859541.0,
      "step": 29605
    },
    {
      "entropy": 5.617239475250244,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004146804881709017,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18257570564746856,
      "num_tokens": 64870816.0,
      "step": 29610
    },
    {
      "entropy": 5.654520893096924,
      "epoch": 2.939162366018261,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1828557714819908,
      "num_tokens": 64882604.0,
      "step": 29615
    },
    {
      "entropy": 5.650551128387451,
      "epoch": 2.939658594680429,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004146249269783867,
      "loss": 4.9901,
      "mean_token_accuracy": 0.19261703938245772,
      "num_tokens": 64893861.0,
      "step": 29620
    },
    {
      "entropy": 5.700672149658203,
      "epoch": 2.9401548233425965,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.0495,
      "mean_token_accuracy": 0.17668808549642562,
      "num_tokens": 64903863.0,
      "step": 29625
    },
    {
      "entropy": 5.678145980834961,
      "epoch": 2.940651052004764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17720865607261657,
      "num_tokens": 64915754.0,
      "step": 29630
    },
    {
      "entropy": 5.569218254089355,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004145415592233569,
      "loss": 4.9362,
      "mean_token_accuracy": 0.18710129708051682,
      "num_tokens": 64926217.0,
      "step": 29635
    },
    {
      "entropy": 5.608903312683106,
      "epoch": 2.941643509329099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041451376305010144,
      "loss": 4.9745,
      "mean_token_accuracy": 0.19088849574327468,
      "num_tokens": 64936478.0,
      "step": 29640
    },
    {
      "entropy": 5.629286193847657,
      "epoch": 2.9421397379912664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004144859634172053,
      "loss": 5.0677,
      "mean_token_accuracy": 0.1784866601228714,
      "num_tokens": 64948195.0,
      "step": 29645
    },
    {
      "entropy": 5.669453573226929,
      "epoch": 2.942635966653434,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041445816032535784,
      "loss": 5.1402,
      "mean_token_accuracy": 0.1752621963620186,
      "num_tokens": 64959019.0,
      "step": 29650
    },
    {
      "entropy": 5.721898174285888,
      "epoch": 2.9431321953156013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.0603,
      "mean_token_accuracy": 0.1786213219165802,
      "num_tokens": 64970147.0,
      "step": 29655
    },
    {
      "entropy": 5.6643171310424805,
      "epoch": 2.943628423977769,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004144025437675668,
      "loss": 4.9546,
      "mean_token_accuracy": 0.1880194440484047,
      "num_tokens": 64980664.0,
      "step": 29660
    },
    {
      "entropy": 5.644007873535156,
      "epoch": 2.9441246526399363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.0916,
      "mean_token_accuracy": 0.18646325021982194,
      "num_tokens": 64992489.0,
      "step": 29665
    },
    {
      "entropy": 5.601555919647216,
      "epoch": 2.9446208813021038,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041434691338224506,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18183703422546388,
      "num_tokens": 65002861.0,
      "step": 29670
    },
    {
      "entropy": 5.638562107086182,
      "epoch": 2.9451171099642717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004143190930059846,
      "loss": 5.051,
      "mean_token_accuracy": 0.1793093889951706,
      "num_tokens": 65013865.0,
      "step": 29675
    },
    {
      "entropy": 5.644855928421021,
      "epoch": 2.945613338626439,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.0592,
      "mean_token_accuracy": 0.1857016295194626,
      "num_tokens": 65026155.0,
      "step": 29680
    },
    {
      "entropy": 5.652260684967041,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17651654928922653,
      "num_tokens": 65037306.0,
      "step": 29685
    },
    {
      "entropy": 5.636915922164917,
      "epoch": 2.946605795950774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142356111510831,
      "loss": 5.0467,
      "mean_token_accuracy": 0.18453503400087357,
      "num_tokens": 65047715.0,
      "step": 29690
    },
    {
      "entropy": 5.64082555770874,
      "epoch": 2.9471020246129416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041420777695970945,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19134028106927872,
      "num_tokens": 65057694.0,
      "step": 29695
    },
    {
      "entropy": 5.654380416870117,
      "epoch": 2.947598253275109,
      "grad_norm": 1.171875,
      "learning_rate": 0.000414179939316283,
      "loss": 5.0642,
      "mean_token_accuracy": 0.174689356982708,
      "num_tokens": 65068131.0,
      "step": 29700
    },
    {
      "entropy": 5.697171354293824,
      "epoch": 2.948094481937277,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041415209822149397,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17234092056751252,
      "num_tokens": 65081044.0,
      "step": 29705
    },
    {
      "entropy": 5.672778224945068,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004141242536760327,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1808453992009163,
      "num_tokens": 65092151.0,
      "step": 29710
    },
    {
      "entropy": 5.657071447372436,
      "epoch": 2.949086939261612,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004140964056805897,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18648619651794435,
      "num_tokens": 65103787.0,
      "step": 29715
    },
    {
      "entropy": 5.638384199142456,
      "epoch": 2.9495831679237794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17900338768959045,
      "num_tokens": 65115157.0,
      "step": 29720
    },
    {
      "entropy": 5.662084913253784,
      "epoch": 2.950079396585947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.0703,
      "mean_token_accuracy": 0.18323520570993423,
      "num_tokens": 65127090.0,
      "step": 29725
    },
    {
      "entropy": 5.6642889976501465,
      "epoch": 2.9505756252481143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004140128410012767,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18108914643526078,
      "num_tokens": 65138346.0,
      "step": 29730
    },
    {
      "entropy": 5.700783443450928,
      "epoch": 2.951071853910282,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041398497921281354,
      "loss": 5.1192,
      "mean_token_accuracy": 0.1741524800658226,
      "num_tokens": 65149510.0,
      "step": 29735
    },
    {
      "entropy": 5.639670467376709,
      "epoch": 2.9515680825724493,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.0919,
      "mean_token_accuracy": 0.17667409628629685,
      "num_tokens": 65159828.0,
      "step": 29740
    },
    {
      "entropy": 5.683807325363159,
      "epoch": 2.9520643112346168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.012,
      "mean_token_accuracy": 0.18445794731378556,
      "num_tokens": 65170594.0,
      "step": 29745
    },
    {
      "entropy": 5.656837463378906,
      "epoch": 2.9525605398967842,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041390137317102,
      "loss": 5.039,
      "mean_token_accuracy": 0.1850636899471283,
      "num_tokens": 65180722.0,
      "step": 29750
    },
    {
      "entropy": 5.66458797454834,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004138734976005911,
      "loss": 5.0027,
      "mean_token_accuracy": 0.18759342581033706,
      "num_tokens": 65191390.0,
      "step": 29755
    },
    {
      "entropy": 5.59180793762207,
      "epoch": 2.9535529972211196,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004138456185863986,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18071270883083343,
      "num_tokens": 65202915.0,
      "step": 29760
    },
    {
      "entropy": 5.521899032592773,
      "epoch": 2.954049225883287,
      "grad_norm": 1.125,
      "learning_rate": 0.00041381773612913413,
      "loss": 4.9618,
      "mean_token_accuracy": 0.1888586699962616,
      "num_tokens": 65213098.0,
      "step": 29765
    },
    {
      "entropy": 5.6426389694213865,
      "epoch": 2.9545454545454546,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004137898502294889,
      "loss": 5.0745,
      "mean_token_accuracy": 0.18199582695960997,
      "num_tokens": 65224003.0,
      "step": 29770
    },
    {
      "entropy": 5.682808828353882,
      "epoch": 2.955041683207622,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004137619608881544,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18030325025320054,
      "num_tokens": 65234520.0,
      "step": 29775
    },
    {
      "entropy": 5.623538064956665,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004137340681058223,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18244570046663283,
      "num_tokens": 65245117.0,
      "step": 29780
    },
    {
      "entropy": 5.523581266403198,
      "epoch": 2.956034140531957,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041370617188318424,
      "loss": 4.9059,
      "mean_token_accuracy": 0.19290920197963715,
      "num_tokens": 65254933.0,
      "step": 29785
    },
    {
      "entropy": 5.638413524627685,
      "epoch": 2.956530369194125,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041367827222093195,
      "loss": 5.0909,
      "mean_token_accuracy": 0.1807855635881424,
      "num_tokens": 65266253.0,
      "step": 29790
    },
    {
      "entropy": 5.606453371047974,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004136503691197574,
      "loss": 4.9838,
      "mean_token_accuracy": 0.18394114524126054,
      "num_tokens": 65276595.0,
      "step": 29795
    },
    {
      "entropy": 5.649300193786621,
      "epoch": 2.95752282651846,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.1185,
      "mean_token_accuracy": 0.18082195073366164,
      "num_tokens": 65286915.0,
      "step": 29800
    },
    {
      "entropy": 5.624021768569946,
      "epoch": 2.9580190551806274,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.981,
      "mean_token_accuracy": 0.18928828090429306,
      "num_tokens": 65297613.0,
      "step": 29805
    },
    {
      "entropy": 5.649372863769531,
      "epoch": 2.958515283842795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18142866343259811,
      "num_tokens": 65309273.0,
      "step": 29810
    },
    {
      "entropy": 5.695127868652344,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.0595,
      "mean_token_accuracy": 0.17774974405765534,
      "num_tokens": 65321241.0,
      "step": 29815
    },
    {
      "entropy": 5.592904710769654,
      "epoch": 2.95950774116713,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004135108020542705,
      "loss": 4.94,
      "mean_token_accuracy": 0.1902812674641609,
      "num_tokens": 65331520.0,
      "step": 29820
    },
    {
      "entropy": 5.584838056564331,
      "epoch": 2.9600039698292973,
      "grad_norm": 1.125,
      "learning_rate": 0.00041348287833409557,
      "loss": 4.9866,
      "mean_token_accuracy": 0.1853605955839157,
      "num_tokens": 65341551.0,
      "step": 29825
    },
    {
      "entropy": 5.593941259384155,
      "epoch": 2.9605001984914647,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004134549511798438,
      "loss": 5.0398,
      "mean_token_accuracy": 0.18158978670835496,
      "num_tokens": 65352489.0,
      "step": 29830
    },
    {
      "entropy": 5.596078824996948,
      "epoch": 2.960996427153632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004134270205922075,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1914202943444252,
      "num_tokens": 65363298.0,
      "step": 29835
    },
    {
      "entropy": 5.72927942276001,
      "epoch": 2.9614926558157997,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041339908657187933,
      "loss": 5.102,
      "mean_token_accuracy": 0.18113802522420883,
      "num_tokens": 65374426.0,
      "step": 29840
    },
    {
      "entropy": 5.665491390228271,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.0312,
      "mean_token_accuracy": 0.19066964834928513,
      "num_tokens": 65386007.0,
      "step": 29845
    },
    {
      "entropy": 5.6309620380401615,
      "epoch": 2.962485113140135,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041334320823591857,
      "loss": 5.0375,
      "mean_token_accuracy": 0.18641480505466462,
      "num_tokens": 65397457.0,
      "step": 29850
    },
    {
      "entropy": 5.603397035598755,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041331526392167155,
      "loss": 4.9596,
      "mean_token_accuracy": 0.1906726211309433,
      "num_tokens": 65407905.0,
      "step": 29855
    },
    {
      "entropy": 5.641343212127685,
      "epoch": 2.96347757046447,
      "grad_norm": 1.125,
      "learning_rate": 0.000413287316177504,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18419494181871415,
      "num_tokens": 65418946.0,
      "step": 29860
    },
    {
      "entropy": 5.606385326385498,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.125,
      "learning_rate": 0.00041325936500410905,
      "loss": 4.9471,
      "mean_token_accuracy": 0.194561967253685,
      "num_tokens": 65428700.0,
      "step": 29865
    },
    {
      "entropy": 5.609182262420655,
      "epoch": 2.964470027788805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041323141040217976,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1866613194346428,
      "num_tokens": 65439473.0,
      "step": 29870
    },
    {
      "entropy": 5.634658861160278,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00041320345237240923,
      "loss": 5.036,
      "mean_token_accuracy": 0.18747060298919677,
      "num_tokens": 65451799.0,
      "step": 29875
    },
    {
      "entropy": 5.693775081634522,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000413175490915491,
      "loss": 5.0309,
      "mean_token_accuracy": 0.17899725288152696,
      "num_tokens": 65462890.0,
      "step": 29880
    },
    {
      "entropy": 5.601750612258911,
      "epoch": 2.965958713775308,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004131475260321183,
      "loss": 4.9906,
      "mean_token_accuracy": 0.1841592788696289,
      "num_tokens": 65473503.0,
      "step": 29885
    },
    {
      "entropy": 5.636192703247071,
      "epoch": 2.9664549424374753,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041311955772298464,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17929145246744155,
      "num_tokens": 65485383.0,
      "step": 29890
    },
    {
      "entropy": 5.548891592025757,
      "epoch": 2.966951171099643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004130915859887837,
      "loss": 4.8696,
      "mean_token_accuracy": 0.19181720912456512,
      "num_tokens": 65496044.0,
      "step": 29895
    },
    {
      "entropy": 5.645783567428589,
      "epoch": 2.9674473997618103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18482993245124818,
      "num_tokens": 65506673.0,
      "step": 29900
    },
    {
      "entropy": 5.589011001586914,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041303563224795393,
      "loss": 4.9896,
      "mean_token_accuracy": 0.17989249974489213,
      "num_tokens": 65516492.0,
      "step": 29905
    },
    {
      "entropy": 5.590347146987915,
      "epoch": 2.968439857086145,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.9662,
      "mean_token_accuracy": 0.18518815636634828,
      "num_tokens": 65528240.0,
      "step": 29910
    },
    {
      "entropy": 5.7083488464355465,
      "epoch": 2.9689360857483127,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.0517,
      "mean_token_accuracy": 0.1810692146420479,
      "num_tokens": 65537861.0,
      "step": 29915
    },
    {
      "entropy": 5.622909688949585,
      "epoch": 2.96943231441048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.031,
      "mean_token_accuracy": 0.17987693399190902,
      "num_tokens": 65549326.0,
      "step": 29920
    },
    {
      "entropy": 5.573264694213867,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041292368369601136,
      "loss": 4.9782,
      "mean_token_accuracy": 0.18260956853628157,
      "num_tokens": 65560962.0,
      "step": 29925
    },
    {
      "entropy": 5.672169637680054,
      "epoch": 2.9704247717348156,
      "grad_norm": 1.078125,
      "learning_rate": 0.000412895688005765,
      "loss": 5.0589,
      "mean_token_accuracy": 0.18619115203619002,
      "num_tokens": 65572411.0,
      "step": 29930
    },
    {
      "entropy": 5.673497200012207,
      "epoch": 2.970921000396983,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004128676888960028,
      "loss": 5.0673,
      "mean_token_accuracy": 0.18207914978265763,
      "num_tokens": 65582423.0,
      "step": 29935
    },
    {
      "entropy": 5.649214601516723,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041283968636741903,
      "loss": 5.067,
      "mean_token_accuracy": 0.1857440173625946,
      "num_tokens": 65594205.0,
      "step": 29940
    },
    {
      "entropy": 5.576479578018189,
      "epoch": 2.971913457721318,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004128116804207081,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18597959727048874,
      "num_tokens": 65605262.0,
      "step": 29945
    },
    {
      "entropy": 5.621347951889038,
      "epoch": 2.9724096863834855,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004127836710565642,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1822081044316292,
      "num_tokens": 65616175.0,
      "step": 29950
    },
    {
      "entropy": 5.6239067077636715,
      "epoch": 2.972905915045653,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.0468,
      "mean_token_accuracy": 0.17871336191892623,
      "num_tokens": 65628128.0,
      "step": 29955
    },
    {
      "entropy": 5.655569124221802,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.203125,
      "learning_rate": 0.000412727642078757,
      "loss": 5.0201,
      "mean_token_accuracy": 0.1829136461019516,
      "num_tokens": 65639605.0,
      "step": 29960
    },
    {
      "entropy": 5.660594034194946,
      "epoch": 2.9738983723699883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.0695,
      "mean_token_accuracy": 0.18392260372638702,
      "num_tokens": 65652552.0,
      "step": 29965
    },
    {
      "entropy": 5.669814205169677,
      "epoch": 2.974394601032156,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041267159943955483,
      "loss": 4.9996,
      "mean_token_accuracy": 0.18274493366479874,
      "num_tokens": 65662423.0,
      "step": 29970
    },
    {
      "entropy": 5.61334342956543,
      "epoch": 2.9748908296943233,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004126435729986678,
      "loss": 5.0037,
      "mean_token_accuracy": 0.18803874552249908,
      "num_tokens": 65674582.0,
      "step": 29975
    },
    {
      "entropy": 5.587045764923095,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041261554314451674,
      "loss": 5.0056,
      "mean_token_accuracy": 0.18265703916549683,
      "num_tokens": 65685321.0,
      "step": 29980
    },
    {
      "entropy": 5.579566144943238,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004125875098777967,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18788570612668992,
      "num_tokens": 65696000.0,
      "step": 29985
    },
    {
      "entropy": 5.635945749282837,
      "epoch": 2.9763795156808257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041255947319920297,
      "loss": 4.9888,
      "mean_token_accuracy": 0.1875760331749916,
      "num_tokens": 65705930.0,
      "step": 29990
    },
    {
      "entropy": 5.589231252670288,
      "epoch": 2.976875744342993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041253143310943064,
      "loss": 4.9653,
      "mean_token_accuracy": 0.19006443321704863,
      "num_tokens": 65717038.0,
      "step": 29995
    },
    {
      "entropy": 5.65604157447815,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041250338960917515,
      "loss": 5.061,
      "mean_token_accuracy": 0.18489857912063598,
      "num_tokens": 65728915.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.4257256635211375,
      "eval_loss": 5.129428386688232,
      "eval_mean_token_accuracy": 0.1852736881298842,
      "eval_num_tokens": 65728915.0,
      "eval_runtime": 20.5226,
      "eval_samples_per_second": 1640.578,
      "eval_steps_per_second": 205.09,
      "step": 30000
    },
    {
      "entropy": 5.655517959594727,
      "epoch": 2.977868201667328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041247534269913197,
      "loss": 4.9862,
      "mean_token_accuracy": 0.1885976657271385,
      "num_tokens": 65739780.0,
      "step": 30005
    },
    {
      "entropy": 5.686596775054932,
      "epoch": 2.9783644303294956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004124472923799964,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1845828041434288,
      "num_tokens": 65750694.0,
      "step": 30010
    },
    {
      "entropy": 5.640625810623169,
      "epoch": 2.9788606589916635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004124192386524642,
      "loss": 5.027,
      "mean_token_accuracy": 0.18567117899656296,
      "num_tokens": 65761972.0,
      "step": 30015
    },
    {
      "entropy": 5.61476092338562,
      "epoch": 2.979356887653831,
      "grad_norm": 1.0625,
      "learning_rate": 0.000412391181517231,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18361756205558777,
      "num_tokens": 65772936.0,
      "step": 30020
    },
    {
      "entropy": 5.68560757637024,
      "epoch": 2.9798531163159985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041236312097499257,
      "loss": 5.0561,
      "mean_token_accuracy": 0.17919246554374696,
      "num_tokens": 65784876.0,
      "step": 30025
    },
    {
      "entropy": 5.659902191162109,
      "epoch": 2.980349344978166,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004123350570264447,
      "loss": 4.9961,
      "mean_token_accuracy": 0.17778465896844864,
      "num_tokens": 65794943.0,
      "step": 30030
    },
    {
      "entropy": 5.675333786010742,
      "epoch": 2.9808455736403334,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041230698967228334,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17547298222780228,
      "num_tokens": 65807073.0,
      "step": 30035
    },
    {
      "entropy": 5.586102533340454,
      "epoch": 2.981341802302501,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004122789189132045,
      "loss": 4.9547,
      "mean_token_accuracy": 0.19434681683778762,
      "num_tokens": 65819148.0,
      "step": 30040
    },
    {
      "entropy": 5.641989994049072,
      "epoch": 2.9818380309646684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004122508447499043,
      "loss": 4.9889,
      "mean_token_accuracy": 0.1793474644422531,
      "num_tokens": 65831813.0,
      "step": 30045
    },
    {
      "entropy": 5.603312253952026,
      "epoch": 2.9823342596268363,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041222276718307896,
      "loss": 5.1014,
      "mean_token_accuracy": 0.16969270557165145,
      "num_tokens": 65843004.0,
      "step": 30050
    },
    {
      "entropy": 5.675017023086548,
      "epoch": 2.9828304882890038,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004121946862134246,
      "loss": 4.9874,
      "mean_token_accuracy": 0.19138624519109726,
      "num_tokens": 65853144.0,
      "step": 30055
    },
    {
      "entropy": 5.669761228561401,
      "epoch": 2.9833267169511712,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041216660184163767,
      "loss": 5.0106,
      "mean_token_accuracy": 0.17959968149662017,
      "num_tokens": 65864187.0,
      "step": 30060
    },
    {
      "entropy": 5.584762239456177,
      "epoch": 2.9838229456133387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041213851406841456,
      "loss": 4.9233,
      "mean_token_accuracy": 0.19246678352355956,
      "num_tokens": 65876197.0,
      "step": 30065
    },
    {
      "entropy": 5.601703882217407,
      "epoch": 2.984319174275506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004121104228944518,
      "loss": 5.035,
      "mean_token_accuracy": 0.18655723482370376,
      "num_tokens": 65887361.0,
      "step": 30070
    },
    {
      "entropy": 5.707655382156372,
      "epoch": 2.9848154029376737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004120823283204461,
      "loss": 5.0929,
      "mean_token_accuracy": 0.18572610020637512,
      "num_tokens": 65898838.0,
      "step": 30075
    },
    {
      "entropy": 5.5973179817199705,
      "epoch": 2.985311631599841,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412054230347094,
      "loss": 4.9679,
      "mean_token_accuracy": 0.19022254198789595,
      "num_tokens": 65908582.0,
      "step": 30080
    },
    {
      "entropy": 5.571391868591308,
      "epoch": 2.9858078602620086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004120261289750923,
      "loss": 4.9451,
      "mean_token_accuracy": 0.1819291353225708,
      "num_tokens": 65918638.0,
      "step": 30085
    },
    {
      "entropy": 5.6350010395050045,
      "epoch": 2.986304088924176,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041199802420513785,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1852958008646965,
      "num_tokens": 65929735.0,
      "step": 30090
    },
    {
      "entropy": 5.618671607971192,
      "epoch": 2.9868003175863436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004119699160379276,
      "loss": 4.9421,
      "mean_token_accuracy": 0.18828670233488082,
      "num_tokens": 65940063.0,
      "step": 30095
    },
    {
      "entropy": 5.613043785095215,
      "epoch": 2.987296546248511,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041194180447415855,
      "loss": 4.9885,
      "mean_token_accuracy": 0.18641282767057418,
      "num_tokens": 65950584.0,
      "step": 30100
    },
    {
      "entropy": 5.650855636596679,
      "epoch": 2.987792774910679,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004119136895145279,
      "loss": 5.0445,
      "mean_token_accuracy": 0.17904165983200074,
      "num_tokens": 65962531.0,
      "step": 30105
    },
    {
      "entropy": 5.618394136428833,
      "epoch": 2.9882890035728464,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004118855711597327,
      "loss": 5.034,
      "mean_token_accuracy": 0.18312027156352997,
      "num_tokens": 65974043.0,
      "step": 30110
    },
    {
      "entropy": 5.733680868148804,
      "epoch": 2.988785232235014,
      "grad_norm": 1.125,
      "learning_rate": 0.00041185744941047033,
      "loss": 5.1713,
      "mean_token_accuracy": 0.17684618532657623,
      "num_tokens": 65984981.0,
      "step": 30115
    },
    {
      "entropy": 5.648280429840088,
      "epoch": 2.9892814608971814,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004118293242674382,
      "loss": 5.041,
      "mean_token_accuracy": 0.1809054806828499,
      "num_tokens": 65994821.0,
      "step": 30120
    },
    {
      "entropy": 5.630240726470947,
      "epoch": 2.989777689559349,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004118011957313336,
      "loss": 4.9623,
      "mean_token_accuracy": 0.18765304684638978,
      "num_tokens": 66006885.0,
      "step": 30125
    },
    {
      "entropy": 5.6095363140106205,
      "epoch": 2.9902739182215163,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041177306380285405,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18156481981277467,
      "num_tokens": 66018790.0,
      "step": 30130
    },
    {
      "entropy": 5.572743940353393,
      "epoch": 2.9907701468836843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004117449284826974,
      "loss": 4.9129,
      "mean_token_accuracy": 0.1868675395846367,
      "num_tokens": 66028442.0,
      "step": 30135
    },
    {
      "entropy": 5.650924110412598,
      "epoch": 2.9912663755458517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041171678977156104,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18109894245862962,
      "num_tokens": 66037959.0,
      "step": 30140
    },
    {
      "entropy": 5.615357828140259,
      "epoch": 2.991762604208019,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041168864767014295,
      "loss": 4.9807,
      "mean_token_accuracy": 0.18967444598674774,
      "num_tokens": 66048465.0,
      "step": 30145
    },
    {
      "entropy": 5.5847734928131105,
      "epoch": 2.9922588328701867,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000411660502179141,
      "loss": 4.9925,
      "mean_token_accuracy": 0.1792124718427658,
      "num_tokens": 66058996.0,
      "step": 30150
    },
    {
      "entropy": 5.650338554382325,
      "epoch": 2.992755061532354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004116323532992529,
      "loss": 5.0363,
      "mean_token_accuracy": 0.1837320640683174,
      "num_tokens": 66070293.0,
      "step": 30155
    },
    {
      "entropy": 5.602009057998657,
      "epoch": 2.9932512901945216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041160420103117704,
      "loss": 4.8989,
      "mean_token_accuracy": 0.19600794315338135,
      "num_tokens": 66080507.0,
      "step": 30160
    },
    {
      "entropy": 5.621094417572022,
      "epoch": 2.993747518856689,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041157604537561123,
      "loss": 5.0765,
      "mean_token_accuracy": 0.18390306383371352,
      "num_tokens": 66090574.0,
      "step": 30165
    },
    {
      "entropy": 5.677659130096435,
      "epoch": 2.9942437475188566,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041154788633325387,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18065542876720428,
      "num_tokens": 66101773.0,
      "step": 30170
    },
    {
      "entropy": 5.6798341274261475,
      "epoch": 2.994739976181024,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004115197239048031,
      "loss": 5.0641,
      "mean_token_accuracy": 0.18194830417633057,
      "num_tokens": 66113805.0,
      "step": 30175
    },
    {
      "entropy": 5.645811223983765,
      "epoch": 2.9952362048431915,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041149155809095734,
      "loss": 5.002,
      "mean_token_accuracy": 0.1791323736310005,
      "num_tokens": 66125397.0,
      "step": 30180
    },
    {
      "entropy": 5.668725681304932,
      "epoch": 2.995732433505359,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041146338889241506,
      "loss": 5.1159,
      "mean_token_accuracy": 0.18074213713407516,
      "num_tokens": 66135393.0,
      "step": 30185
    },
    {
      "entropy": 5.6435866355896,
      "epoch": 2.996228662167527,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041143521630987483,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18870480507612228,
      "num_tokens": 66145572.0,
      "step": 30190
    },
    {
      "entropy": 5.7105920791625975,
      "epoch": 2.9967248908296944,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00041140704034403527,
      "loss": 4.9981,
      "mean_token_accuracy": 0.1884501174092293,
      "num_tokens": 66156062.0,
      "step": 30195
    },
    {
      "entropy": 5.624505186080933,
      "epoch": 2.997221119491862,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000411378860995595,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18262304216623307,
      "num_tokens": 66166322.0,
      "step": 30200
    },
    {
      "entropy": 5.535659456253052,
      "epoch": 2.9977173481540293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004113506782652528,
      "loss": 4.9647,
      "mean_token_accuracy": 0.18651018142700196,
      "num_tokens": 66178110.0,
      "step": 30205
    },
    {
      "entropy": 5.652169990539551,
      "epoch": 2.998213576816197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004113224921537076,
      "loss": 4.9844,
      "mean_token_accuracy": 0.18881457597017287,
      "num_tokens": 66187975.0,
      "step": 30210
    },
    {
      "entropy": 5.608609771728515,
      "epoch": 2.9987098054783643,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041129430266165836,
      "loss": 4.9518,
      "mean_token_accuracy": 0.19382120817899703,
      "num_tokens": 66198149.0,
      "step": 30215
    },
    {
      "entropy": 5.647004222869873,
      "epoch": 2.999206034140532,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004112661097898041,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18068738430738449,
      "num_tokens": 66211187.0,
      "step": 30220
    },
    {
      "entropy": 5.70217547416687,
      "epoch": 2.9997022628026997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041123791353884397,
      "loss": 5.0915,
      "mean_token_accuracy": 0.17839718759059905,
      "num_tokens": 66221659.0,
      "step": 30225
    },
    {
      "entropy": 5.646273422241211,
      "epoch": 3.000198491464867,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004112097139094771,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17963675409555435,
      "num_tokens": 66231997.0,
      "step": 30230
    },
    {
      "entropy": 5.627167797088623,
      "epoch": 3.0006947201270346,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041118151090240284,
      "loss": 4.9352,
      "mean_token_accuracy": 0.1823921397328377,
      "num_tokens": 66242216.0,
      "step": 30235
    },
    {
      "entropy": 5.645984888076782,
      "epoch": 3.001190948789202,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041115330451832065,
      "loss": 5.0492,
      "mean_token_accuracy": 0.18422577530145645,
      "num_tokens": 66255143.0,
      "step": 30240
    },
    {
      "entropy": 5.705660438537597,
      "epoch": 3.0016871774513696,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041112509475792983,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18225701302289962,
      "num_tokens": 66265732.0,
      "step": 30245
    },
    {
      "entropy": 5.692768001556397,
      "epoch": 3.002183406113537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041109688162193,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18495579063892365,
      "num_tokens": 66276781.0,
      "step": 30250
    },
    {
      "entropy": 5.59572205543518,
      "epoch": 3.0026796347757045,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004110686651110207,
      "loss": 4.9378,
      "mean_token_accuracy": 0.1897829920053482,
      "num_tokens": 66287600.0,
      "step": 30255
    },
    {
      "entropy": 5.62666597366333,
      "epoch": 3.003175863437872,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004110404452259017,
      "loss": 4.9829,
      "mean_token_accuracy": 0.18574561029672623,
      "num_tokens": 66299360.0,
      "step": 30260
    },
    {
      "entropy": 5.6855274677276615,
      "epoch": 3.0036720921000395,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004110122219672729,
      "loss": 4.9672,
      "mean_token_accuracy": 0.1960098475217819,
      "num_tokens": 66309268.0,
      "step": 30265
    },
    {
      "entropy": 5.637028503417969,
      "epoch": 3.0041683207622074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000410983995335834,
      "loss": 4.9805,
      "mean_token_accuracy": 0.1830380827188492,
      "num_tokens": 66320367.0,
      "step": 30270
    },
    {
      "entropy": 5.6447395324707035,
      "epoch": 3.004664549424375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000410955765332285,
      "loss": 4.9347,
      "mean_token_accuracy": 0.18921254128217696,
      "num_tokens": 66331478.0,
      "step": 30275
    },
    {
      "entropy": 5.592505931854248,
      "epoch": 3.0051607780865424,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041092753195732607,
      "loss": 4.9162,
      "mean_token_accuracy": 0.18985694348812104,
      "num_tokens": 66342110.0,
      "step": 30280
    },
    {
      "entropy": 5.605686902999878,
      "epoch": 3.00565700674871,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041089929521165725,
      "loss": 5.0365,
      "mean_token_accuracy": 0.177264104783535,
      "num_tokens": 66353189.0,
      "step": 30285
    },
    {
      "entropy": 5.625279760360717,
      "epoch": 3.0061532354108773,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004108710550959787,
      "loss": 4.9685,
      "mean_token_accuracy": 0.18598517775535583,
      "num_tokens": 66363681.0,
      "step": 30290
    },
    {
      "entropy": 5.666809940338135,
      "epoch": 3.006649464073045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041084281161099073,
      "loss": 4.9546,
      "mean_token_accuracy": 0.18897387534379959,
      "num_tokens": 66373200.0,
      "step": 30295
    },
    {
      "entropy": 5.647259473800659,
      "epoch": 3.0071456927352123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004108145647573938,
      "loss": 4.9429,
      "mean_token_accuracy": 0.18803007751703263,
      "num_tokens": 66382717.0,
      "step": 30300
    },
    {
      "entropy": 5.606934976577759,
      "epoch": 3.0076419213973797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041078631453588834,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18914911299943923,
      "num_tokens": 66394471.0,
      "step": 30305
    },
    {
      "entropy": 5.640807199478149,
      "epoch": 3.0081381500595477,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004107580609471748,
      "loss": 4.991,
      "mean_token_accuracy": 0.1862222045660019,
      "num_tokens": 66405739.0,
      "step": 30310
    },
    {
      "entropy": 5.62283148765564,
      "epoch": 3.008634378721715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041072980399195394,
      "loss": 5.0204,
      "mean_token_accuracy": 0.18131803274154662,
      "num_tokens": 66416087.0,
      "step": 30315
    },
    {
      "entropy": 5.7019084930419925,
      "epoch": 3.0091306073838826,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004107015436709264,
      "loss": 5.0726,
      "mean_token_accuracy": 0.18146310299634932,
      "num_tokens": 66425741.0,
      "step": 30320
    },
    {
      "entropy": 5.7050008296966555,
      "epoch": 3.00962683604605,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041067327998479296,
      "loss": 5.0057,
      "mean_token_accuracy": 0.188351009786129,
      "num_tokens": 66437221.0,
      "step": 30325
    },
    {
      "entropy": 5.636007308959961,
      "epoch": 3.0101230647082176,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041064501293425453,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18937138020992278,
      "num_tokens": 66447303.0,
      "step": 30330
    },
    {
      "entropy": 5.655188703536988,
      "epoch": 3.010619293370385,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004106167425200121,
      "loss": 4.9235,
      "mean_token_accuracy": 0.19245097488164903,
      "num_tokens": 66458160.0,
      "step": 30335
    },
    {
      "entropy": 5.628158378601074,
      "epoch": 3.0111155220325525,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004105884687427667,
      "loss": 4.9671,
      "mean_token_accuracy": 0.18588169068098068,
      "num_tokens": 66467989.0,
      "step": 30340
    },
    {
      "entropy": 5.606671762466431,
      "epoch": 3.01161175069472,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004105601916032194,
      "loss": 4.9319,
      "mean_token_accuracy": 0.18799275159835815,
      "num_tokens": 66479210.0,
      "step": 30345
    },
    {
      "entropy": 5.627881240844727,
      "epoch": 3.0121079793568875,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004105319111020714,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18021228909492493,
      "num_tokens": 66491451.0,
      "step": 30350
    },
    {
      "entropy": 5.6067383766174315,
      "epoch": 3.0126042080190554,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041050362724002413,
      "loss": 4.995,
      "mean_token_accuracy": 0.18067421168088912,
      "num_tokens": 66503179.0,
      "step": 30355
    },
    {
      "entropy": 5.598499727249146,
      "epoch": 3.013100436681223,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041047534001777884,
      "loss": 4.9428,
      "mean_token_accuracy": 0.19373321235179902,
      "num_tokens": 66513422.0,
      "step": 30360
    },
    {
      "entropy": 5.577413320541382,
      "epoch": 3.0135966653433903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004104470494360371,
      "loss": 4.8656,
      "mean_token_accuracy": 0.19481759071350097,
      "num_tokens": 66522948.0,
      "step": 30365
    },
    {
      "entropy": 5.528187799453735,
      "epoch": 3.014092894005558,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041041875549550043,
      "loss": 5.0079,
      "mean_token_accuracy": 0.19042045027017593,
      "num_tokens": 66534012.0,
      "step": 30370
    },
    {
      "entropy": 5.618349027633667,
      "epoch": 3.0145891226677253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004103904581968703,
      "loss": 4.965,
      "mean_token_accuracy": 0.1938430979847908,
      "num_tokens": 66544751.0,
      "step": 30375
    },
    {
      "entropy": 5.687828636169433,
      "epoch": 3.0150853513298927,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041036215754084864,
      "loss": 4.9627,
      "mean_token_accuracy": 0.19124780148267745,
      "num_tokens": 66554808.0,
      "step": 30380
    },
    {
      "entropy": 5.6335203647613525,
      "epoch": 3.0155815799920602,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041033385352813714,
      "loss": 5.0142,
      "mean_token_accuracy": 0.18959670662879943,
      "num_tokens": 66566496.0,
      "step": 30385
    },
    {
      "entropy": 5.60981764793396,
      "epoch": 3.0160778086542277,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004103055461594377,
      "loss": 4.924,
      "mean_token_accuracy": 0.19220445901155472,
      "num_tokens": 66577333.0,
      "step": 30390
    },
    {
      "entropy": 5.593558406829834,
      "epoch": 3.0165740373163956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004102772354354522,
      "loss": 4.9086,
      "mean_token_accuracy": 0.1909906357526779,
      "num_tokens": 66587858.0,
      "step": 30395
    },
    {
      "entropy": 5.616844224929809,
      "epoch": 3.017070265978563,
      "grad_norm": 1.125,
      "learning_rate": 0.00041024892135688287,
      "loss": 4.9437,
      "mean_token_accuracy": 0.18993887603282927,
      "num_tokens": 66598799.0,
      "step": 30400
    },
    {
      "entropy": 5.632180547714233,
      "epoch": 3.0175664946407306,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004102206039244316,
      "loss": 4.9573,
      "mean_token_accuracy": 0.19071339517831803,
      "num_tokens": 66608934.0,
      "step": 30405
    },
    {
      "entropy": 5.653420448303223,
      "epoch": 3.018062723302898,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004101922831388008,
      "loss": 5.0529,
      "mean_token_accuracy": 0.18986130654811859,
      "num_tokens": 66621504.0,
      "step": 30410
    },
    {
      "entropy": 5.679101228713989,
      "epoch": 3.0185589519650655,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041016395900069267,
      "loss": 5.0238,
      "mean_token_accuracy": 0.1804334506392479,
      "num_tokens": 66632673.0,
      "step": 30415
    },
    {
      "entropy": 5.687519884109497,
      "epoch": 3.019055180627233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004101356315108097,
      "loss": 5.014,
      "mean_token_accuracy": 0.18128267526626587,
      "num_tokens": 66644227.0,
      "step": 30420
    },
    {
      "entropy": 5.612639617919922,
      "epoch": 3.0195514092894005,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041010730066985414,
      "loss": 4.9882,
      "mean_token_accuracy": 0.196365188062191,
      "num_tokens": 66654532.0,
      "step": 30425
    },
    {
      "entropy": 5.65359148979187,
      "epoch": 3.020047637951568,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004100789664785287,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18313369750976563,
      "num_tokens": 66665781.0,
      "step": 30430
    },
    {
      "entropy": 5.5887833595275875,
      "epoch": 3.0205438666137354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004100506289375359,
      "loss": 4.8775,
      "mean_token_accuracy": 0.19563065767288207,
      "num_tokens": 66675298.0,
      "step": 30435
    },
    {
      "entropy": 5.596146202087402,
      "epoch": 3.0210400952759033,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004100222880475786,
      "loss": 4.9368,
      "mean_token_accuracy": 0.1849667713046074,
      "num_tokens": 66685686.0,
      "step": 30440
    },
    {
      "entropy": 5.570736646652222,
      "epoch": 3.021536323938071,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040999394380935943,
      "loss": 4.8652,
      "mean_token_accuracy": 0.19665919989347458,
      "num_tokens": 66695407.0,
      "step": 30445
    },
    {
      "entropy": 5.625837326049805,
      "epoch": 3.0220325526002383,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040996559622358145,
      "loss": 4.9719,
      "mean_token_accuracy": 0.18310090005397797,
      "num_tokens": 66706339.0,
      "step": 30450
    },
    {
      "entropy": 5.613515138626099,
      "epoch": 3.0225287812624058,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004099372452909474,
      "loss": 4.9539,
      "mean_token_accuracy": 0.18412170857191085,
      "num_tokens": 66716817.0,
      "step": 30455
    },
    {
      "entropy": 5.737480211257934,
      "epoch": 3.0230250099245732,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004099088910121605,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1900569275021553,
      "num_tokens": 66727575.0,
      "step": 30460
    },
    {
      "entropy": 5.765031480789185,
      "epoch": 3.0235212385867407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040988053338792367,
      "loss": 5.1098,
      "mean_token_accuracy": 0.18197411745786668,
      "num_tokens": 66739181.0,
      "step": 30465
    },
    {
      "entropy": 5.657609844207764,
      "epoch": 3.024017467248908,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004098521724189404,
      "loss": 4.9684,
      "mean_token_accuracy": 0.18664852827787398,
      "num_tokens": 66750781.0,
      "step": 30470
    },
    {
      "entropy": 5.662783098220825,
      "epoch": 3.0245136959110757,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040982380810591386,
      "loss": 5.0486,
      "mean_token_accuracy": 0.18300556093454362,
      "num_tokens": 66761843.0,
      "step": 30475
    },
    {
      "entropy": 5.625848865509033,
      "epoch": 3.025009924573243,
      "grad_norm": 1.125,
      "learning_rate": 0.0004097954404495473,
      "loss": 4.941,
      "mean_token_accuracy": 0.18753099143505098,
      "num_tokens": 66772399.0,
      "step": 30480
    },
    {
      "entropy": 5.609835481643676,
      "epoch": 3.025506153235411,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040976706945054434,
      "loss": 4.933,
      "mean_token_accuracy": 0.1917019546031952,
      "num_tokens": 66782797.0,
      "step": 30485
    },
    {
      "entropy": 5.68047776222229,
      "epoch": 3.0260023818975785,
      "grad_norm": 1.125,
      "learning_rate": 0.00040973869510960837,
      "loss": 5.0186,
      "mean_token_accuracy": 0.18602966964244844,
      "num_tokens": 66792996.0,
      "step": 30490
    },
    {
      "entropy": 5.660466432571411,
      "epoch": 3.026498610559746,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040971031742744314,
      "loss": 4.9838,
      "mean_token_accuracy": 0.1910230502486229,
      "num_tokens": 66802962.0,
      "step": 30495
    },
    {
      "entropy": 5.5977123260498045,
      "epoch": 3.0269948392219135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040968193640475236,
      "loss": 4.9395,
      "mean_token_accuracy": 0.19791232347488402,
      "num_tokens": 66815119.0,
      "step": 30500
    },
    {
      "entropy": 5.639046239852905,
      "epoch": 3.027491067884081,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004096535520422397,
      "loss": 5.0577,
      "mean_token_accuracy": 0.18756525069475175,
      "num_tokens": 66826155.0,
      "step": 30505
    },
    {
      "entropy": 5.6412615299224855,
      "epoch": 3.0279872965462484,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040962516434060915,
      "loss": 4.9618,
      "mean_token_accuracy": 0.19276318103075027,
      "num_tokens": 66836494.0,
      "step": 30510
    },
    {
      "entropy": 5.568629932403565,
      "epoch": 3.028483525208416,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004095967733005646,
      "loss": 4.9299,
      "mean_token_accuracy": 0.18910106867551804,
      "num_tokens": 66847721.0,
      "step": 30515
    },
    {
      "entropy": 5.626459360122681,
      "epoch": 3.0289797538705834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040956837892281014,
      "loss": 4.9945,
      "mean_token_accuracy": 0.18208910524845123,
      "num_tokens": 66858982.0,
      "step": 30520
    },
    {
      "entropy": 5.622489309310913,
      "epoch": 3.0294759825327513,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040953998120804976,
      "loss": 4.9602,
      "mean_token_accuracy": 0.19051127582788469,
      "num_tokens": 66870469.0,
      "step": 30525
    },
    {
      "entropy": 5.588276100158692,
      "epoch": 3.0299722111949188,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095115801569878,
      "loss": 4.9335,
      "mean_token_accuracy": 0.18946251422166824,
      "num_tokens": 66880622.0,
      "step": 30530
    },
    {
      "entropy": 5.6450514793396,
      "epoch": 3.0304684398570862,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004094831757703285,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18029366284608841,
      "num_tokens": 66892089.0,
      "step": 30535
    },
    {
      "entropy": 5.535994005203247,
      "epoch": 3.0309646685192537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040945476804877623,
      "loss": 4.8305,
      "mean_token_accuracy": 0.20070065408945084,
      "num_tokens": 66903086.0,
      "step": 30540
    },
    {
      "entropy": 5.574310684204102,
      "epoch": 3.031460897181421,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004094263569930355,
      "loss": 4.9606,
      "mean_token_accuracy": 0.18492578715085983,
      "num_tokens": 66913871.0,
      "step": 30545
    },
    {
      "entropy": 5.667612504959107,
      "epoch": 3.0319571258435887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040939794260381074,
      "loss": 5.0173,
      "mean_token_accuracy": 0.19038795083761215,
      "num_tokens": 66924358.0,
      "step": 30550
    },
    {
      "entropy": 5.701993370056153,
      "epoch": 3.032453354505756,
      "grad_norm": 1.125,
      "learning_rate": 0.0004093695248818065,
      "loss": 5.0322,
      "mean_token_accuracy": 0.19016535133123397,
      "num_tokens": 66934669.0,
      "step": 30555
    },
    {
      "entropy": 5.679090785980224,
      "epoch": 3.0329495831679236,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004093411038277277,
      "loss": 4.9917,
      "mean_token_accuracy": 0.18881332874298096,
      "num_tokens": 66944255.0,
      "step": 30560
    },
    {
      "entropy": 5.63826355934143,
      "epoch": 3.033445811830091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000409312679442279,
      "loss": 4.9752,
      "mean_token_accuracy": 0.1896502524614334,
      "num_tokens": 66955259.0,
      "step": 30565
    },
    {
      "entropy": 5.658491134643555,
      "epoch": 3.033942040492259,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040928425172616524,
      "loss": 5.011,
      "mean_token_accuracy": 0.1847146436572075,
      "num_tokens": 66966098.0,
      "step": 30570
    },
    {
      "entropy": 5.701732206344604,
      "epoch": 3.0344382691544265,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040925582068009145,
      "loss": 5.0619,
      "mean_token_accuracy": 0.18235949724912642,
      "num_tokens": 66976638.0,
      "step": 30575
    },
    {
      "entropy": 5.6026078224182125,
      "epoch": 3.034934497816594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004092273863047625,
      "loss": 4.9286,
      "mean_token_accuracy": 0.19282049983739852,
      "num_tokens": 66986835.0,
      "step": 30580
    },
    {
      "entropy": 5.654129838943481,
      "epoch": 3.0354307264787614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040919894860088366,
      "loss": 5.0322,
      "mean_token_accuracy": 0.18816171288490297,
      "num_tokens": 66998057.0,
      "step": 30585
    },
    {
      "entropy": 5.608001804351806,
      "epoch": 3.035926955140929,
      "grad_norm": 1.0,
      "learning_rate": 0.0004091705075691601,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18589043617248535,
      "num_tokens": 67010067.0,
      "step": 30590
    },
    {
      "entropy": 5.622958517074585,
      "epoch": 3.0364231838030964,
      "grad_norm": 1.703125,
      "learning_rate": 0.000409142063210297,
      "loss": 4.894,
      "mean_token_accuracy": 0.19588921070098878,
      "num_tokens": 67021256.0,
      "step": 30595
    },
    {
      "entropy": 5.677024078369141,
      "epoch": 3.036919412465264,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040911361552499974,
      "loss": 4.9845,
      "mean_token_accuracy": 0.18409974575042726,
      "num_tokens": 67032345.0,
      "step": 30600
    },
    {
      "entropy": 5.603674364089966,
      "epoch": 3.0374156411274313,
      "grad_norm": 1.34375,
      "learning_rate": 0.00040908516451397395,
      "loss": 4.9212,
      "mean_token_accuracy": 0.19324565827846527,
      "num_tokens": 67042756.0,
      "step": 30605
    },
    {
      "entropy": 5.627187299728393,
      "epoch": 3.0379118697895993,
      "grad_norm": 1.25,
      "learning_rate": 0.0004090567101779249,
      "loss": 5.006,
      "mean_token_accuracy": 0.1856330707669258,
      "num_tokens": 67053825.0,
      "step": 30610
    },
    {
      "entropy": 5.627196311950684,
      "epoch": 3.0384080984517667,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004090282525175583,
      "loss": 4.9925,
      "mean_token_accuracy": 0.18639818876981734,
      "num_tokens": 67063547.0,
      "step": 30615
    },
    {
      "entropy": 5.606569051742554,
      "epoch": 3.038904327113934,
      "grad_norm": 1.125,
      "learning_rate": 0.0004089997915335799,
      "loss": 4.9508,
      "mean_token_accuracy": 0.1932611271739006,
      "num_tokens": 67075333.0,
      "step": 30620
    },
    {
      "entropy": 5.706089687347412,
      "epoch": 3.0394005557761017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004089713272266954,
      "loss": 5.0333,
      "mean_token_accuracy": 0.1825876548886299,
      "num_tokens": 67086817.0,
      "step": 30625
    },
    {
      "entropy": 5.598360204696656,
      "epoch": 3.039896784438269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004089428595976106,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19448540061712266,
      "num_tokens": 67098532.0,
      "step": 30630
    },
    {
      "entropy": 5.649130010604859,
      "epoch": 3.0403930131004366,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004089143886470316,
      "loss": 5.0211,
      "mean_token_accuracy": 0.1845538079738617,
      "num_tokens": 67109053.0,
      "step": 30635
    },
    {
      "entropy": 5.637075805664063,
      "epoch": 3.040889241762604,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004088859143756643,
      "loss": 5.0229,
      "mean_token_accuracy": 0.18958151638507842,
      "num_tokens": 67121821.0,
      "step": 30640
    },
    {
      "entropy": 5.65146803855896,
      "epoch": 3.0413854704247716,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004088574367842148,
      "loss": 4.9201,
      "mean_token_accuracy": 0.18795784413814545,
      "num_tokens": 67133248.0,
      "step": 30645
    },
    {
      "entropy": 5.63796706199646,
      "epoch": 3.041881699086939,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004088289558733893,
      "loss": 4.986,
      "mean_token_accuracy": 0.19146591126918794,
      "num_tokens": 67144884.0,
      "step": 30650
    },
    {
      "entropy": 5.651608037948608,
      "epoch": 3.042377927749107,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004088004716438941,
      "loss": 4.9224,
      "mean_token_accuracy": 0.18374719321727753,
      "num_tokens": 67155671.0,
      "step": 30655
    },
    {
      "entropy": 5.6129964828491214,
      "epoch": 3.0428741564112745,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004087719840964355,
      "loss": 4.9689,
      "mean_token_accuracy": 0.1889234662055969,
      "num_tokens": 67166038.0,
      "step": 30660
    },
    {
      "entropy": 5.628834199905396,
      "epoch": 3.043370385073442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040874349323171994,
      "loss": 5.0174,
      "mean_token_accuracy": 0.18342244029045104,
      "num_tokens": 67177235.0,
      "step": 30665
    },
    {
      "entropy": 5.7034827709198,
      "epoch": 3.0438666137356094,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004087149990504539,
      "loss": 5.0338,
      "mean_token_accuracy": 0.18503244519233703,
      "num_tokens": 67187681.0,
      "step": 30670
    },
    {
      "entropy": 5.59188437461853,
      "epoch": 3.044362842397777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040868650155334406,
      "loss": 4.8964,
      "mean_token_accuracy": 0.1948375090956688,
      "num_tokens": 67199303.0,
      "step": 30675
    },
    {
      "entropy": 5.577424955368042,
      "epoch": 3.0448590710599444,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040865800074109713,
      "loss": 4.9794,
      "mean_token_accuracy": 0.19271010607481004,
      "num_tokens": 67210792.0,
      "step": 30680
    },
    {
      "entropy": 5.557029581069946,
      "epoch": 3.045355299722112,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004086294966144196,
      "loss": 4.9735,
      "mean_token_accuracy": 0.18395899683237077,
      "num_tokens": 67222690.0,
      "step": 30685
    },
    {
      "entropy": 5.6125936031341555,
      "epoch": 3.0458515283842793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004086009891740187,
      "loss": 4.9448,
      "mean_token_accuracy": 0.1871587261557579,
      "num_tokens": 67232957.0,
      "step": 30690
    },
    {
      "entropy": 5.5837537288665775,
      "epoch": 3.046347757046447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040857247842060105,
      "loss": 4.8977,
      "mean_token_accuracy": 0.19316088855266572,
      "num_tokens": 67243270.0,
      "step": 30695
    },
    {
      "entropy": 5.637495136260986,
      "epoch": 3.0468439857086147,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004085439643548737,
      "loss": 5.0122,
      "mean_token_accuracy": 0.1858732059597969,
      "num_tokens": 67253620.0,
      "step": 30700
    },
    {
      "entropy": 5.603529357910157,
      "epoch": 3.047340214370782,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004085154469775439,
      "loss": 4.9002,
      "mean_token_accuracy": 0.19014730900526047,
      "num_tokens": 67263474.0,
      "step": 30705
    },
    {
      "entropy": 5.611464977264404,
      "epoch": 3.0478364430329496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004084869262893187,
      "loss": 4.9717,
      "mean_token_accuracy": 0.1916084259748459,
      "num_tokens": 67274344.0,
      "step": 30710
    },
    {
      "entropy": 5.667855644226075,
      "epoch": 3.048332671695117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004084584022909053,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18332767784595488,
      "num_tokens": 67284425.0,
      "step": 30715
    },
    {
      "entropy": 5.586198711395264,
      "epoch": 3.0488289003572846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040842987498301115,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19987361431121825,
      "num_tokens": 67295127.0,
      "step": 30720
    },
    {
      "entropy": 5.658857727050782,
      "epoch": 3.049325129019452,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004084013443663437,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18110522478818894,
      "num_tokens": 67307420.0,
      "step": 30725
    },
    {
      "entropy": 5.586784934997558,
      "epoch": 3.0498213576816195,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004083728104416102,
      "loss": 4.9342,
      "mean_token_accuracy": 0.18986494690179825,
      "num_tokens": 67319309.0,
      "step": 30730
    },
    {
      "entropy": 5.646332836151123,
      "epoch": 3.050317586343787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004083442732095185,
      "loss": 5.0203,
      "mean_token_accuracy": 0.18145884126424788,
      "num_tokens": 67332214.0,
      "step": 30735
    },
    {
      "entropy": 5.624533176422119,
      "epoch": 3.050813815005955,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004083157326707762,
      "loss": 4.9588,
      "mean_token_accuracy": 0.19117061495780946,
      "num_tokens": 67345287.0,
      "step": 30740
    },
    {
      "entropy": 5.628713941574096,
      "epoch": 3.0513100436681224,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040828718882609085,
      "loss": 4.9134,
      "mean_token_accuracy": 0.1880913645029068,
      "num_tokens": 67355503.0,
      "step": 30745
    },
    {
      "entropy": 5.62474684715271,
      "epoch": 3.05180627233029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040825864167617047,
      "loss": 5.0237,
      "mean_token_accuracy": 0.18269980996847152,
      "num_tokens": 67367216.0,
      "step": 30750
    },
    {
      "entropy": 5.5587821960449215,
      "epoch": 3.0523025009924574,
      "grad_norm": 1.125,
      "learning_rate": 0.000408230091221723,
      "loss": 4.9428,
      "mean_token_accuracy": 0.19062643200159074,
      "num_tokens": 67378985.0,
      "step": 30755
    },
    {
      "entropy": 5.613412761688233,
      "epoch": 3.052798729654625,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004082015374634563,
      "loss": 4.96,
      "mean_token_accuracy": 0.19569221436977385,
      "num_tokens": 67390308.0,
      "step": 30760
    },
    {
      "entropy": 5.662452602386475,
      "epoch": 3.0532949583167923,
      "grad_norm": 1.125,
      "learning_rate": 0.0004081729804020785,
      "loss": 5.0445,
      "mean_token_accuracy": 0.18697765171527864,
      "num_tokens": 67401153.0,
      "step": 30765
    },
    {
      "entropy": 5.566605854034424,
      "epoch": 3.05379118697896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004081444200382978,
      "loss": 4.9343,
      "mean_token_accuracy": 0.19012220203876495,
      "num_tokens": 67412205.0,
      "step": 30770
    },
    {
      "entropy": 5.612611770629883,
      "epoch": 3.0542874156411273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040811585637282234,
      "loss": 4.975,
      "mean_token_accuracy": 0.18630310148000717,
      "num_tokens": 67422994.0,
      "step": 30775
    },
    {
      "entropy": 5.6818108558654785,
      "epoch": 3.0547836443032947,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004080872894063605,
      "loss": 5.0135,
      "mean_token_accuracy": 0.1855757787823677,
      "num_tokens": 67432666.0,
      "step": 30780
    },
    {
      "entropy": 5.655606698989868,
      "epoch": 3.0552798729654627,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040805871913962056,
      "loss": 4.9819,
      "mean_token_accuracy": 0.19281812906265258,
      "num_tokens": 67442480.0,
      "step": 30785
    },
    {
      "entropy": 5.591313982009888,
      "epoch": 3.05577610162763,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004080301455733112,
      "loss": 4.9574,
      "mean_token_accuracy": 0.1856932520866394,
      "num_tokens": 67453142.0,
      "step": 30790
    },
    {
      "entropy": 5.687723541259766,
      "epoch": 3.0562723302897976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040800156870814087,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18907384127378463,
      "num_tokens": 67464525.0,
      "step": 30795
    },
    {
      "entropy": 5.657848834991455,
      "epoch": 3.056768558951965,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040797298854481826,
      "loss": 4.9592,
      "mean_token_accuracy": 0.18947216421365737,
      "num_tokens": 67474282.0,
      "step": 30800
    },
    {
      "entropy": 5.590458583831787,
      "epoch": 3.0572647876141326,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040794440508405195,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18557136803865432,
      "num_tokens": 67485499.0,
      "step": 30805
    },
    {
      "entropy": 5.612951469421387,
      "epoch": 3.0577610162763,
      "grad_norm": 1.125,
      "learning_rate": 0.000407915818326551,
      "loss": 4.9788,
      "mean_token_accuracy": 0.1874785229563713,
      "num_tokens": 67495306.0,
      "step": 30810
    },
    {
      "entropy": 5.633898830413818,
      "epoch": 3.0582572449384675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000407887228273024,
      "loss": 4.9282,
      "mean_token_accuracy": 0.18699049353599548,
      "num_tokens": 67506044.0,
      "step": 30815
    },
    {
      "entropy": 5.6624541759490965,
      "epoch": 3.058753473600635,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004078586349241802,
      "loss": 5.0359,
      "mean_token_accuracy": 0.19167490452528,
      "num_tokens": 67518028.0,
      "step": 30820
    },
    {
      "entropy": 5.631259298324585,
      "epoch": 3.059249702262803,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040783003828072855,
      "loss": 4.966,
      "mean_token_accuracy": 0.19185076653957367,
      "num_tokens": 67528350.0,
      "step": 30825
    },
    {
      "entropy": 5.6741101264953615,
      "epoch": 3.0597459309249704,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004078014383433782,
      "loss": 5.0907,
      "mean_token_accuracy": 0.17759838700294495,
      "num_tokens": 67540341.0,
      "step": 30830
    },
    {
      "entropy": 5.6512072563171385,
      "epoch": 3.060242159587138,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004077728351128383,
      "loss": 4.9384,
      "mean_token_accuracy": 0.19312791973352433,
      "num_tokens": 67551185.0,
      "step": 30835
    },
    {
      "entropy": 5.7169403553009035,
      "epoch": 3.0607383882493053,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004077442285898181,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17734378427267075,
      "num_tokens": 67561262.0,
      "step": 30840
    },
    {
      "entropy": 5.643795776367187,
      "epoch": 3.061234616911473,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040771561877502727,
      "loss": 4.9532,
      "mean_token_accuracy": 0.19547418057918547,
      "num_tokens": 67572342.0,
      "step": 30845
    },
    {
      "entropy": 5.6745171546936035,
      "epoch": 3.0617308455736403,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004076870056691749,
      "loss": 5.0126,
      "mean_token_accuracy": 0.1825897827744484,
      "num_tokens": 67582452.0,
      "step": 30850
    },
    {
      "entropy": 5.619982433319092,
      "epoch": 3.0622270742358078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040765838927297077,
      "loss": 4.9946,
      "mean_token_accuracy": 0.19018048644065857,
      "num_tokens": 67594665.0,
      "step": 30855
    },
    {
      "entropy": 5.664794492721557,
      "epoch": 3.0627233028979752,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004076297695871244,
      "loss": 5.0387,
      "mean_token_accuracy": 0.18076572120189666,
      "num_tokens": 67605419.0,
      "step": 30860
    },
    {
      "entropy": 5.661939525604248,
      "epoch": 3.0632195315601427,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040760114661234553,
      "loss": 5.0001,
      "mean_token_accuracy": 0.1821002945303917,
      "num_tokens": 67617320.0,
      "step": 30865
    },
    {
      "entropy": 5.728667306900024,
      "epoch": 3.0637157602223106,
      "grad_norm": 1.03125,
      "learning_rate": 0.000407572520349344,
      "loss": 5.0522,
      "mean_token_accuracy": 0.1841203063726425,
      "num_tokens": 67629464.0,
      "step": 30870
    },
    {
      "entropy": 5.653510522842407,
      "epoch": 3.064211988884478,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004075438907988296,
      "loss": 4.9923,
      "mean_token_accuracy": 0.18542294353246688,
      "num_tokens": 67640972.0,
      "step": 30875
    },
    {
      "entropy": 5.635851573944092,
      "epoch": 3.0647082175466456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004075152579615124,
      "loss": 4.9498,
      "mean_token_accuracy": 0.1968692809343338,
      "num_tokens": 67651261.0,
      "step": 30880
    },
    {
      "entropy": 5.719280195236206,
      "epoch": 3.065204446208813,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004074866218381023,
      "loss": 5.0684,
      "mean_token_accuracy": 0.17963047623634337,
      "num_tokens": 67661699.0,
      "step": 30885
    },
    {
      "entropy": 5.598703670501709,
      "epoch": 3.0657006748709805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040745798242930947,
      "loss": 4.9223,
      "mean_token_accuracy": 0.19116218835115434,
      "num_tokens": 67671451.0,
      "step": 30890
    },
    {
      "entropy": 5.64216833114624,
      "epoch": 3.066196903533148,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000407429339735844,
      "loss": 4.9817,
      "mean_token_accuracy": 0.1816502705216408,
      "num_tokens": 67681494.0,
      "step": 30895
    },
    {
      "entropy": 5.631985664367676,
      "epoch": 3.0666931321953155,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040740069375841637,
      "loss": 4.9738,
      "mean_token_accuracy": 0.1868586078286171,
      "num_tokens": 67691490.0,
      "step": 30900
    },
    {
      "entropy": 5.590925931930542,
      "epoch": 3.067189360857483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004073720444977368,
      "loss": 4.9389,
      "mean_token_accuracy": 0.1848338782787323,
      "num_tokens": 67701803.0,
      "step": 30905
    },
    {
      "entropy": 5.72981219291687,
      "epoch": 3.067685589519651,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004073433919545157,
      "loss": 5.1161,
      "mean_token_accuracy": 0.18041064888238906,
      "num_tokens": 67713718.0,
      "step": 30910
    },
    {
      "entropy": 5.602089595794678,
      "epoch": 3.0681818181818183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004073147361294638,
      "loss": 4.9099,
      "mean_token_accuracy": 0.1941933110356331,
      "num_tokens": 67725731.0,
      "step": 30915
    },
    {
      "entropy": 5.692307710647583,
      "epoch": 3.068678046843986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004072860770232915,
      "loss": 5.0167,
      "mean_token_accuracy": 0.19417853206396102,
      "num_tokens": 67736527.0,
      "step": 30920
    },
    {
      "entropy": 5.7080260753631595,
      "epoch": 3.0691742755061533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004072574146367095,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18577313274145127,
      "num_tokens": 67747157.0,
      "step": 30925
    },
    {
      "entropy": 5.668478918075562,
      "epoch": 3.0696705041683208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040722874897042863,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18620286881923676,
      "num_tokens": 67758344.0,
      "step": 30930
    },
    {
      "entropy": 5.55354380607605,
      "epoch": 3.0701667328304882,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004072000800251597,
      "loss": 4.9489,
      "mean_token_accuracy": 0.18932622969150542,
      "num_tokens": 67769011.0,
      "step": 30935
    },
    {
      "entropy": 5.590779495239258,
      "epoch": 3.0706629614926557,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040717140780161367,
      "loss": 5.0083,
      "mean_token_accuracy": 0.18521087914705275,
      "num_tokens": 67779916.0,
      "step": 30940
    },
    {
      "entropy": 5.651741981506348,
      "epoch": 3.071159190154823,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040714273230050153,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18912963718175888,
      "num_tokens": 67790061.0,
      "step": 30945
    },
    {
      "entropy": 5.6285484790802,
      "epoch": 3.0716554188169907,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004071140535225343,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18623408079147338,
      "num_tokens": 67801729.0,
      "step": 30950
    },
    {
      "entropy": 5.627841472625732,
      "epoch": 3.0721516474791586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004070853714684233,
      "loss": 5.0397,
      "mean_token_accuracy": 0.18440906405448915,
      "num_tokens": 67812913.0,
      "step": 30955
    },
    {
      "entropy": 5.636021709442138,
      "epoch": 3.072647876141326,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004070566861388797,
      "loss": 4.9479,
      "mean_token_accuracy": 0.19627901315689086,
      "num_tokens": 67824098.0,
      "step": 30960
    },
    {
      "entropy": 5.615736722946167,
      "epoch": 3.0731441048034935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040702799753461477,
      "loss": 4.9767,
      "mean_token_accuracy": 0.18788774609565734,
      "num_tokens": 67835965.0,
      "step": 30965
    },
    {
      "entropy": 5.626912307739258,
      "epoch": 3.073640333465661,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004069993056563401,
      "loss": 4.9453,
      "mean_token_accuracy": 0.1929906576871872,
      "num_tokens": 67847599.0,
      "step": 30970
    },
    {
      "entropy": 5.742709732055664,
      "epoch": 3.0741365621278285,
      "grad_norm": 1.15625,
      "learning_rate": 0.000406970610504767,
      "loss": 5.0362,
      "mean_token_accuracy": 0.1785975843667984,
      "num_tokens": 67857455.0,
      "step": 30975
    },
    {
      "entropy": 5.685127305984497,
      "epoch": 3.074632790789996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004069419120806072,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18091536015272142,
      "num_tokens": 67868571.0,
      "step": 30980
    },
    {
      "entropy": 5.630264091491699,
      "epoch": 3.0751290194521634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040691321038457224,
      "loss": 4.9966,
      "mean_token_accuracy": 0.19155528843402864,
      "num_tokens": 67880412.0,
      "step": 30985
    },
    {
      "entropy": 5.6437445163726805,
      "epoch": 3.075625248114331,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004068845054173739,
      "loss": 4.9258,
      "mean_token_accuracy": 0.18491489440202713,
      "num_tokens": 67889649.0,
      "step": 30990
    },
    {
      "entropy": 5.597090482711792,
      "epoch": 3.076121476776499,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004068557971797241,
      "loss": 4.8817,
      "mean_token_accuracy": 0.1953936070203781,
      "num_tokens": 67901036.0,
      "step": 30995
    },
    {
      "entropy": 5.678354024887085,
      "epoch": 3.0766177054386663,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040682708567233464,
      "loss": 5.0232,
      "mean_token_accuracy": 0.17726509124040604,
      "num_tokens": 67911958.0,
      "step": 31000
    },
    {
      "entropy": 5.623543071746826,
      "epoch": 3.077113934100834,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040679837089591743,
      "loss": 4.9264,
      "mean_token_accuracy": 0.1860221207141876,
      "num_tokens": 67923501.0,
      "step": 31005
    },
    {
      "entropy": 5.629914665222168,
      "epoch": 3.0776101627630013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040676965285118465,
      "loss": 4.9878,
      "mean_token_accuracy": 0.19251036942005156,
      "num_tokens": 67933240.0,
      "step": 31010
    },
    {
      "entropy": 5.608217334747314,
      "epoch": 3.0781063914251687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004067409315388484,
      "loss": 4.9794,
      "mean_token_accuracy": 0.1934982180595398,
      "num_tokens": 67943976.0,
      "step": 31015
    },
    {
      "entropy": 5.659818124771118,
      "epoch": 3.078602620087336,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040671220695962104,
      "loss": 4.9948,
      "mean_token_accuracy": 0.18868274986743927,
      "num_tokens": 67955397.0,
      "step": 31020
    },
    {
      "entropy": 5.6027027606964115,
      "epoch": 3.0790988487495037,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004066834791142146,
      "loss": 5.0463,
      "mean_token_accuracy": 0.18508544117212294,
      "num_tokens": 67965818.0,
      "step": 31025
    },
    {
      "entropy": 5.670929718017578,
      "epoch": 3.079595077411671,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004066547480033417,
      "loss": 5.042,
      "mean_token_accuracy": 0.180736143887043,
      "num_tokens": 67975763.0,
      "step": 31030
    },
    {
      "entropy": 5.617731046676636,
      "epoch": 3.0800913060738386,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004066260136277148,
      "loss": 4.9351,
      "mean_token_accuracy": 0.1899925574660301,
      "num_tokens": 67985982.0,
      "step": 31035
    },
    {
      "entropy": 5.594274854660034,
      "epoch": 3.0805875347360065,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040659727598804635,
      "loss": 4.9587,
      "mean_token_accuracy": 0.19618308693170547,
      "num_tokens": 67997115.0,
      "step": 31040
    },
    {
      "entropy": 5.561894321441651,
      "epoch": 3.081083763398174,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040656853508504897,
      "loss": 4.8967,
      "mean_token_accuracy": 0.20075039565563202,
      "num_tokens": 68008054.0,
      "step": 31045
    },
    {
      "entropy": 5.634236240386963,
      "epoch": 3.0815799920603415,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040653979091943555,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1795438751578331,
      "num_tokens": 68018675.0,
      "step": 31050
    },
    {
      "entropy": 5.632723808288574,
      "epoch": 3.082076220722509,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040651104349191864,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18309897184371948,
      "num_tokens": 68029998.0,
      "step": 31055
    },
    {
      "entropy": 5.7192244052886965,
      "epoch": 3.0825724493846764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040648229280321126,
      "loss": 5.0981,
      "mean_token_accuracy": 0.18125752061605455,
      "num_tokens": 68040436.0,
      "step": 31060
    },
    {
      "entropy": 5.599119186401367,
      "epoch": 3.083068678046844,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004064535388540264,
      "loss": 4.9448,
      "mean_token_accuracy": 0.18884424716234208,
      "num_tokens": 68051479.0,
      "step": 31065
    },
    {
      "entropy": 5.649576091766358,
      "epoch": 3.0835649067090114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004064247816450769,
      "loss": 5.032,
      "mean_token_accuracy": 0.18869945108890535,
      "num_tokens": 68062773.0,
      "step": 31070
    },
    {
      "entropy": 5.574585437774658,
      "epoch": 3.084061135371179,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004063960211770761,
      "loss": 4.9322,
      "mean_token_accuracy": 0.19504079669713975,
      "num_tokens": 68074212.0,
      "step": 31075
    },
    {
      "entropy": 5.57750129699707,
      "epoch": 3.084557364033347,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040636725745073714,
      "loss": 4.8625,
      "mean_token_accuracy": 0.19186593741178512,
      "num_tokens": 68084684.0,
      "step": 31080
    },
    {
      "entropy": 5.591439580917358,
      "epoch": 3.0850535926955143,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004063384904667732,
      "loss": 4.9562,
      "mean_token_accuracy": 0.18428155183792114,
      "num_tokens": 68095481.0,
      "step": 31085
    },
    {
      "entropy": 5.708949184417724,
      "epoch": 3.0855498213576817,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040630972022589774,
      "loss": 5.0196,
      "mean_token_accuracy": 0.19321672469377518,
      "num_tokens": 68107244.0,
      "step": 31090
    },
    {
      "entropy": 5.59012131690979,
      "epoch": 3.086046050019849,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040628094672882415,
      "loss": 4.9635,
      "mean_token_accuracy": 0.1885400265455246,
      "num_tokens": 68120233.0,
      "step": 31095
    },
    {
      "entropy": 5.626113319396973,
      "epoch": 3.0865422786820167,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000406252169976266,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1875481054186821,
      "num_tokens": 68131338.0,
      "step": 31100
    },
    {
      "entropy": 5.653605222702026,
      "epoch": 3.087038507344184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004062233899689368,
      "loss": 4.9693,
      "mean_token_accuracy": 0.1870253100991249,
      "num_tokens": 68143113.0,
      "step": 31105
    },
    {
      "entropy": 5.656967639923096,
      "epoch": 3.0875347360063516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004061946067075503,
      "loss": 5.0459,
      "mean_token_accuracy": 0.17956080585718154,
      "num_tokens": 68153199.0,
      "step": 31110
    },
    {
      "entropy": 5.6514787673950195,
      "epoch": 3.088030964668519,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004061658201928203,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18874994814395904,
      "num_tokens": 68164230.0,
      "step": 31115
    },
    {
      "entropy": 5.575189304351807,
      "epoch": 3.0885271933306866,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004061370304254606,
      "loss": 4.9383,
      "mean_token_accuracy": 0.19383935630321503,
      "num_tokens": 68174612.0,
      "step": 31120
    },
    {
      "entropy": 5.708833265304565,
      "epoch": 3.0890234219928545,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000406108237406185,
      "loss": 4.9674,
      "mean_token_accuracy": 0.1873927891254425,
      "num_tokens": 68184904.0,
      "step": 31125
    },
    {
      "entropy": 5.66887674331665,
      "epoch": 3.089519650655022,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00040607944113570777,
      "loss": 5.083,
      "mean_token_accuracy": 0.18395161181688308,
      "num_tokens": 68195312.0,
      "step": 31130
    },
    {
      "entropy": 5.674493408203125,
      "epoch": 3.0900158793171895,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004060506416147428,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18494034558534622,
      "num_tokens": 68206642.0,
      "step": 31135
    },
    {
      "entropy": 5.62971248626709,
      "epoch": 3.090512107979357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004060218388440043,
      "loss": 4.9285,
      "mean_token_accuracy": 0.18618354499340056,
      "num_tokens": 68217608.0,
      "step": 31140
    },
    {
      "entropy": 5.6361188888549805,
      "epoch": 3.0910083366415244,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004059930328242065,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18650540262460708,
      "num_tokens": 68228649.0,
      "step": 31145
    },
    {
      "entropy": 5.601342391967774,
      "epoch": 3.091504565303692,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040596422355606383,
      "loss": 4.9404,
      "mean_token_accuracy": 0.19019017219543458,
      "num_tokens": 68240157.0,
      "step": 31150
    },
    {
      "entropy": 5.633986711502075,
      "epoch": 3.0920007939658594,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004059354110402905,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1872507616877556,
      "num_tokens": 68250973.0,
      "step": 31155
    },
    {
      "entropy": 5.618847942352295,
      "epoch": 3.092497022628027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040590659527760123,
      "loss": 4.9803,
      "mean_token_accuracy": 0.1898741289973259,
      "num_tokens": 68262498.0,
      "step": 31160
    },
    {
      "entropy": 5.688375234603882,
      "epoch": 3.0929932512901943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040587777626871046,
      "loss": 5.0764,
      "mean_token_accuracy": 0.18066901266574859,
      "num_tokens": 68273122.0,
      "step": 31165
    },
    {
      "entropy": 5.694882440567016,
      "epoch": 3.0934894799523622,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004058489540143328,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18373402804136277,
      "num_tokens": 68283584.0,
      "step": 31170
    },
    {
      "entropy": 5.635001516342163,
      "epoch": 3.0939857086145297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004058201285151831,
      "loss": 4.948,
      "mean_token_accuracy": 0.19324699491262437,
      "num_tokens": 68296381.0,
      "step": 31175
    },
    {
      "entropy": 5.622324466705322,
      "epoch": 3.094481937276697,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004057912997719761,
      "loss": 4.9745,
      "mean_token_accuracy": 0.18292448818683624,
      "num_tokens": 68307292.0,
      "step": 31180
    },
    {
      "entropy": 5.60841932296753,
      "epoch": 3.0949781659388647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004057624677854266,
      "loss": 5.0323,
      "mean_token_accuracy": 0.18437555879354478,
      "num_tokens": 68318418.0,
      "step": 31185
    },
    {
      "entropy": 5.665901517868042,
      "epoch": 3.095474394601032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040573363255624976,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18320154547691345,
      "num_tokens": 68330815.0,
      "step": 31190
    },
    {
      "entropy": 5.706272029876709,
      "epoch": 3.0959706232631996,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004057047940851605,
      "loss": 5.034,
      "mean_token_accuracy": 0.18629360646009446,
      "num_tokens": 68341277.0,
      "step": 31195
    },
    {
      "entropy": 5.653310489654541,
      "epoch": 3.096466851925367,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004056759523728739,
      "loss": 5.0622,
      "mean_token_accuracy": 0.19353277683258058,
      "num_tokens": 68352480.0,
      "step": 31200
    },
    {
      "entropy": 5.596419429779052,
      "epoch": 3.0969630805875346,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040564710742010545,
      "loss": 4.8881,
      "mean_token_accuracy": 0.19002746641635895,
      "num_tokens": 68363327.0,
      "step": 31205
    },
    {
      "entropy": 5.630442333221436,
      "epoch": 3.0974593092497025,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004056182592275702,
      "loss": 5.007,
      "mean_token_accuracy": 0.18722266256809234,
      "num_tokens": 68376039.0,
      "step": 31210
    },
    {
      "entropy": 5.598615741729736,
      "epoch": 3.09795553791187,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004055894077959835,
      "loss": 5.0097,
      "mean_token_accuracy": 0.19075351506471633,
      "num_tokens": 68386485.0,
      "step": 31215
    },
    {
      "entropy": 5.623859214782715,
      "epoch": 3.0984517665740374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004055605531260609,
      "loss": 4.8737,
      "mean_token_accuracy": 0.1976034536957741,
      "num_tokens": 68396889.0,
      "step": 31220
    },
    {
      "entropy": 5.66358208656311,
      "epoch": 3.098947995236205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040553169521851794,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18419030606746672,
      "num_tokens": 68408847.0,
      "step": 31225
    },
    {
      "entropy": 5.638728618621826,
      "epoch": 3.0994442238983724,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040550283407407025,
      "loss": 4.9209,
      "mean_token_accuracy": 0.19268135279417037,
      "num_tokens": 68418347.0,
      "step": 31230
    },
    {
      "entropy": 5.627539110183716,
      "epoch": 3.09994045256054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004054739696934334,
      "loss": 4.9555,
      "mean_token_accuracy": 0.1923694610595703,
      "num_tokens": 68429064.0,
      "step": 31235
    },
    {
      "entropy": 5.659124040603638,
      "epoch": 3.1004366812227073,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040544510207732337,
      "loss": 5.0223,
      "mean_token_accuracy": 0.1906321093440056,
      "num_tokens": 68440946.0,
      "step": 31240
    },
    {
      "entropy": 5.590301513671875,
      "epoch": 3.100932909884875,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040541623122645576,
      "loss": 4.932,
      "mean_token_accuracy": 0.1915037915110588,
      "num_tokens": 68452030.0,
      "step": 31245
    },
    {
      "entropy": 5.616900873184204,
      "epoch": 3.1014291385470423,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00040538735714154675,
      "loss": 4.9438,
      "mean_token_accuracy": 0.19550321102142335,
      "num_tokens": 68462643.0,
      "step": 31250
    },
    {
      "entropy": 5.559059524536133,
      "epoch": 3.10192536720921,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004053584798233122,
      "loss": 4.943,
      "mean_token_accuracy": 0.18800892382860185,
      "num_tokens": 68473169.0,
      "step": 31255
    },
    {
      "entropy": 5.658758926391601,
      "epoch": 3.1024215958713777,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040532959927246814,
      "loss": 4.9728,
      "mean_token_accuracy": 0.18912568390369416,
      "num_tokens": 68483529.0,
      "step": 31260
    },
    {
      "entropy": 5.5945127487182615,
      "epoch": 3.102917824533545,
      "grad_norm": 1.09375,
      "learning_rate": 0.000405300715489731,
      "loss": 4.9077,
      "mean_token_accuracy": 0.1954020455479622,
      "num_tokens": 68494293.0,
      "step": 31265
    },
    {
      "entropy": 5.582448196411133,
      "epoch": 3.1034140531957126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004052718284758169,
      "loss": 4.962,
      "mean_token_accuracy": 0.19044020175933837,
      "num_tokens": 68505153.0,
      "step": 31270
    },
    {
      "entropy": 5.578580665588379,
      "epoch": 3.10391028185788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000405242938231442,
      "loss": 4.927,
      "mean_token_accuracy": 0.18804333806037904,
      "num_tokens": 68515545.0,
      "step": 31275
    },
    {
      "entropy": 5.568288946151734,
      "epoch": 3.1044065105200476,
      "grad_norm": 1.203125,
      "learning_rate": 0.000405214044757323,
      "loss": 4.9537,
      "mean_token_accuracy": 0.19389913678169252,
      "num_tokens": 68526051.0,
      "step": 31280
    },
    {
      "entropy": 5.674172639846802,
      "epoch": 3.104902739182215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004051851480541763,
      "loss": 5.0886,
      "mean_token_accuracy": 0.18339601904153824,
      "num_tokens": 68538141.0,
      "step": 31285
    },
    {
      "entropy": 5.626538038253784,
      "epoch": 3.1053989678443825,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004051562481227184,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18582859188318251,
      "num_tokens": 68548844.0,
      "step": 31290
    },
    {
      "entropy": 5.637701082229614,
      "epoch": 3.10589519650655,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000405127344963666,
      "loss": 4.9506,
      "mean_token_accuracy": 0.19066590964794158,
      "num_tokens": 68558591.0,
      "step": 31295
    },
    {
      "entropy": 5.611349487304688,
      "epoch": 3.106391425168718,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004050984385777359,
      "loss": 4.9327,
      "mean_token_accuracy": 0.19042217433452607,
      "num_tokens": 68569523.0,
      "step": 31300
    },
    {
      "entropy": 5.660407972335816,
      "epoch": 3.1068876538308854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004050695289656448,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18347012400627136,
      "num_tokens": 68580497.0,
      "step": 31305
    },
    {
      "entropy": 5.689936304092408,
      "epoch": 3.107383882493053,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040504061612810964,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18467676490545273,
      "num_tokens": 68590801.0,
      "step": 31310
    },
    {
      "entropy": 5.622058248519897,
      "epoch": 3.1078801111552203,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004050117000658475,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18805556148290634,
      "num_tokens": 68600760.0,
      "step": 31315
    },
    {
      "entropy": 5.572716474533081,
      "epoch": 3.108376339817388,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040498278077957534,
      "loss": 4.9556,
      "mean_token_accuracy": 0.18759870082139968,
      "num_tokens": 68612062.0,
      "step": 31320
    },
    {
      "entropy": 5.711012649536133,
      "epoch": 3.1088725684795553,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004049538582700102,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17764299213886262,
      "num_tokens": 68621980.0,
      "step": 31325
    },
    {
      "entropy": 5.712452125549317,
      "epoch": 3.1093687971417228,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004049249325378695,
      "loss": 4.9918,
      "mean_token_accuracy": 0.1884749114513397,
      "num_tokens": 68632225.0,
      "step": 31330
    },
    {
      "entropy": 5.669777154922485,
      "epoch": 3.1098650258038902,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040489600358387036,
      "loss": 4.9819,
      "mean_token_accuracy": 0.18516893088817596,
      "num_tokens": 68641894.0,
      "step": 31335
    },
    {
      "entropy": 5.642133617401123,
      "epoch": 3.110361254466058,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004048670714087303,
      "loss": 5.0427,
      "mean_token_accuracy": 0.18521158695220946,
      "num_tokens": 68652085.0,
      "step": 31340
    },
    {
      "entropy": 5.632580518722534,
      "epoch": 3.1108574831282256,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004048381360131667,
      "loss": 5.0114,
      "mean_token_accuracy": 0.18957686573266982,
      "num_tokens": 68661486.0,
      "step": 31345
    },
    {
      "entropy": 5.595360851287841,
      "epoch": 3.111353711790393,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004048091973978971,
      "loss": 4.9509,
      "mean_token_accuracy": 0.1961502403020859,
      "num_tokens": 68671504.0,
      "step": 31350
    },
    {
      "entropy": 5.607140445709229,
      "epoch": 3.1118499404525606,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004047802555636391,
      "loss": 4.9308,
      "mean_token_accuracy": 0.18853892832994462,
      "num_tokens": 68681454.0,
      "step": 31355
    },
    {
      "entropy": 5.687619543075561,
      "epoch": 3.112346169114728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004047513105111104,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1812593936920166,
      "num_tokens": 68693356.0,
      "step": 31360
    },
    {
      "entropy": 5.659380340576172,
      "epoch": 3.1128423977768955,
      "grad_norm": 1.328125,
      "learning_rate": 0.00040472236224102883,
      "loss": 4.9211,
      "mean_token_accuracy": 0.1905170664191246,
      "num_tokens": 68704499.0,
      "step": 31365
    },
    {
      "entropy": 5.715338134765625,
      "epoch": 3.113338626439063,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040469341075411217,
      "loss": 5.0674,
      "mean_token_accuracy": 0.18335809409618378,
      "num_tokens": 68716381.0,
      "step": 31370
    },
    {
      "entropy": 5.6687164306640625,
      "epoch": 3.1138348551012305,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004046644560510783,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18315690457820893,
      "num_tokens": 68728042.0,
      "step": 31375
    },
    {
      "entropy": 5.653932332992554,
      "epoch": 3.114331083763398,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040463549813264557,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1865084409713745,
      "num_tokens": 68739296.0,
      "step": 31380
    },
    {
      "entropy": 5.6743687152862545,
      "epoch": 3.114827312425566,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040460653699953166,
      "loss": 4.988,
      "mean_token_accuracy": 0.1874071717262268,
      "num_tokens": 68749375.0,
      "step": 31385
    },
    {
      "entropy": 5.669315195083618,
      "epoch": 3.1153235410877333,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040457757265245485,
      "loss": 4.9234,
      "mean_token_accuracy": 0.19250575751066207,
      "num_tokens": 68760336.0,
      "step": 31390
    },
    {
      "entropy": 5.580797004699707,
      "epoch": 3.115819769749901,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004045486050921336,
      "loss": 4.9741,
      "mean_token_accuracy": 0.19101016223430634,
      "num_tokens": 68771498.0,
      "step": 31395
    },
    {
      "entropy": 5.630480337142944,
      "epoch": 3.1163159984120683,
      "grad_norm": 1.125,
      "learning_rate": 0.000404519634319286,
      "loss": 4.9909,
      "mean_token_accuracy": 0.1921900510787964,
      "num_tokens": 68782180.0,
      "step": 31400
    },
    {
      "entropy": 5.645281410217285,
      "epoch": 3.1168122270742358,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040449066033463053,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18724946230649947,
      "num_tokens": 68792561.0,
      "step": 31405
    },
    {
      "entropy": 5.648979520797729,
      "epoch": 3.1173084557364032,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004044616831388858,
      "loss": 4.9443,
      "mean_token_accuracy": 0.18725570738315583,
      "num_tokens": 68802179.0,
      "step": 31410
    },
    {
      "entropy": 5.5729231357574465,
      "epoch": 3.1178046843985707,
      "grad_norm": 1.125,
      "learning_rate": 0.0004044327027327703,
      "loss": 4.9253,
      "mean_token_accuracy": 0.1917438566684723,
      "num_tokens": 68813652.0,
      "step": 31415
    },
    {
      "entropy": 5.608116483688354,
      "epoch": 3.118300913060738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004044037191170026,
      "loss": 4.9426,
      "mean_token_accuracy": 0.18240556120872498,
      "num_tokens": 68823389.0,
      "step": 31420
    },
    {
      "entropy": 5.591274547576904,
      "epoch": 3.118797141722906,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040437473229230155,
      "loss": 4.9298,
      "mean_token_accuracy": 0.18915115743875505,
      "num_tokens": 68834393.0,
      "step": 31425
    },
    {
      "entropy": 5.555749177932739,
      "epoch": 3.1192933703850736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004043457422593859,
      "loss": 4.8923,
      "mean_token_accuracy": 0.19545914679765702,
      "num_tokens": 68844945.0,
      "step": 31430
    },
    {
      "entropy": 5.605589914321899,
      "epoch": 3.119789599047241,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004043167490189746,
      "loss": 5.0253,
      "mean_token_accuracy": 0.18274757117033005,
      "num_tokens": 68855412.0,
      "step": 31435
    },
    {
      "entropy": 5.647127485275268,
      "epoch": 3.1202858277094085,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040428775257178655,
      "loss": 5.0431,
      "mean_token_accuracy": 0.19185117334127427,
      "num_tokens": 68866976.0,
      "step": 31440
    },
    {
      "entropy": 5.646087503433227,
      "epoch": 3.120782056371576,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040425875291854083,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1852948933839798,
      "num_tokens": 68877670.0,
      "step": 31445
    },
    {
      "entropy": 5.6120623588562015,
      "epoch": 3.1212782850337435,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040422975005995656,
      "loss": 4.9863,
      "mean_token_accuracy": 0.1850049689412117,
      "num_tokens": 68888209.0,
      "step": 31450
    },
    {
      "entropy": 5.658448314666748,
      "epoch": 3.121774513695911,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004042007439967529,
      "loss": 4.9429,
      "mean_token_accuracy": 0.19016427546739578,
      "num_tokens": 68899552.0,
      "step": 31455
    },
    {
      "entropy": 5.639508485794067,
      "epoch": 3.1222707423580784,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004041717347296493,
      "loss": 4.9813,
      "mean_token_accuracy": 0.18661072850227356,
      "num_tokens": 68908990.0,
      "step": 31460
    },
    {
      "entropy": 5.675248479843139,
      "epoch": 3.122766971020246,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004041427222593649,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17926451861858367,
      "num_tokens": 68921320.0,
      "step": 31465
    },
    {
      "entropy": 5.5834095001220705,
      "epoch": 3.123263199682414,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040411370658661935,
      "loss": 4.9738,
      "mean_token_accuracy": 0.18814538270235062,
      "num_tokens": 68931117.0,
      "step": 31470
    },
    {
      "entropy": 5.580458927154541,
      "epoch": 3.1237594283445813,
      "grad_norm": 1.125,
      "learning_rate": 0.00040408468771213206,
      "loss": 4.9523,
      "mean_token_accuracy": 0.191773921251297,
      "num_tokens": 68941168.0,
      "step": 31475
    },
    {
      "entropy": 5.653632688522339,
      "epoch": 3.124255657006749,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040405566563662265,
      "loss": 4.9698,
      "mean_token_accuracy": 0.187161822617054,
      "num_tokens": 68951708.0,
      "step": 31480
    },
    {
      "entropy": 5.649270868301391,
      "epoch": 3.1247518856689163,
      "grad_norm": 1.734375,
      "learning_rate": 0.00040402664036081086,
      "loss": 4.9724,
      "mean_token_accuracy": 0.18970058113336563,
      "num_tokens": 68961013.0,
      "step": 31485
    },
    {
      "entropy": 5.7511796951293945,
      "epoch": 3.1252481143310837,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004039976118854164,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1821807935833931,
      "num_tokens": 68971656.0,
      "step": 31490
    },
    {
      "entropy": 5.678013849258423,
      "epoch": 3.125744342993251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004039685802111591,
      "loss": 4.9904,
      "mean_token_accuracy": 0.1834453210234642,
      "num_tokens": 68983260.0,
      "step": 31495
    },
    {
      "entropy": 5.6840167999267575,
      "epoch": 3.1262405716554187,
      "grad_norm": 1.125,
      "learning_rate": 0.00040393954533875904,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1881084218621254,
      "num_tokens": 68994049.0,
      "step": 31500
    },
    {
      "entropy": 5.6001804828643795,
      "epoch": 3.126736800317586,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040391050726893595,
      "loss": 4.9076,
      "mean_token_accuracy": 0.18974436819553375,
      "num_tokens": 69004697.0,
      "step": 31505
    },
    {
      "entropy": 5.572524547576904,
      "epoch": 3.127233028979754,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004038814660024101,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18325024843215942,
      "num_tokens": 69016362.0,
      "step": 31510
    },
    {
      "entropy": 5.55459508895874,
      "epoch": 3.1277292576419216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004038524215399016,
      "loss": 4.8885,
      "mean_token_accuracy": 0.19281895011663436,
      "num_tokens": 69026836.0,
      "step": 31515
    },
    {
      "entropy": 5.624519968032837,
      "epoch": 3.128225486304089,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004038233738821308,
      "loss": 4.962,
      "mean_token_accuracy": 0.18919406831264496,
      "num_tokens": 69039197.0,
      "step": 31520
    },
    {
      "entropy": 5.671218395233154,
      "epoch": 3.1287217149662565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004037943230298178,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18486249148845674,
      "num_tokens": 69050558.0,
      "step": 31525
    },
    {
      "entropy": 5.657033205032349,
      "epoch": 3.129217943628424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040376526898368314,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18207802027463912,
      "num_tokens": 69061183.0,
      "step": 31530
    },
    {
      "entropy": 5.6422830581665036,
      "epoch": 3.1297141722905915,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040373621174444736,
      "loss": 4.8914,
      "mean_token_accuracy": 0.19724217653274537,
      "num_tokens": 69071471.0,
      "step": 31535
    },
    {
      "entropy": 5.668665361404419,
      "epoch": 3.130210400952759,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040370715131283076,
      "loss": 4.9997,
      "mean_token_accuracy": 0.18604392558336258,
      "num_tokens": 69082066.0,
      "step": 31540
    },
    {
      "entropy": 5.620624017715454,
      "epoch": 3.1307066296149264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004036780876895543,
      "loss": 4.971,
      "mean_token_accuracy": 0.18902055472135543,
      "num_tokens": 69093998.0,
      "step": 31545
    },
    {
      "entropy": 5.573972845077515,
      "epoch": 3.131202858277094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004036490208753385,
      "loss": 4.9085,
      "mean_token_accuracy": 0.19161683470010757,
      "num_tokens": 69105257.0,
      "step": 31550
    },
    {
      "entropy": 5.532020330429077,
      "epoch": 3.131699086939262,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040361995087090407,
      "loss": 4.9755,
      "mean_token_accuracy": 0.19120337665081025,
      "num_tokens": 69114929.0,
      "step": 31555
    },
    {
      "entropy": 5.680608606338501,
      "epoch": 3.1321953156014293,
      "grad_norm": 1.125,
      "learning_rate": 0.00040359087767697223,
      "loss": 4.94,
      "mean_token_accuracy": 0.1955512583255768,
      "num_tokens": 69124894.0,
      "step": 31560
    },
    {
      "entropy": 5.655366230010986,
      "epoch": 3.1326915442635968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040356180129426356,
      "loss": 4.991,
      "mean_token_accuracy": 0.1894986405968666,
      "num_tokens": 69136363.0,
      "step": 31565
    },
    {
      "entropy": 5.681346893310547,
      "epoch": 3.1331877729257642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004035327217234993,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1815449506044388,
      "num_tokens": 69148455.0,
      "step": 31570
    },
    {
      "entropy": 5.640151262283325,
      "epoch": 3.1336840015879317,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040350363896540054,
      "loss": 5.0571,
      "mean_token_accuracy": 0.18686581701040267,
      "num_tokens": 69158381.0,
      "step": 31575
    },
    {
      "entropy": 5.5815941333770756,
      "epoch": 3.134180230250099,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040347455302068835,
      "loss": 4.9173,
      "mean_token_accuracy": 0.18982015699148178,
      "num_tokens": 69171250.0,
      "step": 31580
    },
    {
      "entropy": 5.6253856182098385,
      "epoch": 3.1346764589122667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040344546389008406,
      "loss": 5.0362,
      "mean_token_accuracy": 0.18804918080568314,
      "num_tokens": 69182936.0,
      "step": 31585
    },
    {
      "entropy": 5.651669454574585,
      "epoch": 3.135172687574434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040341637157430916,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18681079447269439,
      "num_tokens": 69193197.0,
      "step": 31590
    },
    {
      "entropy": 5.634455394744873,
      "epoch": 3.135668916236602,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004033872760740849,
      "loss": 4.9595,
      "mean_token_accuracy": 0.19389351606369018,
      "num_tokens": 69204170.0,
      "step": 31595
    },
    {
      "entropy": 5.588116836547852,
      "epoch": 3.1361651448987695,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004033581773901328,
      "loss": 4.978,
      "mean_token_accuracy": 0.18782264441251756,
      "num_tokens": 69217184.0,
      "step": 31600
    },
    {
      "entropy": 5.6485240936279295,
      "epoch": 3.136661373560937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040332907552317453,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17719289064407348,
      "num_tokens": 69227783.0,
      "step": 31605
    },
    {
      "entropy": 5.618285846710205,
      "epoch": 3.1371576022231045,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004032999704739317,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1915598526597023,
      "num_tokens": 69239391.0,
      "step": 31610
    },
    {
      "entropy": 5.692057514190674,
      "epoch": 3.137653830885272,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004032708622431261,
      "loss": 5.0821,
      "mean_token_accuracy": 0.1762917622923851,
      "num_tokens": 69251171.0,
      "step": 31615
    },
    {
      "entropy": 5.71341986656189,
      "epoch": 3.1381500595474394,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004032417508314795,
      "loss": 5.0346,
      "mean_token_accuracy": 0.18416742533445357,
      "num_tokens": 69263833.0,
      "step": 31620
    },
    {
      "entropy": 5.646372604370117,
      "epoch": 3.138646288209607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040321263623971376,
      "loss": 4.9313,
      "mean_token_accuracy": 0.1939586102962494,
      "num_tokens": 69275065.0,
      "step": 31625
    },
    {
      "entropy": 5.49307804107666,
      "epoch": 3.1391425168717744,
      "grad_norm": 1.140625,
      "learning_rate": 0.000403183518468551,
      "loss": 4.8756,
      "mean_token_accuracy": 0.19811997264623643,
      "num_tokens": 69285084.0,
      "step": 31630
    },
    {
      "entropy": 5.663730525970459,
      "epoch": 3.139638745533942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004031543975187132,
      "loss": 5.0009,
      "mean_token_accuracy": 0.18746813237667084,
      "num_tokens": 69295500.0,
      "step": 31635
    },
    {
      "entropy": 5.694372463226318,
      "epoch": 3.1401349741961098,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040312527339092237,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1843213215470314,
      "num_tokens": 69307776.0,
      "step": 31640
    },
    {
      "entropy": 5.601633548736572,
      "epoch": 3.1406312028582772,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040309614608590095,
      "loss": 4.9651,
      "mean_token_accuracy": 0.19165512919425964,
      "num_tokens": 69320011.0,
      "step": 31645
    },
    {
      "entropy": 5.62839093208313,
      "epoch": 3.1411274315204447,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004030670156043711,
      "loss": 4.9421,
      "mean_token_accuracy": 0.1971704363822937,
      "num_tokens": 69329427.0,
      "step": 31650
    },
    {
      "entropy": 5.610344171524048,
      "epoch": 3.141623660182612,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040303788194705523,
      "loss": 4.8846,
      "mean_token_accuracy": 0.19701789021492006,
      "num_tokens": 69340916.0,
      "step": 31655
    },
    {
      "entropy": 5.643756771087647,
      "epoch": 3.1421198888447797,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040300874511467574,
      "loss": 4.8958,
      "mean_token_accuracy": 0.19414390474557877,
      "num_tokens": 69351363.0,
      "step": 31660
    },
    {
      "entropy": 5.650760698318481,
      "epoch": 3.142616117506947,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040297960510795526,
      "loss": 5.0018,
      "mean_token_accuracy": 0.1832362800836563,
      "num_tokens": 69363306.0,
      "step": 31665
    },
    {
      "entropy": 5.600221395492554,
      "epoch": 3.1431123461691146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004029504619276163,
      "loss": 4.9544,
      "mean_token_accuracy": 0.18309537172317505,
      "num_tokens": 69374660.0,
      "step": 31670
    },
    {
      "entropy": 5.6041419982910154,
      "epoch": 3.143608574831282,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040292131557438166,
      "loss": 4.9407,
      "mean_token_accuracy": 0.1920928180217743,
      "num_tokens": 69384537.0,
      "step": 31675
    },
    {
      "entropy": 5.619178342819214,
      "epoch": 3.14410480349345,
      "grad_norm": 1.234375,
      "learning_rate": 0.000402892166048974,
      "loss": 4.9957,
      "mean_token_accuracy": 0.19190824180841445,
      "num_tokens": 69395893.0,
      "step": 31680
    },
    {
      "entropy": 5.634036922454834,
      "epoch": 3.1446010321556175,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004028630133521162,
      "loss": 4.9898,
      "mean_token_accuracy": 0.18968677520751953,
      "num_tokens": 69407829.0,
      "step": 31685
    },
    {
      "entropy": 5.707188367843628,
      "epoch": 3.145097260817785,
      "grad_norm": 1.125,
      "learning_rate": 0.0004028338574845312,
      "loss": 5.1277,
      "mean_token_accuracy": 0.18251728862524033,
      "num_tokens": 69419460.0,
      "step": 31690
    },
    {
      "entropy": 5.6584985733032225,
      "epoch": 3.1455934894799524,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040280469844694203,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18912708312273024,
      "num_tokens": 69429255.0,
      "step": 31695
    },
    {
      "entropy": 5.648843002319336,
      "epoch": 3.14608971814212,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004027755362400716,
      "loss": 5.0013,
      "mean_token_accuracy": 0.1913452684879303,
      "num_tokens": 69439987.0,
      "step": 31700
    },
    {
      "entropy": 5.677603435516358,
      "epoch": 3.1465859468042874,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004027463708646434,
      "loss": 4.9855,
      "mean_token_accuracy": 0.1881941482424736,
      "num_tokens": 69452428.0,
      "step": 31705
    },
    {
      "entropy": 5.651621150970459,
      "epoch": 3.147082175466455,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004027172023213803,
      "loss": 4.9212,
      "mean_token_accuracy": 0.18809480220079422,
      "num_tokens": 69462543.0,
      "step": 31710
    },
    {
      "entropy": 5.674396705627442,
      "epoch": 3.1475784041286223,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004026880306110058,
      "loss": 5.036,
      "mean_token_accuracy": 0.18452268689870835,
      "num_tokens": 69473908.0,
      "step": 31715
    },
    {
      "entropy": 5.648675203323364,
      "epoch": 3.14807463279079,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040265885573424336,
      "loss": 4.9812,
      "mean_token_accuracy": 0.1920025169849396,
      "num_tokens": 69483677.0,
      "step": 31720
    },
    {
      "entropy": 5.6175275325775145,
      "epoch": 3.1485708614529577,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040262967769181634,
      "loss": 4.964,
      "mean_token_accuracy": 0.1908070534467697,
      "num_tokens": 69495390.0,
      "step": 31725
    },
    {
      "entropy": 5.655036354064942,
      "epoch": 3.149067090115125,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040260049648444836,
      "loss": 5.0313,
      "mean_token_accuracy": 0.17699674963951112,
      "num_tokens": 69506867.0,
      "step": 31730
    },
    {
      "entropy": 5.647692108154297,
      "epoch": 3.1495633187772927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040257131211286307,
      "loss": 4.93,
      "mean_token_accuracy": 0.19260971397161483,
      "num_tokens": 69516888.0,
      "step": 31735
    },
    {
      "entropy": 5.749320077896118,
      "epoch": 3.15005954743946,
      "grad_norm": 1.171875,
      "learning_rate": 0.000402542124577784,
      "loss": 5.0856,
      "mean_token_accuracy": 0.18164819180965425,
      "num_tokens": 69528054.0,
      "step": 31740
    },
    {
      "entropy": 5.657619285583496,
      "epoch": 3.1505557761016276,
      "grad_norm": 1.296875,
      "learning_rate": 0.00040251293387993513,
      "loss": 5.0656,
      "mean_token_accuracy": 0.17783531099557875,
      "num_tokens": 69537042.0,
      "step": 31745
    },
    {
      "entropy": 5.617402076721191,
      "epoch": 3.151052004763795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004024837400200403,
      "loss": 4.998,
      "mean_token_accuracy": 0.18778745830059052,
      "num_tokens": 69547835.0,
      "step": 31750
    },
    {
      "entropy": 5.595995903015137,
      "epoch": 3.1515482334259626,
      "grad_norm": 1.125,
      "learning_rate": 0.00040245454299882343,
      "loss": 4.9144,
      "mean_token_accuracy": 0.18555198907852172,
      "num_tokens": 69558442.0,
      "step": 31755
    },
    {
      "entropy": 5.559958839416504,
      "epoch": 3.15204446208813,
      "grad_norm": 1.125,
      "learning_rate": 0.00040242534281700843,
      "loss": 4.9235,
      "mean_token_accuracy": 0.19056300371885299,
      "num_tokens": 69570572.0,
      "step": 31760
    },
    {
      "entropy": 5.649234104156494,
      "epoch": 3.152540690750298,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040239613947531965,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18657036125659943,
      "num_tokens": 69581028.0,
      "step": 31765
    },
    {
      "entropy": 5.643005228042602,
      "epoch": 3.1530369194124654,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040236693297448107,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1858994856476784,
      "num_tokens": 69591347.0,
      "step": 31770
    },
    {
      "entropy": 5.609976148605346,
      "epoch": 3.153533148074633,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004023377233152169,
      "loss": 4.9898,
      "mean_token_accuracy": 0.1877061203122139,
      "num_tokens": 69602599.0,
      "step": 31775
    },
    {
      "entropy": 5.567247438430786,
      "epoch": 3.1540293767368004,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040230851049825173,
      "loss": 4.9591,
      "mean_token_accuracy": 0.19270767867565156,
      "num_tokens": 69614871.0,
      "step": 31780
    },
    {
      "entropy": 5.626673984527588,
      "epoch": 3.154525605398968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040227929452430974,
      "loss": 4.9738,
      "mean_token_accuracy": 0.1872774064540863,
      "num_tokens": 69626384.0,
      "step": 31785
    },
    {
      "entropy": 5.643950080871582,
      "epoch": 3.1550218340611353,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040225007539411553,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18209747076034546,
      "num_tokens": 69637230.0,
      "step": 31790
    },
    {
      "entropy": 5.653741312026978,
      "epoch": 3.155518062723303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004022208531083936,
      "loss": 5.0415,
      "mean_token_accuracy": 0.18203838020563126,
      "num_tokens": 69648718.0,
      "step": 31795
    },
    {
      "entropy": 5.682896852493286,
      "epoch": 3.1560142913854703,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004021916276678688,
      "loss": 5.0451,
      "mean_token_accuracy": 0.18594916611909867,
      "num_tokens": 69658303.0,
      "step": 31800
    },
    {
      "entropy": 5.633527851104736,
      "epoch": 3.1565105200476378,
      "grad_norm": 1.125,
      "learning_rate": 0.00040216239907326555,
      "loss": 5.0213,
      "mean_token_accuracy": 0.1853281080722809,
      "num_tokens": 69669693.0,
      "step": 31805
    },
    {
      "entropy": 5.6459613800048825,
      "epoch": 3.1570067487098052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004021331673253089,
      "loss": 5.0009,
      "mean_token_accuracy": 0.18507904559373856,
      "num_tokens": 69679763.0,
      "step": 31810
    },
    {
      "entropy": 5.559248685836792,
      "epoch": 3.157502977371973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004021039324247236,
      "loss": 4.8419,
      "mean_token_accuracy": 0.19497805386781691,
      "num_tokens": 69690901.0,
      "step": 31815
    },
    {
      "entropy": 5.616631269454956,
      "epoch": 3.1579992060341406,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040207469437223466,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18054184466600418,
      "num_tokens": 69700548.0,
      "step": 31820
    },
    {
      "entropy": 5.661181259155273,
      "epoch": 3.158495434696308,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040204545316856714,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1864597961306572,
      "num_tokens": 69711258.0,
      "step": 31825
    },
    {
      "entropy": 5.602327346801758,
      "epoch": 3.1589916633584756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004020162088144461,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18820270597934724,
      "num_tokens": 69721800.0,
      "step": 31830
    },
    {
      "entropy": 5.689257717132568,
      "epoch": 3.159487892020643,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040198696131059684,
      "loss": 5.0613,
      "mean_token_accuracy": 0.1849336951971054,
      "num_tokens": 69733508.0,
      "step": 31835
    },
    {
      "entropy": 5.6760986804962155,
      "epoch": 3.1599841206828105,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040195771065774455,
      "loss": 4.976,
      "mean_token_accuracy": 0.1878679320216179,
      "num_tokens": 69744410.0,
      "step": 31840
    },
    {
      "entropy": 5.631263780593872,
      "epoch": 3.160480349344978,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004019284568566146,
      "loss": 4.8978,
      "mean_token_accuracy": 0.19475563019514083,
      "num_tokens": 69754755.0,
      "step": 31845
    },
    {
      "entropy": 5.601552820205688,
      "epoch": 3.1609765780071455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040189919990793243,
      "loss": 5.0571,
      "mean_token_accuracy": 0.18420334309339523,
      "num_tokens": 69765634.0,
      "step": 31850
    },
    {
      "entropy": 5.675172472000122,
      "epoch": 3.1614728066693134,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040186993981242353,
      "loss": 5.076,
      "mean_token_accuracy": 0.17103953808546066,
      "num_tokens": 69777635.0,
      "step": 31855
    },
    {
      "entropy": 5.672176122665405,
      "epoch": 3.161969035331481,
      "grad_norm": 1.125,
      "learning_rate": 0.0004018406765708135,
      "loss": 4.9022,
      "mean_token_accuracy": 0.19083661884069442,
      "num_tokens": 69788571.0,
      "step": 31860
    },
    {
      "entropy": 5.658387994766235,
      "epoch": 3.1624652639936484,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040181141018382806,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18861045092344284,
      "num_tokens": 69798685.0,
      "step": 31865
    },
    {
      "entropy": 5.656968212127685,
      "epoch": 3.162961492655816,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040178214065219285,
      "loss": 4.9698,
      "mean_token_accuracy": 0.18588539808988572,
      "num_tokens": 69808828.0,
      "step": 31870
    },
    {
      "entropy": 5.558004379272461,
      "epoch": 3.1634577213179833,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004017528679766337,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19313247203826905,
      "num_tokens": 69820687.0,
      "step": 31875
    },
    {
      "entropy": 5.684401750564575,
      "epoch": 3.163953949980151,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004017235921578767,
      "loss": 5.088,
      "mean_token_accuracy": 0.18027301281690597,
      "num_tokens": 69834302.0,
      "step": 31880
    },
    {
      "entropy": 5.607842016220093,
      "epoch": 3.1644501786423183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040169431319664756,
      "loss": 4.8876,
      "mean_token_accuracy": 0.19474564641714096,
      "num_tokens": 69844707.0,
      "step": 31885
    },
    {
      "entropy": 5.642267894744873,
      "epoch": 3.1649464073044857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004016650310936725,
      "loss": 5.0046,
      "mean_token_accuracy": 0.19344507455825805,
      "num_tokens": 69855064.0,
      "step": 31890
    },
    {
      "entropy": 5.604533672332764,
      "epoch": 3.165442635966653,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004016357458496776,
      "loss": 4.9414,
      "mean_token_accuracy": 0.19819669872522355,
      "num_tokens": 69865643.0,
      "step": 31895
    },
    {
      "entropy": 5.576957416534424,
      "epoch": 3.165938864628821,
      "grad_norm": 1.171875,
      "learning_rate": 0.000401606457465389,
      "loss": 4.9433,
      "mean_token_accuracy": 0.1900026872754097,
      "num_tokens": 69876504.0,
      "step": 31900
    },
    {
      "entropy": 5.648366022109985,
      "epoch": 3.1664350932909886,
      "grad_norm": 1.25,
      "learning_rate": 0.0004015771659415332,
      "loss": 5.004,
      "mean_token_accuracy": 0.18652251213788987,
      "num_tokens": 69886922.0,
      "step": 31905
    },
    {
      "entropy": 5.699649906158447,
      "epoch": 3.166931321953156,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004015478712788364,
      "loss": 5.0128,
      "mean_token_accuracy": 0.18527907133102417,
      "num_tokens": 69897516.0,
      "step": 31910
    },
    {
      "entropy": 5.637491178512573,
      "epoch": 3.1674275506153236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040151857347802506,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1853422537446022,
      "num_tokens": 69908391.0,
      "step": 31915
    },
    {
      "entropy": 5.614643907546997,
      "epoch": 3.167923779277491,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040148927253982577,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18749719858169556,
      "num_tokens": 69919274.0,
      "step": 31920
    },
    {
      "entropy": 5.622378492355347,
      "epoch": 3.1684200079396585,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040145996846496497,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18850743323564528,
      "num_tokens": 69929500.0,
      "step": 31925
    },
    {
      "entropy": 5.661145448684692,
      "epoch": 3.168916236601826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004014306612541695,
      "loss": 4.9821,
      "mean_token_accuracy": 0.19006522446870805,
      "num_tokens": 69940130.0,
      "step": 31930
    },
    {
      "entropy": 5.7338062763214115,
      "epoch": 3.1694124652639935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004014013509081661,
      "loss": 5.0806,
      "mean_token_accuracy": 0.18012312352657317,
      "num_tokens": 69951222.0,
      "step": 31935
    },
    {
      "entropy": 5.6786301612854,
      "epoch": 3.1699086939261614,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040137203742768155,
      "loss": 5.0668,
      "mean_token_accuracy": 0.18208354562520981,
      "num_tokens": 69961542.0,
      "step": 31940
    },
    {
      "entropy": 5.64868311882019,
      "epoch": 3.170404922588329,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004013427208134427,
      "loss": 4.9872,
      "mean_token_accuracy": 0.1802208259701729,
      "num_tokens": 69970671.0,
      "step": 31945
    },
    {
      "entropy": 5.667693996429444,
      "epoch": 3.1709011512504963,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040131340106617673,
      "loss": 4.9592,
      "mean_token_accuracy": 0.18623486608266832,
      "num_tokens": 69982205.0,
      "step": 31950
    },
    {
      "entropy": 5.645547866821289,
      "epoch": 3.171397379912664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004012840781866106,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17920104563236236,
      "num_tokens": 69991888.0,
      "step": 31955
    },
    {
      "entropy": 5.588000535964966,
      "epoch": 3.1718936085748313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040125475217547135,
      "loss": 4.9741,
      "mean_token_accuracy": 0.18367499411106109,
      "num_tokens": 70004001.0,
      "step": 31960
    },
    {
      "entropy": 5.676630258560181,
      "epoch": 3.1723898372369987,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040122542303348634,
      "loss": 4.9505,
      "mean_token_accuracy": 0.18872714340686797,
      "num_tokens": 70014864.0,
      "step": 31965
    },
    {
      "entropy": 5.670525074005127,
      "epoch": 3.172886065899166,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004011960907613828,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1828980028629303,
      "num_tokens": 70025294.0,
      "step": 31970
    },
    {
      "entropy": 5.632301664352417,
      "epoch": 3.1733822945613337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040116675535988825,
      "loss": 5.0236,
      "mean_token_accuracy": 0.1848171681165695,
      "num_tokens": 70036560.0,
      "step": 31975
    },
    {
      "entropy": 5.641480445861816,
      "epoch": 3.173878523223501,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040113741682972994,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18473788797855378,
      "num_tokens": 70047771.0,
      "step": 31980
    },
    {
      "entropy": 5.578259992599487,
      "epoch": 3.174374751885669,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004011080751716356,
      "loss": 4.9096,
      "mean_token_accuracy": 0.1986730992794037,
      "num_tokens": 70058288.0,
      "step": 31985
    },
    {
      "entropy": 5.622065162658691,
      "epoch": 3.1748709805478366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040107873038633255,
      "loss": 4.9772,
      "mean_token_accuracy": 0.1877088338136673,
      "num_tokens": 70069896.0,
      "step": 31990
    },
    {
      "entropy": 5.630265235900879,
      "epoch": 3.175367209210004,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004010493824745487,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18370037972927095,
      "num_tokens": 70080640.0,
      "step": 31995
    },
    {
      "entropy": 5.5925556182861325,
      "epoch": 3.1758634378721715,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040102003143701194,
      "loss": 4.9435,
      "mean_token_accuracy": 0.188647997379303,
      "num_tokens": 70090885.0,
      "step": 32000
    },
    {
      "entropy": 5.618486595153809,
      "epoch": 3.176359666534339,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040099067727444984,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18602531552314758,
      "num_tokens": 70102652.0,
      "step": 32005
    },
    {
      "entropy": 5.615683841705322,
      "epoch": 3.1768558951965065,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004009613199875904,
      "loss": 5.033,
      "mean_token_accuracy": 0.17936387658119202,
      "num_tokens": 70113219.0,
      "step": 32010
    },
    {
      "entropy": 5.699048948287964,
      "epoch": 3.177352123858674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004009319595771617,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18401361107826233,
      "num_tokens": 70124963.0,
      "step": 32015
    },
    {
      "entropy": 5.654812908172607,
      "epoch": 3.1778483525208414,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004009025960438918,
      "loss": 4.9564,
      "mean_token_accuracy": 0.19492590725421904,
      "num_tokens": 70136094.0,
      "step": 32020
    },
    {
      "entropy": 5.681875944137573,
      "epoch": 3.1783445811830093,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040087322938850867,
      "loss": 5.043,
      "mean_token_accuracy": 0.18389632403850556,
      "num_tokens": 70146298.0,
      "step": 32025
    },
    {
      "entropy": 5.607324123382568,
      "epoch": 3.178840809845177,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004008438596117408,
      "loss": 4.966,
      "mean_token_accuracy": 0.18744442760944366,
      "num_tokens": 70159405.0,
      "step": 32030
    },
    {
      "entropy": 5.696175670623779,
      "epoch": 3.1793370385073443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040081448671431635,
      "loss": 5.0555,
      "mean_token_accuracy": 0.18270031958818436,
      "num_tokens": 70170480.0,
      "step": 32035
    },
    {
      "entropy": 5.6372888565063475,
      "epoch": 3.1798332671695118,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040078511069696377,
      "loss": 4.9684,
      "mean_token_accuracy": 0.19276729971170425,
      "num_tokens": 70182539.0,
      "step": 32040
    },
    {
      "entropy": 5.664832687377929,
      "epoch": 3.1803294958316792,
      "grad_norm": 1.125,
      "learning_rate": 0.00040075573156041146,
      "loss": 4.9791,
      "mean_token_accuracy": 0.19468171894550323,
      "num_tokens": 70193750.0,
      "step": 32045
    },
    {
      "entropy": 5.6440342426300045,
      "epoch": 3.1808257244938467,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040072634930538796,
      "loss": 4.9954,
      "mean_token_accuracy": 0.1873960793018341,
      "num_tokens": 70203849.0,
      "step": 32050
    },
    {
      "entropy": 5.617047691345215,
      "epoch": 3.181321953156014,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040069696393262196,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18903241902589799,
      "num_tokens": 70214667.0,
      "step": 32055
    },
    {
      "entropy": 5.576152801513672,
      "epoch": 3.1818181818181817,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004006675754428421,
      "loss": 4.9097,
      "mean_token_accuracy": 0.19865494966506958,
      "num_tokens": 70225369.0,
      "step": 32060
    },
    {
      "entropy": 5.614527225494385,
      "epoch": 3.182314410480349,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040063818383677716,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18349191248416902,
      "num_tokens": 70235764.0,
      "step": 32065
    },
    {
      "entropy": 5.711842679977417,
      "epoch": 3.182810639142517,
      "grad_norm": 1.1875,
      "learning_rate": 0.000400608789115156,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1786705181002617,
      "num_tokens": 70246745.0,
      "step": 32070
    },
    {
      "entropy": 5.646728372573852,
      "epoch": 3.1833068678046845,
      "grad_norm": 1.125,
      "learning_rate": 0.00040057939127870754,
      "loss": 4.9705,
      "mean_token_accuracy": 0.1852160796523094,
      "num_tokens": 70257129.0,
      "step": 32075
    },
    {
      "entropy": 5.661528730392456,
      "epoch": 3.183803096466852,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004005499903281607,
      "loss": 5.0276,
      "mean_token_accuracy": 0.18788153678178787,
      "num_tokens": 70267986.0,
      "step": 32080
    },
    {
      "entropy": 5.6443366527557375,
      "epoch": 3.1842993251290195,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040052058626424464,
      "loss": 4.9682,
      "mean_token_accuracy": 0.19078468531370163,
      "num_tokens": 70279362.0,
      "step": 32085
    },
    {
      "entropy": 5.637982940673828,
      "epoch": 3.184795553791187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004004911790876886,
      "loss": 4.9293,
      "mean_token_accuracy": 0.19518563151359558,
      "num_tokens": 70289522.0,
      "step": 32090
    },
    {
      "entropy": 5.61222882270813,
      "epoch": 3.1852917824533544,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040046176879922156,
      "loss": 4.9151,
      "mean_token_accuracy": 0.19347090423107147,
      "num_tokens": 70301372.0,
      "step": 32095
    },
    {
      "entropy": 5.639336776733399,
      "epoch": 3.185788011115522,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040043235539957315,
      "loss": 4.9932,
      "mean_token_accuracy": 0.19129723459482192,
      "num_tokens": 70311382.0,
      "step": 32100
    },
    {
      "entropy": 5.582602834701538,
      "epoch": 3.1862842397776894,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004004029388894726,
      "loss": 4.8964,
      "mean_token_accuracy": 0.1958796888589859,
      "num_tokens": 70321418.0,
      "step": 32105
    },
    {
      "entropy": 5.614124870300293,
      "epoch": 3.1867804684398573,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004003735192696494,
      "loss": 4.9522,
      "mean_token_accuracy": 0.19103971272706985,
      "num_tokens": 70331396.0,
      "step": 32110
    },
    {
      "entropy": 5.666194248199463,
      "epoch": 3.1872766971020248,
      "grad_norm": 1.109375,
      "learning_rate": 0.000400344096540833,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18597547709941864,
      "num_tokens": 70343382.0,
      "step": 32115
    },
    {
      "entropy": 5.634309959411621,
      "epoch": 3.1877729257641922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040031467070375315,
      "loss": 5.0204,
      "mean_token_accuracy": 0.18289321213960646,
      "num_tokens": 70355414.0,
      "step": 32120
    },
    {
      "entropy": 5.692996501922607,
      "epoch": 3.1882691544263597,
      "grad_norm": 1.25,
      "learning_rate": 0.0004002852417591395,
      "loss": 4.986,
      "mean_token_accuracy": 0.18810425996780394,
      "num_tokens": 70366490.0,
      "step": 32125
    },
    {
      "entropy": 5.625392627716065,
      "epoch": 3.188765383088527,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004002558097077218,
      "loss": 4.9532,
      "mean_token_accuracy": 0.18944575190544127,
      "num_tokens": 70377175.0,
      "step": 32130
    },
    {
      "entropy": 5.594571876525879,
      "epoch": 3.1892616117506947,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004002263745502299,
      "loss": 4.9542,
      "mean_token_accuracy": 0.19223009645938874,
      "num_tokens": 70388110.0,
      "step": 32135
    },
    {
      "entropy": 5.626569318771362,
      "epoch": 3.189757840412862,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040019693628739377,
      "loss": 4.9668,
      "mean_token_accuracy": 0.1941126689314842,
      "num_tokens": 70398336.0,
      "step": 32140
    },
    {
      "entropy": 5.642003393173217,
      "epoch": 3.1902540690750296,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004001674949199434,
      "loss": 5.0574,
      "mean_token_accuracy": 0.18018237948417665,
      "num_tokens": 70408967.0,
      "step": 32145
    },
    {
      "entropy": 5.641814851760865,
      "epoch": 3.190750297737197,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004001380504486089,
      "loss": 5.0198,
      "mean_token_accuracy": 0.19225274175405502,
      "num_tokens": 70419054.0,
      "step": 32150
    },
    {
      "entropy": 5.67811918258667,
      "epoch": 3.191246526399365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004001086028741204,
      "loss": 4.9665,
      "mean_token_accuracy": 0.18711631447076799,
      "num_tokens": 70429312.0,
      "step": 32155
    },
    {
      "entropy": 5.656784534454346,
      "epoch": 3.1917427550615325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004000791521972081,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18271855264902115,
      "num_tokens": 70440233.0,
      "step": 32160
    },
    {
      "entropy": 5.640780687332153,
      "epoch": 3.1922389837237,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040004969841860233,
      "loss": 4.9437,
      "mean_token_accuracy": 0.19305364340543746,
      "num_tokens": 70451642.0,
      "step": 32165
    },
    {
      "entropy": 5.694652509689331,
      "epoch": 3.1927352123858674,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040002024153903347,
      "loss": 4.9896,
      "mean_token_accuracy": 0.18902652561664582,
      "num_tokens": 70462433.0,
      "step": 32170
    },
    {
      "entropy": 5.58720760345459,
      "epoch": 3.193231441048035,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003999907815592321,
      "loss": 4.992,
      "mean_token_accuracy": 0.18389304280281066,
      "num_tokens": 70472723.0,
      "step": 32175
    },
    {
      "entropy": 5.659954643249511,
      "epoch": 3.1937276697102024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039996131847992863,
      "loss": 5.0169,
      "mean_token_accuracy": 0.18721013367176056,
      "num_tokens": 70484250.0,
      "step": 32180
    },
    {
      "entropy": 5.6174828052520756,
      "epoch": 3.19422389837237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003999318523018537,
      "loss": 4.975,
      "mean_token_accuracy": 0.18685137778520583,
      "num_tokens": 70495910.0,
      "step": 32185
    },
    {
      "entropy": 5.5930397510528564,
      "epoch": 3.1947201270345373,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003999023830257382,
      "loss": 4.9147,
      "mean_token_accuracy": 0.18939875215291976,
      "num_tokens": 70506745.0,
      "step": 32190
    },
    {
      "entropy": 5.661487197875976,
      "epoch": 3.1952163556967053,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003998729106523126,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1843820184469223,
      "num_tokens": 70517263.0,
      "step": 32195
    },
    {
      "entropy": 5.606229448318482,
      "epoch": 3.1957125843588727,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003998434351823078,
      "loss": 4.9177,
      "mean_token_accuracy": 0.18996207416057587,
      "num_tokens": 70527741.0,
      "step": 32200
    },
    {
      "entropy": 5.665555047988891,
      "epoch": 3.19620881302104,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000399813956616455,
      "loss": 4.9945,
      "mean_token_accuracy": 0.18340473175048827,
      "num_tokens": 70537819.0,
      "step": 32205
    },
    {
      "entropy": 5.712918615341186,
      "epoch": 3.1967050416832077,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039978447495548497,
      "loss": 5.0525,
      "mean_token_accuracy": 0.18065420538187027,
      "num_tokens": 70548448.0,
      "step": 32210
    },
    {
      "entropy": 5.699454259872437,
      "epoch": 3.197201270345375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039975499020012886,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18925203382968903,
      "num_tokens": 70559997.0,
      "step": 32215
    },
    {
      "entropy": 5.610911846160889,
      "epoch": 3.1976974990075426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039972550235111776,
      "loss": 4.9324,
      "mean_token_accuracy": 0.18652448654174805,
      "num_tokens": 70571312.0,
      "step": 32220
    },
    {
      "entropy": 5.61423544883728,
      "epoch": 3.19819372766971,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000399696011409183,
      "loss": 4.9225,
      "mean_token_accuracy": 0.1956249937415123,
      "num_tokens": 70581806.0,
      "step": 32225
    },
    {
      "entropy": 5.639820909500122,
      "epoch": 3.1986899563318776,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039966651737505585,
      "loss": 4.9659,
      "mean_token_accuracy": 0.1907321557402611,
      "num_tokens": 70592984.0,
      "step": 32230
    },
    {
      "entropy": 5.689223051071167,
      "epoch": 3.199186184994045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003996370202494677,
      "loss": 5.0685,
      "mean_token_accuracy": 0.18300080448389053,
      "num_tokens": 70603945.0,
      "step": 32235
    },
    {
      "entropy": 5.66500678062439,
      "epoch": 3.199682413656213,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039960752003315,
      "loss": 5.016,
      "mean_token_accuracy": 0.18349252343177797,
      "num_tokens": 70614358.0,
      "step": 32240
    },
    {
      "entropy": 5.6492218494415285,
      "epoch": 3.2001786423183805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003995780167268343,
      "loss": 4.8767,
      "mean_token_accuracy": 0.1907762423157692,
      "num_tokens": 70625124.0,
      "step": 32245
    },
    {
      "entropy": 5.623841381072998,
      "epoch": 3.200674870980548,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003995485103312523,
      "loss": 4.9654,
      "mean_token_accuracy": 0.18887520879507064,
      "num_tokens": 70634832.0,
      "step": 32250
    },
    {
      "entropy": 5.666232538223267,
      "epoch": 3.2011710996427154,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039951900084713555,
      "loss": 5.028,
      "mean_token_accuracy": 0.18036964237689973,
      "num_tokens": 70646602.0,
      "step": 32255
    },
    {
      "entropy": 5.610412979125977,
      "epoch": 3.201667328304883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003994894882752159,
      "loss": 4.8869,
      "mean_token_accuracy": 0.1952275887131691,
      "num_tokens": 70656092.0,
      "step": 32260
    },
    {
      "entropy": 5.583690643310547,
      "epoch": 3.2021635569670504,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039945997261622526,
      "loss": 4.8783,
      "mean_token_accuracy": 0.18943430930376054,
      "num_tokens": 70668735.0,
      "step": 32265
    },
    {
      "entropy": 5.668617153167725,
      "epoch": 3.202659785629218,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003994304538708954,
      "loss": 4.998,
      "mean_token_accuracy": 0.1887151703238487,
      "num_tokens": 70680981.0,
      "step": 32270
    },
    {
      "entropy": 5.603886270523072,
      "epoch": 3.2031560142913853,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003994009320399584,
      "loss": 4.9744,
      "mean_token_accuracy": 0.18597880154848098,
      "num_tokens": 70691721.0,
      "step": 32275
    },
    {
      "entropy": 5.707055139541626,
      "epoch": 3.203652242953553,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039937140712414644,
      "loss": 5.0486,
      "mean_token_accuracy": 0.1838463082909584,
      "num_tokens": 70702405.0,
      "step": 32280
    },
    {
      "entropy": 5.6132182598114015,
      "epoch": 3.2041484716157207,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003993418791241915,
      "loss": 4.9186,
      "mean_token_accuracy": 0.19564815908670424,
      "num_tokens": 70713563.0,
      "step": 32285
    },
    {
      "entropy": 5.707144689559937,
      "epoch": 3.204644700277888,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003993123480408259,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17502713501453399,
      "num_tokens": 70726114.0,
      "step": 32290
    },
    {
      "entropy": 5.599138355255127,
      "epoch": 3.2051409289400556,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039928281387478207,
      "loss": 4.9675,
      "mean_token_accuracy": 0.1912006214261055,
      "num_tokens": 70737072.0,
      "step": 32295
    },
    {
      "entropy": 5.628244876861572,
      "epoch": 3.205637157602223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003992532766267921,
      "loss": 4.9723,
      "mean_token_accuracy": 0.18704378455877305,
      "num_tokens": 70747099.0,
      "step": 32300
    },
    {
      "entropy": 5.654053258895874,
      "epoch": 3.2061333862643906,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003992237362975887,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1858128786087036,
      "num_tokens": 70757163.0,
      "step": 32305
    },
    {
      "entropy": 5.616705369949341,
      "epoch": 3.206629614926558,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039919419288790423,
      "loss": 4.9523,
      "mean_token_accuracy": 0.1827022522687912,
      "num_tokens": 70768898.0,
      "step": 32310
    },
    {
      "entropy": 5.670943307876587,
      "epoch": 3.2071258435887255,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039916464639847145,
      "loss": 4.9524,
      "mean_token_accuracy": 0.1877250701189041,
      "num_tokens": 70780208.0,
      "step": 32315
    },
    {
      "entropy": 5.720714139938354,
      "epoch": 3.207622072250893,
      "grad_norm": 1.125,
      "learning_rate": 0.0003991350968300229,
      "loss": 5.1309,
      "mean_token_accuracy": 0.17481465488672257,
      "num_tokens": 70791575.0,
      "step": 32320
    },
    {
      "entropy": 5.672789096832275,
      "epoch": 3.208118300913061,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003991055441832916,
      "loss": 4.9576,
      "mean_token_accuracy": 0.19232695549726486,
      "num_tokens": 70801535.0,
      "step": 32325
    },
    {
      "entropy": 5.6433452606201175,
      "epoch": 3.2086145295752284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039907598845901016,
      "loss": 5.0352,
      "mean_token_accuracy": 0.18444890677928924,
      "num_tokens": 70813979.0,
      "step": 32330
    },
    {
      "entropy": 5.586486864089966,
      "epoch": 3.209110758237396,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003990464296579114,
      "loss": 4.9551,
      "mean_token_accuracy": 0.18591040819883348,
      "num_tokens": 70824598.0,
      "step": 32335
    },
    {
      "entropy": 5.653820991516113,
      "epoch": 3.2096069868995634,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003990168677807287,
      "loss": 5.1121,
      "mean_token_accuracy": 0.1859035387635231,
      "num_tokens": 70836297.0,
      "step": 32340
    },
    {
      "entropy": 5.641550064086914,
      "epoch": 3.210103215561731,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003989873028281948,
      "loss": 4.9882,
      "mean_token_accuracy": 0.17937104254961014,
      "num_tokens": 70846729.0,
      "step": 32345
    },
    {
      "entropy": 5.658050918579102,
      "epoch": 3.2105994442238983,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000398957734801043,
      "loss": 4.9521,
      "mean_token_accuracy": 0.1924817904829979,
      "num_tokens": 70856994.0,
      "step": 32350
    },
    {
      "entropy": 5.594233846664428,
      "epoch": 3.211095672886066,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039892816370000657,
      "loss": 4.9559,
      "mean_token_accuracy": 0.18555313348770142,
      "num_tokens": 70867436.0,
      "step": 32355
    },
    {
      "entropy": 5.6259551525115965,
      "epoch": 3.2115919015482333,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003988985895258186,
      "loss": 4.9989,
      "mean_token_accuracy": 0.19111654460430144,
      "num_tokens": 70878079.0,
      "step": 32360
    },
    {
      "entropy": 5.656444311141968,
      "epoch": 3.212088130210401,
      "grad_norm": 1.125,
      "learning_rate": 0.00039886901227921263,
      "loss": 5.0612,
      "mean_token_accuracy": 0.17842593491077424,
      "num_tokens": 70887722.0,
      "step": 32365
    },
    {
      "entropy": 5.625882291793824,
      "epoch": 3.2125843588725687,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003988394319609221,
      "loss": 4.9164,
      "mean_token_accuracy": 0.19405217468738556,
      "num_tokens": 70899353.0,
      "step": 32370
    },
    {
      "entropy": 5.642453527450561,
      "epoch": 3.213080587534736,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003988098485716804,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18416234254837036,
      "num_tokens": 70910653.0,
      "step": 32375
    },
    {
      "entropy": 5.607425117492676,
      "epoch": 3.2135768161969036,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039878026211222136,
      "loss": 4.991,
      "mean_token_accuracy": 0.18882764577865602,
      "num_tokens": 70921891.0,
      "step": 32380
    },
    {
      "entropy": 5.661868572235107,
      "epoch": 3.214073044859071,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039875067258327853,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18878328949213027,
      "num_tokens": 70932723.0,
      "step": 32385
    },
    {
      "entropy": 5.630725908279419,
      "epoch": 3.2145692735212386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003987210799855856,
      "loss": 4.982,
      "mean_token_accuracy": 0.1877628743648529,
      "num_tokens": 70944033.0,
      "step": 32390
    },
    {
      "entropy": 5.647501134872437,
      "epoch": 3.215065502183406,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039869148431987653,
      "loss": 4.9541,
      "mean_token_accuracy": 0.1877082735300064,
      "num_tokens": 70954867.0,
      "step": 32395
    },
    {
      "entropy": 5.562717247009277,
      "epoch": 3.2155617308455735,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003986618855868852,
      "loss": 4.9506,
      "mean_token_accuracy": 0.18952571004629135,
      "num_tokens": 70965662.0,
      "step": 32400
    },
    {
      "entropy": 5.592810249328613,
      "epoch": 3.216057959507741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003986322837873455,
      "loss": 4.9199,
      "mean_token_accuracy": 0.1827988088130951,
      "num_tokens": 70976093.0,
      "step": 32405
    },
    {
      "entropy": 5.697250556945801,
      "epoch": 3.216554188169909,
      "grad_norm": 1.125,
      "learning_rate": 0.00039860267892199167,
      "loss": 5.0889,
      "mean_token_accuracy": 0.18561056405305862,
      "num_tokens": 70986497.0,
      "step": 32410
    },
    {
      "entropy": 5.648070049285889,
      "epoch": 3.2170504168320764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039857307099155773,
      "loss": 4.9639,
      "mean_token_accuracy": 0.18990270793437958,
      "num_tokens": 70995597.0,
      "step": 32415
    },
    {
      "entropy": 5.68447470664978,
      "epoch": 3.217546645494244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039854345999677777,
      "loss": 4.9911,
      "mean_token_accuracy": 0.19400025755167008,
      "num_tokens": 71006602.0,
      "step": 32420
    },
    {
      "entropy": 5.621898889541626,
      "epoch": 3.2180428741564113,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039851384593838637,
      "loss": 4.9665,
      "mean_token_accuracy": 0.18509878665208818,
      "num_tokens": 71018378.0,
      "step": 32425
    },
    {
      "entropy": 5.659513139724732,
      "epoch": 3.218539102818579,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039848422881711765,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18880328834056853,
      "num_tokens": 71030390.0,
      "step": 32430
    },
    {
      "entropy": 5.605237197875977,
      "epoch": 3.2190353314807463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003984546086337061,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18674559593200685,
      "num_tokens": 71041333.0,
      "step": 32435
    },
    {
      "entropy": 5.624614000320435,
      "epoch": 3.2195315601429138,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039842498538888646,
      "loss": 5.0179,
      "mean_token_accuracy": 0.1887422651052475,
      "num_tokens": 71052346.0,
      "step": 32440
    },
    {
      "entropy": 5.643225860595703,
      "epoch": 3.2200277888050812,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039839535908339297,
      "loss": 4.9868,
      "mean_token_accuracy": 0.18925818502902986,
      "num_tokens": 71062522.0,
      "step": 32445
    },
    {
      "entropy": 5.687436294555664,
      "epoch": 3.2205240174672487,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039836572971796045,
      "loss": 5.1093,
      "mean_token_accuracy": 0.17616080790758132,
      "num_tokens": 71073678.0,
      "step": 32450
    },
    {
      "entropy": 5.59967188835144,
      "epoch": 3.2210202461294166,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003983360972933239,
      "loss": 4.9684,
      "mean_token_accuracy": 0.19189002960920334,
      "num_tokens": 71085579.0,
      "step": 32455
    },
    {
      "entropy": 5.61641674041748,
      "epoch": 3.221516474791584,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003983064618102177,
      "loss": 4.9807,
      "mean_token_accuracy": 0.18705473840236664,
      "num_tokens": 71096477.0,
      "step": 32460
    },
    {
      "entropy": 5.628661632537842,
      "epoch": 3.2220127034537516,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003982768232693769,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18807086795568467,
      "num_tokens": 71106421.0,
      "step": 32465
    },
    {
      "entropy": 5.60895733833313,
      "epoch": 3.222508932115919,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039824718167153666,
      "loss": 4.9554,
      "mean_token_accuracy": 0.1880482017993927,
      "num_tokens": 71116778.0,
      "step": 32470
    },
    {
      "entropy": 5.681633567810058,
      "epoch": 3.2230051607780865,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003982175370174319,
      "loss": 5.0849,
      "mean_token_accuracy": 0.18406376838684083,
      "num_tokens": 71128083.0,
      "step": 32475
    },
    {
      "entropy": 5.617866945266724,
      "epoch": 3.223501389440254,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003981878893077976,
      "loss": 4.8866,
      "mean_token_accuracy": 0.19097762703895568,
      "num_tokens": 71139322.0,
      "step": 32480
    },
    {
      "entropy": 5.644269704818726,
      "epoch": 3.2239976181024215,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039815823854336916,
      "loss": 4.9284,
      "mean_token_accuracy": 0.1910408318042755,
      "num_tokens": 71150117.0,
      "step": 32485
    },
    {
      "entropy": 5.63714017868042,
      "epoch": 3.224493846764589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003981285847248818,
      "loss": 4.9768,
      "mean_token_accuracy": 0.18768416345119476,
      "num_tokens": 71160507.0,
      "step": 32490
    },
    {
      "entropy": 5.5162999629974365,
      "epoch": 3.2249900754267564,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003980989278530709,
      "loss": 4.901,
      "mean_token_accuracy": 0.1849822372198105,
      "num_tokens": 71171409.0,
      "step": 32495
    },
    {
      "entropy": 5.5713128566741945,
      "epoch": 3.2254863040889243,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003980692679286718,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18523908108472825,
      "num_tokens": 71182311.0,
      "step": 32500
    },
    {
      "entropy": 5.6584413051605225,
      "epoch": 3.225982532751092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039803960495241996,
      "loss": 5.0987,
      "mean_token_accuracy": 0.17691277861595153,
      "num_tokens": 71194072.0,
      "step": 32505
    },
    {
      "entropy": 5.754081583023071,
      "epoch": 3.2264787614132593,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003980099389250511,
      "loss": 5.0168,
      "mean_token_accuracy": 0.1804109439253807,
      "num_tokens": 71204609.0,
      "step": 32510
    },
    {
      "entropy": 5.646346712112427,
      "epoch": 3.2269749900754268,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039798026984730084,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19318155497312545,
      "num_tokens": 71216066.0,
      "step": 32515
    },
    {
      "entropy": 5.5469717502594,
      "epoch": 3.2274712187375942,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003979505977199049,
      "loss": 4.8908,
      "mean_token_accuracy": 0.18876945227384567,
      "num_tokens": 71227621.0,
      "step": 32520
    },
    {
      "entropy": 5.662465906143188,
      "epoch": 3.2279674473997617,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039792092254359903,
      "loss": 5.0347,
      "mean_token_accuracy": 0.1859391823410988,
      "num_tokens": 71238057.0,
      "step": 32525
    },
    {
      "entropy": 5.605532836914063,
      "epoch": 3.228463676061929,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039789124431911923,
      "loss": 4.9424,
      "mean_token_accuracy": 0.1824689358472824,
      "num_tokens": 71249423.0,
      "step": 32530
    },
    {
      "entropy": 5.663258504867554,
      "epoch": 3.2289599047240967,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039786156304720126,
      "loss": 4.9746,
      "mean_token_accuracy": 0.19362290501594542,
      "num_tokens": 71260188.0,
      "step": 32535
    },
    {
      "entropy": 5.656825733184815,
      "epoch": 3.2294561333862646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003978318787285814,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18373458683490754,
      "num_tokens": 71272782.0,
      "step": 32540
    },
    {
      "entropy": 5.648364782333374,
      "epoch": 3.229952362048432,
      "grad_norm": 1.125,
      "learning_rate": 0.00039780219136399556,
      "loss": 4.9399,
      "mean_token_accuracy": 0.18810399174690245,
      "num_tokens": 71283560.0,
      "step": 32545
    },
    {
      "entropy": 5.644419527053833,
      "epoch": 3.2304485907105995,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039777250095418,
      "loss": 4.964,
      "mean_token_accuracy": 0.18926470279693602,
      "num_tokens": 71294026.0,
      "step": 32550
    },
    {
      "entropy": 5.547709560394287,
      "epoch": 3.230944819372767,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003977428074998709,
      "loss": 4.9549,
      "mean_token_accuracy": 0.18795726746320723,
      "num_tokens": 71305067.0,
      "step": 32555
    },
    {
      "entropy": 5.659856557846069,
      "epoch": 3.2314410480349345,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039771311100180476,
      "loss": 5.0454,
      "mean_token_accuracy": 0.1830158308148384,
      "num_tokens": 71315306.0,
      "step": 32560
    },
    {
      "entropy": 5.621788358688354,
      "epoch": 3.231937276697102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039768341146071784,
      "loss": 4.9425,
      "mean_token_accuracy": 0.18844232261180877,
      "num_tokens": 71326900.0,
      "step": 32565
    },
    {
      "entropy": 5.621851968765259,
      "epoch": 3.2324335053592694,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003976537088773467,
      "loss": 4.9602,
      "mean_token_accuracy": 0.19345081150531768,
      "num_tokens": 71336719.0,
      "step": 32570
    },
    {
      "entropy": 5.635453176498413,
      "epoch": 3.232929734021437,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039762400325242785,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18746505677700043,
      "num_tokens": 71348287.0,
      "step": 32575
    },
    {
      "entropy": 5.724239683151245,
      "epoch": 3.2334259626836044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039759429458669805,
      "loss": 5.0688,
      "mean_token_accuracy": 0.18508398532867432,
      "num_tokens": 71361472.0,
      "step": 32580
    },
    {
      "entropy": 5.699393272399902,
      "epoch": 3.2339221913457723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039756458288089377,
      "loss": 5.0862,
      "mean_token_accuracy": 0.18051770329475403,
      "num_tokens": 71373781.0,
      "step": 32585
    },
    {
      "entropy": 5.642323160171509,
      "epoch": 3.2344184200079398,
      "grad_norm": 1.234375,
      "learning_rate": 0.000397534868135752,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18711821585893632,
      "num_tokens": 71383519.0,
      "step": 32590
    },
    {
      "entropy": 5.624776077270508,
      "epoch": 3.2349146486701073,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039750515035200947,
      "loss": 4.96,
      "mean_token_accuracy": 0.186074860394001,
      "num_tokens": 71393959.0,
      "step": 32595
    },
    {
      "entropy": 5.623622035980224,
      "epoch": 3.2354108773322747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039747542953040327,
      "loss": 5.042,
      "mean_token_accuracy": 0.18613225370645523,
      "num_tokens": 71405848.0,
      "step": 32600
    },
    {
      "entropy": 5.677010822296142,
      "epoch": 3.235907105994442,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003974457056716703,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18715206980705262,
      "num_tokens": 71418129.0,
      "step": 32605
    },
    {
      "entropy": 5.637772893905639,
      "epoch": 3.2364033346566097,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003974159787765477,
      "loss": 5.0763,
      "mean_token_accuracy": 0.17979386448860168,
      "num_tokens": 71430355.0,
      "step": 32610
    },
    {
      "entropy": 5.729110908508301,
      "epoch": 3.236899563318777,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003973862488457726,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18033471703529358,
      "num_tokens": 71441293.0,
      "step": 32615
    },
    {
      "entropy": 5.636927556991577,
      "epoch": 3.2373957919809446,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003973565158800822,
      "loss": 5.0048,
      "mean_token_accuracy": 0.17910830974578856,
      "num_tokens": 71452092.0,
      "step": 32620
    },
    {
      "entropy": 5.640512990951538,
      "epoch": 3.2378920206431125,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039732677988021394,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18666412979364394,
      "num_tokens": 71462664.0,
      "step": 32625
    },
    {
      "entropy": 5.587694597244263,
      "epoch": 3.23838824930528,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003972970408469051,
      "loss": 4.9095,
      "mean_token_accuracy": 0.18896968215703963,
      "num_tokens": 71472900.0,
      "step": 32630
    },
    {
      "entropy": 5.639098882675171,
      "epoch": 3.2388844779674475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003972672987808932,
      "loss": 5.0237,
      "mean_token_accuracy": 0.18796109706163405,
      "num_tokens": 71484922.0,
      "step": 32635
    },
    {
      "entropy": 5.582996892929077,
      "epoch": 3.239380706629615,
      "grad_norm": 1.125,
      "learning_rate": 0.00039723755368291564,
      "loss": 4.8656,
      "mean_token_accuracy": 0.19934059232473372,
      "num_tokens": 71496951.0,
      "step": 32640
    },
    {
      "entropy": 5.648402070999145,
      "epoch": 3.2398769352917824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003972078055537102,
      "loss": 4.928,
      "mean_token_accuracy": 0.18727451115846633,
      "num_tokens": 71507127.0,
      "step": 32645
    },
    {
      "entropy": 5.666545820236206,
      "epoch": 3.24037316395395,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003971780543940146,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18733966797590257,
      "num_tokens": 71517950.0,
      "step": 32650
    },
    {
      "entropy": 5.603724145889283,
      "epoch": 3.2408693926161174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003971483002045665,
      "loss": 4.8979,
      "mean_token_accuracy": 0.19538461416959763,
      "num_tokens": 71528710.0,
      "step": 32655
    },
    {
      "entropy": 5.591885662078857,
      "epoch": 3.241365621278285,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003971185429861037,
      "loss": 4.9444,
      "mean_token_accuracy": 0.18489827513694762,
      "num_tokens": 71540780.0,
      "step": 32660
    },
    {
      "entropy": 5.562308645248413,
      "epoch": 3.2418618499404523,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039708878273936424,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18689810186624528,
      "num_tokens": 71551487.0,
      "step": 32665
    },
    {
      "entropy": 5.580245637893677,
      "epoch": 3.2423580786026203,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003970590194650859,
      "loss": 4.9265,
      "mean_token_accuracy": 0.18932240307331086,
      "num_tokens": 71561972.0,
      "step": 32670
    },
    {
      "entropy": 5.582523250579834,
      "epoch": 3.2428543072647877,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003970292531640071,
      "loss": 4.9013,
      "mean_token_accuracy": 0.1984691008925438,
      "num_tokens": 71571477.0,
      "step": 32675
    },
    {
      "entropy": 5.561653709411621,
      "epoch": 3.243350535926955,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039699948383686563,
      "loss": 4.933,
      "mean_token_accuracy": 0.190906885266304,
      "num_tokens": 71581941.0,
      "step": 32680
    },
    {
      "entropy": 5.586683225631714,
      "epoch": 3.2438467645891227,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003969697114843998,
      "loss": 4.9272,
      "mean_token_accuracy": 0.19615738093852997,
      "num_tokens": 71593427.0,
      "step": 32685
    },
    {
      "entropy": 5.615177583694458,
      "epoch": 3.24434299325129,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039693993610734807,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18318105936050416,
      "num_tokens": 71604334.0,
      "step": 32690
    },
    {
      "entropy": 5.664994859695435,
      "epoch": 3.2448392219134576,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039691015770644875,
      "loss": 5.0524,
      "mean_token_accuracy": 0.18665938526391984,
      "num_tokens": 71616235.0,
      "step": 32695
    },
    {
      "entropy": 5.700814342498779,
      "epoch": 3.245335450575625,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039688037628244,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18961752951145172,
      "num_tokens": 71627449.0,
      "step": 32700
    },
    {
      "entropy": 5.733562660217285,
      "epoch": 3.2458316792377926,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039685059183606065,
      "loss": 5.0789,
      "mean_token_accuracy": 0.183964404463768,
      "num_tokens": 71639230.0,
      "step": 32705
    },
    {
      "entropy": 5.670098352432251,
      "epoch": 3.2463279078999605,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039682080436804926,
      "loss": 5.1073,
      "mean_token_accuracy": 0.18045780211687087,
      "num_tokens": 71649432.0,
      "step": 32710
    },
    {
      "entropy": 5.602513074874878,
      "epoch": 3.246824136562128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003967910138791442,
      "loss": 4.9623,
      "mean_token_accuracy": 0.19279069304466248,
      "num_tokens": 71660914.0,
      "step": 32715
    },
    {
      "entropy": 5.652623414993286,
      "epoch": 3.2473203652242955,
      "grad_norm": 1.125,
      "learning_rate": 0.00039676122037008473,
      "loss": 4.9442,
      "mean_token_accuracy": 0.18756783306598662,
      "num_tokens": 71671170.0,
      "step": 32720
    },
    {
      "entropy": 5.6357667446136475,
      "epoch": 3.247816593886463,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003967314238416092,
      "loss": 4.9926,
      "mean_token_accuracy": 0.1929889887571335,
      "num_tokens": 71682775.0,
      "step": 32725
    },
    {
      "entropy": 5.578785467147827,
      "epoch": 3.2483128225486304,
      "grad_norm": 1.125,
      "learning_rate": 0.0003967016242944567,
      "loss": 4.9092,
      "mean_token_accuracy": 0.18409010469913484,
      "num_tokens": 71692806.0,
      "step": 32730
    },
    {
      "entropy": 5.616469192504883,
      "epoch": 3.248809051210798,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003966718217293661,
      "loss": 4.9915,
      "mean_token_accuracy": 0.18365826159715654,
      "num_tokens": 71702750.0,
      "step": 32735
    },
    {
      "entropy": 5.646382570266724,
      "epoch": 3.2493052798729654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039664201614707657,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18913742452859877,
      "num_tokens": 71714435.0,
      "step": 32740
    },
    {
      "entropy": 5.708691501617432,
      "epoch": 3.249801508535133,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003966122075483271,
      "loss": 5.0813,
      "mean_token_accuracy": 0.1811624512076378,
      "num_tokens": 71725172.0,
      "step": 32745
    },
    {
      "entropy": 5.669057607650757,
      "epoch": 3.2502977371973003,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039658239593385697,
      "loss": 5.0403,
      "mean_token_accuracy": 0.18772354871034622,
      "num_tokens": 71736337.0,
      "step": 32750
    },
    {
      "entropy": 5.663173627853394,
      "epoch": 3.2507939658594682,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039655258130440543,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18668074160814285,
      "num_tokens": 71747757.0,
      "step": 32755
    },
    {
      "entropy": 5.598687744140625,
      "epoch": 3.2512901945216357,
      "grad_norm": 1.25,
      "learning_rate": 0.0003965227636607117,
      "loss": 4.9421,
      "mean_token_accuracy": 0.19209012687206267,
      "num_tokens": 71757779.0,
      "step": 32760
    },
    {
      "entropy": 5.6172089099884035,
      "epoch": 3.251786423183803,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039649294300351547,
      "loss": 4.9581,
      "mean_token_accuracy": 0.194033682346344,
      "num_tokens": 71769072.0,
      "step": 32765
    },
    {
      "entropy": 5.590090227127075,
      "epoch": 3.2522826518459707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003964631193335559,
      "loss": 4.9613,
      "mean_token_accuracy": 0.1880505934357643,
      "num_tokens": 71781011.0,
      "step": 32770
    },
    {
      "entropy": 5.586590528488159,
      "epoch": 3.252778880508138,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039643329265157275,
      "loss": 4.9225,
      "mean_token_accuracy": 0.190388922393322,
      "num_tokens": 71792868.0,
      "step": 32775
    },
    {
      "entropy": 5.621415996551514,
      "epoch": 3.2532751091703056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039640346295830557,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18410478979349137,
      "num_tokens": 71804047.0,
      "step": 32780
    },
    {
      "entropy": 5.695394277572632,
      "epoch": 3.253771337832473,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003963736302544942,
      "loss": 5.0459,
      "mean_token_accuracy": 0.19170524924993515,
      "num_tokens": 71814437.0,
      "step": 32785
    },
    {
      "entropy": 5.671091938018799,
      "epoch": 3.2542675664946406,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039634379454087814,
      "loss": 5.0235,
      "mean_token_accuracy": 0.18173322081565857,
      "num_tokens": 71824519.0,
      "step": 32790
    },
    {
      "entropy": 5.692321300506592,
      "epoch": 3.2547637951568085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003963139558181977,
      "loss": 5.0591,
      "mean_token_accuracy": 0.18471547961235046,
      "num_tokens": 71835956.0,
      "step": 32795
    },
    {
      "entropy": 5.629510974884033,
      "epoch": 3.255260023818976,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003962841140871924,
      "loss": 4.978,
      "mean_token_accuracy": 0.1856645181775093,
      "num_tokens": 71846952.0,
      "step": 32800
    },
    {
      "entropy": 5.623966026306152,
      "epoch": 3.2557562524811434,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003962542693486024,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19379084259271623,
      "num_tokens": 71858092.0,
      "step": 32805
    },
    {
      "entropy": 5.580258560180664,
      "epoch": 3.256252481143311,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003962244216031679,
      "loss": 4.9728,
      "mean_token_accuracy": 0.18504140973091127,
      "num_tokens": 71870158.0,
      "step": 32810
    },
    {
      "entropy": 5.695288038253784,
      "epoch": 3.2567487098054784,
      "grad_norm": 1.203125,
      "learning_rate": 0.000396194570851629,
      "loss": 5.1098,
      "mean_token_accuracy": 0.17824763804674149,
      "num_tokens": 71880466.0,
      "step": 32815
    },
    {
      "entropy": 5.6925801753997805,
      "epoch": 3.257244938467646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039616471709472573,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18110008239746095,
      "num_tokens": 71890463.0,
      "step": 32820
    },
    {
      "entropy": 5.6539078712463375,
      "epoch": 3.2577411671298133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003961348603331987,
      "loss": 5.0245,
      "mean_token_accuracy": 0.18273636996746062,
      "num_tokens": 71903626.0,
      "step": 32825
    },
    {
      "entropy": 5.601417970657349,
      "epoch": 3.258237395791981,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039610500056778817,
      "loss": 4.9658,
      "mean_token_accuracy": 0.18975120186805725,
      "num_tokens": 71913205.0,
      "step": 32830
    },
    {
      "entropy": 5.677207517623901,
      "epoch": 3.2587336244541483,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003960751377992344,
      "loss": 5.0422,
      "mean_token_accuracy": 0.18306779861450195,
      "num_tokens": 71924524.0,
      "step": 32835
    },
    {
      "entropy": 5.632422828674317,
      "epoch": 3.259229853116316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039604527202827825,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18282397240400314,
      "num_tokens": 71934944.0,
      "step": 32840
    },
    {
      "entropy": 5.604023408889771,
      "epoch": 3.2597260817784837,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039601540325566014,
      "loss": 4.9627,
      "mean_token_accuracy": 0.1865327388048172,
      "num_tokens": 71944845.0,
      "step": 32845
    },
    {
      "entropy": 5.5958672046661375,
      "epoch": 3.260222310440651,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003959855314821208,
      "loss": 4.979,
      "mean_token_accuracy": 0.18362404257059098,
      "num_tokens": 71956370.0,
      "step": 32850
    },
    {
      "entropy": 5.6854499816894535,
      "epoch": 3.2607185391028186,
      "grad_norm": 1.078125,
      "learning_rate": 0.000395955656708401,
      "loss": 5.0018,
      "mean_token_accuracy": 0.19169698357582093,
      "num_tokens": 71967236.0,
      "step": 32855
    },
    {
      "entropy": 5.718086767196655,
      "epoch": 3.261214767764986,
      "grad_norm": 1.25,
      "learning_rate": 0.00039592577893524155,
      "loss": 5.0569,
      "mean_token_accuracy": 0.18383050262928008,
      "num_tokens": 71977241.0,
      "step": 32860
    },
    {
      "entropy": 5.635334634780884,
      "epoch": 3.2617109964271536,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039589589816338335,
      "loss": 5.0584,
      "mean_token_accuracy": 0.18433181047439576,
      "num_tokens": 71988694.0,
      "step": 32865
    },
    {
      "entropy": 5.579209804534912,
      "epoch": 3.262207225089321,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003958660143935674,
      "loss": 4.9123,
      "mean_token_accuracy": 0.19245415627956391,
      "num_tokens": 71999942.0,
      "step": 32870
    },
    {
      "entropy": 5.569094610214234,
      "epoch": 3.2627034537514885,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003958361276265347,
      "loss": 4.9882,
      "mean_token_accuracy": 0.18555572628974915,
      "num_tokens": 72010517.0,
      "step": 32875
    },
    {
      "entropy": 5.6058014869689945,
      "epoch": 3.2631996824136564,
      "grad_norm": 1.125,
      "learning_rate": 0.00039580623786302643,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1881626009941101,
      "num_tokens": 72022389.0,
      "step": 32880
    },
    {
      "entropy": 5.673985815048217,
      "epoch": 3.263695911075824,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039577634510378374,
      "loss": 5.0085,
      "mean_token_accuracy": 0.19019452780485152,
      "num_tokens": 72033521.0,
      "step": 32885
    },
    {
      "entropy": 5.6352582454681395,
      "epoch": 3.2641921397379914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000395746449349548,
      "loss": 4.9445,
      "mean_token_accuracy": 0.18836968392133713,
      "num_tokens": 72044530.0,
      "step": 32890
    },
    {
      "entropy": 5.616096544265747,
      "epoch": 3.264688368400159,
      "grad_norm": 1.125,
      "learning_rate": 0.0003957165506010605,
      "loss": 4.9771,
      "mean_token_accuracy": 0.18270126581192017,
      "num_tokens": 72055476.0,
      "step": 32895
    },
    {
      "entropy": 5.68297643661499,
      "epoch": 3.2651845970623263,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003956866488590626,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1768271267414093,
      "num_tokens": 72065623.0,
      "step": 32900
    },
    {
      "entropy": 5.664883041381836,
      "epoch": 3.265680825724494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003956567441242959,
      "loss": 4.9199,
      "mean_token_accuracy": 0.18982960730791093,
      "num_tokens": 72076396.0,
      "step": 32905
    },
    {
      "entropy": 5.593866062164307,
      "epoch": 3.2661770543866613,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000395626836397502,
      "loss": 4.9673,
      "mean_token_accuracy": 0.1902082622051239,
      "num_tokens": 72087882.0,
      "step": 32910
    },
    {
      "entropy": 5.688396406173706,
      "epoch": 3.2666732830488288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003955969256794224,
      "loss": 5.0151,
      "mean_token_accuracy": 0.18660623878240584,
      "num_tokens": 72099661.0,
      "step": 32915
    },
    {
      "entropy": 5.662267398834229,
      "epoch": 3.2671695117109962,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039556701197079896,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18463416248559952,
      "num_tokens": 72110516.0,
      "step": 32920
    },
    {
      "entropy": 5.637697696685791,
      "epoch": 3.2676657403731637,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039553709527237344,
      "loss": 4.9678,
      "mean_token_accuracy": 0.19022239297628402,
      "num_tokens": 72121854.0,
      "step": 32925
    },
    {
      "entropy": 5.629413270950318,
      "epoch": 3.2681619690353316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003955071755848876,
      "loss": 4.9482,
      "mean_token_accuracy": 0.1895323634147644,
      "num_tokens": 72132780.0,
      "step": 32930
    },
    {
      "entropy": 5.671151256561279,
      "epoch": 3.268658197697499,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039547725290908347,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18655835688114167,
      "num_tokens": 72144187.0,
      "step": 32935
    },
    {
      "entropy": 5.6699076175689695,
      "epoch": 3.2691544263596666,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039544732724570313,
      "loss": 5.0081,
      "mean_token_accuracy": 0.1851753905415535,
      "num_tokens": 72155325.0,
      "step": 32940
    },
    {
      "entropy": 5.717227792739868,
      "epoch": 3.269650655021834,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003954173985954887,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18325461596250534,
      "num_tokens": 72165773.0,
      "step": 32945
    },
    {
      "entropy": 5.668445730209351,
      "epoch": 3.2701468836840015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003953874669591822,
      "loss": 5.0159,
      "mean_token_accuracy": 0.18086966425180434,
      "num_tokens": 72176811.0,
      "step": 32950
    },
    {
      "entropy": 5.6251569271087645,
      "epoch": 3.270643112346169,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039535753233752583,
      "loss": 4.9297,
      "mean_token_accuracy": 0.18912919163703917,
      "num_tokens": 72186657.0,
      "step": 32955
    },
    {
      "entropy": 5.605354642868042,
      "epoch": 3.2711393410083365,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003953275947312621,
      "loss": 4.9871,
      "mean_token_accuracy": 0.1867014229297638,
      "num_tokens": 72196965.0,
      "step": 32960
    },
    {
      "entropy": 5.644795608520508,
      "epoch": 3.2716355696705044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003952976541411332,
      "loss": 4.9706,
      "mean_token_accuracy": 0.1902957409620285,
      "num_tokens": 72207590.0,
      "step": 32965
    },
    {
      "entropy": 5.64307804107666,
      "epoch": 3.272131798332672,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039526771056788186,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1847620815038681,
      "num_tokens": 72219035.0,
      "step": 32970
    },
    {
      "entropy": 5.694469976425171,
      "epoch": 3.2726280269948393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039523776401225043,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18302350491285324,
      "num_tokens": 72229367.0,
      "step": 32975
    },
    {
      "entropy": 5.6318238258361815,
      "epoch": 3.273124255657007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039520781447498145,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17885203361511232,
      "num_tokens": 72241414.0,
      "step": 32980
    },
    {
      "entropy": 5.663037538528442,
      "epoch": 3.2736204843191743,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003951778619568177,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18743847906589509,
      "num_tokens": 72253634.0,
      "step": 32985
    },
    {
      "entropy": 5.687233304977417,
      "epoch": 3.2741167129813418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000395147906458502,
      "loss": 4.9776,
      "mean_token_accuracy": 0.19004595130681992,
      "num_tokens": 72265241.0,
      "step": 32990
    },
    {
      "entropy": 5.654079961776733,
      "epoch": 3.2746129416435092,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039511794798077705,
      "loss": 4.9247,
      "mean_token_accuracy": 0.19295441657304763,
      "num_tokens": 72275978.0,
      "step": 32995
    },
    {
      "entropy": 5.579015731811523,
      "epoch": 3.2751091703056767,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003950879865243858,
      "loss": 4.954,
      "mean_token_accuracy": 0.19632344096899032,
      "num_tokens": 72286676.0,
      "step": 33000
    },
    {
      "epoch": 3.2751091703056767,
      "eval_entropy": 5.421151312775724,
      "eval_loss": 5.118552207946777,
      "eval_mean_token_accuracy": 0.1872995477147625,
      "eval_num_tokens": 72286676.0,
      "eval_runtime": 20.2041,
      "eval_samples_per_second": 1666.443,
      "eval_steps_per_second": 208.324,
      "step": 33000
    },
    {
      "entropy": 5.657505655288697,
      "epoch": 3.275605398967844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039505802209007136,
      "loss": 4.9282,
      "mean_token_accuracy": 0.18212915658950807,
      "num_tokens": 72297939.0,
      "step": 33005
    },
    {
      "entropy": 5.610721921920776,
      "epoch": 3.2761016276300117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003950280546785766,
      "loss": 4.9544,
      "mean_token_accuracy": 0.1829577624797821,
      "num_tokens": 72308092.0,
      "step": 33010
    },
    {
      "entropy": 5.660811424255371,
      "epoch": 3.2765978562921796,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003949980842906447,
      "loss": 5.0072,
      "mean_token_accuracy": 0.18931556791067122,
      "num_tokens": 72318451.0,
      "step": 33015
    },
    {
      "entropy": 5.668350267410278,
      "epoch": 3.277094084954347,
      "grad_norm": 1.0625,
      "learning_rate": 0.000394968110927019,
      "loss": 5.032,
      "mean_token_accuracy": 0.17932357788085937,
      "num_tokens": 72330730.0,
      "step": 33020
    },
    {
      "entropy": 5.568353605270386,
      "epoch": 3.2775903136165145,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039493813458844256,
      "loss": 4.9092,
      "mean_token_accuracy": 0.19906430542469025,
      "num_tokens": 72343201.0,
      "step": 33025
    },
    {
      "entropy": 5.60282711982727,
      "epoch": 3.278086542278682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039490815527565883,
      "loss": 4.9383,
      "mean_token_accuracy": 0.19173253178596497,
      "num_tokens": 72353669.0,
      "step": 33030
    },
    {
      "entropy": 5.6555225372314455,
      "epoch": 3.2785827709408495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039487817298941124,
      "loss": 5.0224,
      "mean_token_accuracy": 0.1757867380976677,
      "num_tokens": 72365451.0,
      "step": 33035
    },
    {
      "entropy": 5.655001449584961,
      "epoch": 3.279078999603017,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003948481877304433,
      "loss": 5.0516,
      "mean_token_accuracy": 0.18618767261505126,
      "num_tokens": 72377152.0,
      "step": 33040
    },
    {
      "entropy": 5.6146632671356205,
      "epoch": 3.2795752282651844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003948181994994985,
      "loss": 4.9629,
      "mean_token_accuracy": 0.18815121799707413,
      "num_tokens": 72387152.0,
      "step": 33045
    },
    {
      "entropy": 5.686898231506348,
      "epoch": 3.2800714569273524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003947882082973206,
      "loss": 5.031,
      "mean_token_accuracy": 0.17960758805274962,
      "num_tokens": 72399479.0,
      "step": 33050
    },
    {
      "entropy": 5.61782169342041,
      "epoch": 3.28056768558952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039475821412465317,
      "loss": 4.8965,
      "mean_token_accuracy": 0.19524168372154235,
      "num_tokens": 72410507.0,
      "step": 33055
    },
    {
      "entropy": 5.664690732955933,
      "epoch": 3.2810639142516873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039472821698224023,
      "loss": 5.0652,
      "mean_token_accuracy": 0.18344541639089584,
      "num_tokens": 72422951.0,
      "step": 33060
    },
    {
      "entropy": 5.6400954246521,
      "epoch": 3.281560142913855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003946982168708254,
      "loss": 5.0065,
      "mean_token_accuracy": 0.18350086361169815,
      "num_tokens": 72434181.0,
      "step": 33065
    },
    {
      "entropy": 5.690362453460693,
      "epoch": 3.2820563715760223,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003946682137911527,
      "loss": 5.045,
      "mean_token_accuracy": 0.18347444832324983,
      "num_tokens": 72444709.0,
      "step": 33070
    },
    {
      "entropy": 5.619188928604126,
      "epoch": 3.2825526002381897,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039463820774396623,
      "loss": 4.9528,
      "mean_token_accuracy": 0.18085064888000488,
      "num_tokens": 72455729.0,
      "step": 33075
    },
    {
      "entropy": 5.601679563522339,
      "epoch": 3.283048828900357,
      "grad_norm": 1.125,
      "learning_rate": 0.00039460819873000995,
      "loss": 4.8938,
      "mean_token_accuracy": 0.20063570886850357,
      "num_tokens": 72466724.0,
      "step": 33080
    },
    {
      "entropy": 5.6607362747192385,
      "epoch": 3.2835450575625247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039457818675002807,
      "loss": 4.9753,
      "mean_token_accuracy": 0.1901071175932884,
      "num_tokens": 72477267.0,
      "step": 33085
    },
    {
      "entropy": 5.678512048721314,
      "epoch": 3.284041286224692,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003945481718047648,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18176218569278718,
      "num_tokens": 72488291.0,
      "step": 33090
    },
    {
      "entropy": 5.754631614685058,
      "epoch": 3.2845375148868596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003945181538949646,
      "loss": 5.0752,
      "mean_token_accuracy": 0.17900153398513793,
      "num_tokens": 72499898.0,
      "step": 33095
    },
    {
      "entropy": 5.671319818496704,
      "epoch": 3.2850337435490276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003944881330213716,
      "loss": 5.0012,
      "mean_token_accuracy": 0.19022804349660874,
      "num_tokens": 72512204.0,
      "step": 33100
    },
    {
      "entropy": 5.628046560287475,
      "epoch": 3.285529972211195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003944581091847304,
      "loss": 4.9469,
      "mean_token_accuracy": 0.18721635341644288,
      "num_tokens": 72523319.0,
      "step": 33105
    },
    {
      "entropy": 5.590186691284179,
      "epoch": 3.2860262008733625,
      "grad_norm": 1.125,
      "learning_rate": 0.00039442808238578553,
      "loss": 4.9604,
      "mean_token_accuracy": 0.19117907881736756,
      "num_tokens": 72534107.0,
      "step": 33110
    },
    {
      "entropy": 5.630396890640259,
      "epoch": 3.28652242953553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039439805262528153,
      "loss": 4.9848,
      "mean_token_accuracy": 0.1843181297183037,
      "num_tokens": 72544768.0,
      "step": 33115
    },
    {
      "entropy": 5.596683502197266,
      "epoch": 3.2870186581976975,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003943680199039631,
      "loss": 4.9181,
      "mean_token_accuracy": 0.19279842525720597,
      "num_tokens": 72555419.0,
      "step": 33120
    },
    {
      "entropy": 5.610516738891602,
      "epoch": 3.287514886859865,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039433798422257505,
      "loss": 4.9261,
      "mean_token_accuracy": 0.18830245584249497,
      "num_tokens": 72566423.0,
      "step": 33125
    },
    {
      "entropy": 5.6210860252380375,
      "epoch": 3.2880111155220324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003943079455818621,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18815473467111588,
      "num_tokens": 72577591.0,
      "step": 33130
    },
    {
      "entropy": 5.660691261291504,
      "epoch": 3.2885073441842003,
      "grad_norm": 1.25,
      "learning_rate": 0.00039427790398256925,
      "loss": 4.8957,
      "mean_token_accuracy": 0.19340712428092957,
      "num_tokens": 72587633.0,
      "step": 33135
    },
    {
      "entropy": 5.669234275817871,
      "epoch": 3.289003572846368,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039424785942544126,
      "loss": 4.9905,
      "mean_token_accuracy": 0.18750974386930466,
      "num_tokens": 72597850.0,
      "step": 33140
    },
    {
      "entropy": 5.59806981086731,
      "epoch": 3.2894998015085353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003942178119112234,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18810254484415054,
      "num_tokens": 72608988.0,
      "step": 33145
    },
    {
      "entropy": 5.642009115219116,
      "epoch": 3.2899960301707027,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003941877614406607,
      "loss": 4.962,
      "mean_token_accuracy": 0.18939609825611115,
      "num_tokens": 72620438.0,
      "step": 33150
    },
    {
      "entropy": 5.6398646354675295,
      "epoch": 3.29049225883287,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003941577080144984,
      "loss": 4.9777,
      "mean_token_accuracy": 0.19096466302871704,
      "num_tokens": 72630584.0,
      "step": 33155
    },
    {
      "entropy": 5.589508867263794,
      "epoch": 3.2909884874950377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003941276516334816,
      "loss": 4.9261,
      "mean_token_accuracy": 0.19118501096963883,
      "num_tokens": 72640631.0,
      "step": 33160
    },
    {
      "entropy": 5.694455337524414,
      "epoch": 3.291484716157205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039409759229835586,
      "loss": 5.0378,
      "mean_token_accuracy": 0.18910957127809525,
      "num_tokens": 72653021.0,
      "step": 33165
    },
    {
      "entropy": 5.601432943344117,
      "epoch": 3.2919809448193726,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003940675300098663,
      "loss": 4.983,
      "mean_token_accuracy": 0.18757562190294266,
      "num_tokens": 72664405.0,
      "step": 33170
    },
    {
      "entropy": 5.611312770843506,
      "epoch": 3.29247717348154,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003940374647687586,
      "loss": 4.9608,
      "mean_token_accuracy": 0.1908017009496689,
      "num_tokens": 72676040.0,
      "step": 33175
    },
    {
      "entropy": 5.570847988128662,
      "epoch": 3.2929734021437076,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003940073965757784,
      "loss": 4.9184,
      "mean_token_accuracy": 0.19160217940807342,
      "num_tokens": 72686472.0,
      "step": 33180
    },
    {
      "entropy": 5.6096240997314455,
      "epoch": 3.2934696308058755,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003939773254316711,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18953447490930558,
      "num_tokens": 72698228.0,
      "step": 33185
    },
    {
      "entropy": 5.624305152893067,
      "epoch": 3.293965859468043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003939472513371826,
      "loss": 4.9358,
      "mean_token_accuracy": 0.19072733223438262,
      "num_tokens": 72708859.0,
      "step": 33190
    },
    {
      "entropy": 5.677036094665527,
      "epoch": 3.2944620881302105,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003939171742930584,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18709117770195008,
      "num_tokens": 72719369.0,
      "step": 33195
    },
    {
      "entropy": 5.658312082290649,
      "epoch": 3.294958316792378,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039388709430004465,
      "loss": 5.023,
      "mean_token_accuracy": 0.18808368295431138,
      "num_tokens": 72729876.0,
      "step": 33200
    },
    {
      "entropy": 5.572378301620484,
      "epoch": 3.2954545454545454,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039385701135888715,
      "loss": 4.9662,
      "mean_token_accuracy": 0.19205163568258285,
      "num_tokens": 72741152.0,
      "step": 33205
    },
    {
      "entropy": 5.62379789352417,
      "epoch": 3.295950774116713,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003938269254703318,
      "loss": 4.9788,
      "mean_token_accuracy": 0.1908908799290657,
      "num_tokens": 72751317.0,
      "step": 33210
    },
    {
      "entropy": 5.631848573684692,
      "epoch": 3.2964470027788804,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039379683663512475,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18327351957559584,
      "num_tokens": 72761570.0,
      "step": 33215
    },
    {
      "entropy": 5.557180404663086,
      "epoch": 3.2969432314410483,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003937667448540123,
      "loss": 4.8547,
      "mean_token_accuracy": 0.18797520995140077,
      "num_tokens": 72771361.0,
      "step": 33220
    },
    {
      "entropy": 5.678770875930786,
      "epoch": 3.2974394601032158,
      "grad_norm": 1.125,
      "learning_rate": 0.00039373665012774027,
      "loss": 5.0351,
      "mean_token_accuracy": 0.18418409079313278,
      "num_tokens": 72781639.0,
      "step": 33225
    },
    {
      "entropy": 5.625521802902222,
      "epoch": 3.2979356887653832,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039370655245705527,
      "loss": 4.9658,
      "mean_token_accuracy": 0.195190691947937,
      "num_tokens": 72791794.0,
      "step": 33230
    },
    {
      "entropy": 5.608374214172363,
      "epoch": 3.2984319174275507,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039367645184270364,
      "loss": 4.9488,
      "mean_token_accuracy": 0.18708093762397765,
      "num_tokens": 72802000.0,
      "step": 33235
    },
    {
      "entropy": 5.61562213897705,
      "epoch": 3.298928146089718,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003936463482854317,
      "loss": 4.9378,
      "mean_token_accuracy": 0.191559037566185,
      "num_tokens": 72813845.0,
      "step": 33240
    },
    {
      "entropy": 5.670791673660278,
      "epoch": 3.2994243747518857,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003936162417859858,
      "loss": 5.0169,
      "mean_token_accuracy": 0.18532804548740386,
      "num_tokens": 72825192.0,
      "step": 33245
    },
    {
      "entropy": 5.713557863235474,
      "epoch": 3.299920603414053,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039358613234511286,
      "loss": 5.0206,
      "mean_token_accuracy": 0.19125026762485503,
      "num_tokens": 72836102.0,
      "step": 33250
    },
    {
      "entropy": 5.622823190689087,
      "epoch": 3.3004168320762206,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039355601996355944,
      "loss": 5.0115,
      "mean_token_accuracy": 0.1793859198689461,
      "num_tokens": 72846564.0,
      "step": 33255
    },
    {
      "entropy": 5.678429317474365,
      "epoch": 3.300913060738388,
      "grad_norm": 1.171875,
      "learning_rate": 0.000393525904642072,
      "loss": 5.0324,
      "mean_token_accuracy": 0.18918611109256744,
      "num_tokens": 72857271.0,
      "step": 33260
    },
    {
      "entropy": 5.6426843166351315,
      "epoch": 3.3014092894005556,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039349578638139766,
      "loss": 5.0299,
      "mean_token_accuracy": 0.18043742924928666,
      "num_tokens": 72869314.0,
      "step": 33265
    },
    {
      "entropy": 5.69200325012207,
      "epoch": 3.3019055180627235,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003934656651822832,
      "loss": 5.0553,
      "mean_token_accuracy": 0.18990725725889207,
      "num_tokens": 72880531.0,
      "step": 33270
    },
    {
      "entropy": 5.666029739379883,
      "epoch": 3.302401746724891,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003934355410454754,
      "loss": 4.9744,
      "mean_token_accuracy": 0.18718209862709045,
      "num_tokens": 72889948.0,
      "step": 33275
    },
    {
      "entropy": 5.634522485733032,
      "epoch": 3.3028979753870584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039340541397172137,
      "loss": 5.0271,
      "mean_token_accuracy": 0.18402965366840363,
      "num_tokens": 72901249.0,
      "step": 33280
    },
    {
      "entropy": 5.6198952198028564,
      "epoch": 3.303394204049226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003933752839617682,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18270187377929686,
      "num_tokens": 72912285.0,
      "step": 33285
    },
    {
      "entropy": 5.668452596664428,
      "epoch": 3.3038904327113934,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003933451510163632,
      "loss": 4.9637,
      "mean_token_accuracy": 0.19451491832733153,
      "num_tokens": 72922887.0,
      "step": 33290
    },
    {
      "entropy": 5.6434544086456295,
      "epoch": 3.304386661373561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003933150151362534,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1797687068581581,
      "num_tokens": 72933597.0,
      "step": 33295
    },
    {
      "entropy": 5.65287070274353,
      "epoch": 3.3048828900357283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039328487632218613,
      "loss": 4.9651,
      "mean_token_accuracy": 0.18802445530891418,
      "num_tokens": 72943813.0,
      "step": 33300
    },
    {
      "entropy": 5.7058624744415285,
      "epoch": 3.305379118697896,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003932547345749088,
      "loss": 4.9939,
      "mean_token_accuracy": 0.18433925807476043,
      "num_tokens": 72954574.0,
      "step": 33305
    },
    {
      "entropy": 5.664068269729614,
      "epoch": 3.3058753473600637,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039322458989516897,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1914246201515198,
      "num_tokens": 72964049.0,
      "step": 33310
    },
    {
      "entropy": 5.714265584945679,
      "epoch": 3.306371576022231,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039319444228371397,
      "loss": 5.0115,
      "mean_token_accuracy": 0.1825781434774399,
      "num_tokens": 72974548.0,
      "step": 33315
    },
    {
      "entropy": 5.646406936645508,
      "epoch": 3.3068678046843987,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003931642917412916,
      "loss": 4.988,
      "mean_token_accuracy": 0.18627510815858841,
      "num_tokens": 72985300.0,
      "step": 33320
    },
    {
      "entropy": 5.691929340362549,
      "epoch": 3.307364033346566,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003931341382686493,
      "loss": 4.9936,
      "mean_token_accuracy": 0.18619215190410615,
      "num_tokens": 72996896.0,
      "step": 33325
    },
    {
      "entropy": 5.688660383224487,
      "epoch": 3.3078602620087336,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039310398186653497,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1790834426879883,
      "num_tokens": 73007224.0,
      "step": 33330
    },
    {
      "entropy": 5.6588245868682865,
      "epoch": 3.308356490670901,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039307382253569637,
      "loss": 5.0039,
      "mean_token_accuracy": 0.18629260212183,
      "num_tokens": 73017051.0,
      "step": 33335
    },
    {
      "entropy": 5.672658681869507,
      "epoch": 3.3088527193330686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003930436602768814,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18592535257339476,
      "num_tokens": 73028492.0,
      "step": 33340
    },
    {
      "entropy": 5.646051263809204,
      "epoch": 3.309348947995236,
      "grad_norm": 1.09375,
      "learning_rate": 0.000393013495090838,
      "loss": 4.9751,
      "mean_token_accuracy": 0.19489786624908448,
      "num_tokens": 73040888.0,
      "step": 33345
    },
    {
      "entropy": 5.603049087524414,
      "epoch": 3.3098451766574035,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003929833269783143,
      "loss": 4.9464,
      "mean_token_accuracy": 0.192742820084095,
      "num_tokens": 73050305.0,
      "step": 33350
    },
    {
      "entropy": 5.6503379344940186,
      "epoch": 3.3103414053195714,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003929531559400582,
      "loss": 4.9797,
      "mean_token_accuracy": 0.18510229438543319,
      "num_tokens": 73062333.0,
      "step": 33355
    },
    {
      "entropy": 5.710484409332276,
      "epoch": 3.310837633981739,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039292298197681803,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18795848339796067,
      "num_tokens": 73073735.0,
      "step": 33360
    },
    {
      "entropy": 5.692732095718384,
      "epoch": 3.3113338626439064,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039289280508934203,
      "loss": 4.9367,
      "mean_token_accuracy": 0.18878621160984038,
      "num_tokens": 73083678.0,
      "step": 33365
    },
    {
      "entropy": 5.650693416595459,
      "epoch": 3.311830091306074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039286262527837856,
      "loss": 4.9939,
      "mean_token_accuracy": 0.1849561884999275,
      "num_tokens": 73094943.0,
      "step": 33370
    },
    {
      "entropy": 5.587219953536987,
      "epoch": 3.3123263199682413,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003928324425446759,
      "loss": 4.9257,
      "mean_token_accuracy": 0.19329022169113158,
      "num_tokens": 73105082.0,
      "step": 33375
    },
    {
      "entropy": 5.594746828079224,
      "epoch": 3.312822548630409,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039280225688898256,
      "loss": 4.9848,
      "mean_token_accuracy": 0.1931580275297165,
      "num_tokens": 73118324.0,
      "step": 33380
    },
    {
      "entropy": 5.630127239227295,
      "epoch": 3.3133187772925763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039277206831204715,
      "loss": 4.972,
      "mean_token_accuracy": 0.18406714200973512,
      "num_tokens": 73130099.0,
      "step": 33385
    },
    {
      "entropy": 5.721219825744629,
      "epoch": 3.3138150059547438,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003927418768146182,
      "loss": 5.0747,
      "mean_token_accuracy": 0.18478987812995912,
      "num_tokens": 73141475.0,
      "step": 33390
    },
    {
      "entropy": 5.619292449951172,
      "epoch": 3.3143112346169117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039271168239744434,
      "loss": 4.9258,
      "mean_token_accuracy": 0.1911814734339714,
      "num_tokens": 73153028.0,
      "step": 33395
    },
    {
      "entropy": 5.594281625747681,
      "epoch": 3.314807463279079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003926814850612745,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18311475217342377,
      "num_tokens": 73164267.0,
      "step": 33400
    },
    {
      "entropy": 5.654182434082031,
      "epoch": 3.3153036919412466,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003926512848068573,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18580484986305237,
      "num_tokens": 73176130.0,
      "step": 33405
    },
    {
      "entropy": 5.630902194976807,
      "epoch": 3.315799920603414,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003926210816349418,
      "loss": 4.9443,
      "mean_token_accuracy": 0.19973783791065217,
      "num_tokens": 73186064.0,
      "step": 33410
    },
    {
      "entropy": 5.613414716720581,
      "epoch": 3.3162961492655816,
      "grad_norm": 1.21875,
      "learning_rate": 0.000392590875546277,
      "loss": 4.8704,
      "mean_token_accuracy": 0.2000698357820511,
      "num_tokens": 73196194.0,
      "step": 33415
    },
    {
      "entropy": 5.773526191711426,
      "epoch": 3.316792377927749,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039256066654161175,
      "loss": 5.0902,
      "mean_token_accuracy": 0.18679990470409394,
      "num_tokens": 73207815.0,
      "step": 33420
    },
    {
      "entropy": 5.629611492156982,
      "epoch": 3.3172886065899165,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003925304546216954,
      "loss": 4.9772,
      "mean_token_accuracy": 0.19085075110197067,
      "num_tokens": 73219699.0,
      "step": 33425
    },
    {
      "entropy": 5.566459846496582,
      "epoch": 3.317784835252084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000392500239787277,
      "loss": 4.9393,
      "mean_token_accuracy": 0.18457242995500564,
      "num_tokens": 73230905.0,
      "step": 33430
    },
    {
      "entropy": 5.637936735153199,
      "epoch": 3.3182810639142515,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003924700220391058,
      "loss": 5.0452,
      "mean_token_accuracy": 0.18737471252679824,
      "num_tokens": 73240937.0,
      "step": 33435
    },
    {
      "entropy": 5.652959394454956,
      "epoch": 3.3187772925764194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003924398013779312,
      "loss": 5.0013,
      "mean_token_accuracy": 0.19021625369787215,
      "num_tokens": 73251599.0,
      "step": 33440
    },
    {
      "entropy": 5.641945505142212,
      "epoch": 3.319273521238587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003924095778045026,
      "loss": 4.9645,
      "mean_token_accuracy": 0.1935942992568016,
      "num_tokens": 73262553.0,
      "step": 33445
    },
    {
      "entropy": 5.627207040786743,
      "epoch": 3.3197697499007544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003923793513195695,
      "loss": 4.9499,
      "mean_token_accuracy": 0.19292019307613373,
      "num_tokens": 73272718.0,
      "step": 33450
    },
    {
      "entropy": 5.638853931427002,
      "epoch": 3.320265978562922,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039234912192388137,
      "loss": 4.9359,
      "mean_token_accuracy": 0.20040459632873536,
      "num_tokens": 73283374.0,
      "step": 33455
    },
    {
      "entropy": 5.60073561668396,
      "epoch": 3.3207622072250893,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003923188896181879,
      "loss": 4.9185,
      "mean_token_accuracy": 0.1888345181941986,
      "num_tokens": 73293941.0,
      "step": 33460
    },
    {
      "entropy": 5.676907873153686,
      "epoch": 3.3212584358872568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003922886544032388,
      "loss": 4.9848,
      "mean_token_accuracy": 0.1871674582362175,
      "num_tokens": 73307005.0,
      "step": 33465
    },
    {
      "entropy": 5.645968580245972,
      "epoch": 3.3217546645494243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039225841627978374,
      "loss": 5.0048,
      "mean_token_accuracy": 0.1895370215177536,
      "num_tokens": 73318234.0,
      "step": 33470
    },
    {
      "entropy": 5.6270664691925045,
      "epoch": 3.3222508932115917,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003922281752485727,
      "loss": 5.0117,
      "mean_token_accuracy": 0.18191763162612914,
      "num_tokens": 73329239.0,
      "step": 33475
    },
    {
      "entropy": 5.665296745300293,
      "epoch": 3.3227471218737596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039219793131035544,
      "loss": 4.951,
      "mean_token_accuracy": 0.195887990295887,
      "num_tokens": 73339136.0,
      "step": 33480
    },
    {
      "entropy": 5.695586061477661,
      "epoch": 3.323243350535927,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003921676844658821,
      "loss": 5.0694,
      "mean_token_accuracy": 0.18058290481567382,
      "num_tokens": 73351279.0,
      "step": 33485
    },
    {
      "entropy": 5.661474609375,
      "epoch": 3.3237395791980946,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003921374347159026,
      "loss": 4.9922,
      "mean_token_accuracy": 0.18810681700706483,
      "num_tokens": 73362292.0,
      "step": 33490
    },
    {
      "entropy": 5.71282114982605,
      "epoch": 3.324235807860262,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003921071820611671,
      "loss": 5.0076,
      "mean_token_accuracy": 0.18082569390535355,
      "num_tokens": 73372370.0,
      "step": 33495
    },
    {
      "entropy": 5.653966093063355,
      "epoch": 3.3247320365224295,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003920769265024259,
      "loss": 4.9342,
      "mean_token_accuracy": 0.18882776647806168,
      "num_tokens": 73382783.0,
      "step": 33500
    },
    {
      "entropy": 5.654179334640503,
      "epoch": 3.325228265184597,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039204666804042913,
      "loss": 5.0133,
      "mean_token_accuracy": 0.1887337237596512,
      "num_tokens": 73394782.0,
      "step": 33505
    },
    {
      "entropy": 5.65034589767456,
      "epoch": 3.3257244938467645,
      "grad_norm": 1.125,
      "learning_rate": 0.0003920164066759273,
      "loss": 4.9476,
      "mean_token_accuracy": 0.19077588468790055,
      "num_tokens": 73405163.0,
      "step": 33510
    },
    {
      "entropy": 5.664493989944458,
      "epoch": 3.326220722508932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003919861424096707,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18901480585336686,
      "num_tokens": 73416554.0,
      "step": 33515
    },
    {
      "entropy": 5.609721755981445,
      "epoch": 3.3267169511710994,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003919558752424098,
      "loss": 4.9284,
      "mean_token_accuracy": 0.19494493901729584,
      "num_tokens": 73426818.0,
      "step": 33520
    },
    {
      "entropy": 5.586587858200073,
      "epoch": 3.327213179833267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003919256051748952,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18964338302612305,
      "num_tokens": 73438316.0,
      "step": 33525
    },
    {
      "entropy": 5.58681092262268,
      "epoch": 3.327709408495435,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039189533220787765,
      "loss": 4.9212,
      "mean_token_accuracy": 0.18978943675756454,
      "num_tokens": 73448799.0,
      "step": 33530
    },
    {
      "entropy": 5.63323016166687,
      "epoch": 3.3282056371576023,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003918650563421076,
      "loss": 4.9463,
      "mean_token_accuracy": 0.18965350836515427,
      "num_tokens": 73459959.0,
      "step": 33535
    },
    {
      "entropy": 5.628065729141236,
      "epoch": 3.32870186581977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000391834777578336,
      "loss": 4.9416,
      "mean_token_accuracy": 0.19083720296621323,
      "num_tokens": 73470537.0,
      "step": 33540
    },
    {
      "entropy": 5.649710607528687,
      "epoch": 3.3291980944819373,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003918044959173138,
      "loss": 5.0417,
      "mean_token_accuracy": 0.17898984551429747,
      "num_tokens": 73480602.0,
      "step": 33545
    },
    {
      "entropy": 5.733684825897217,
      "epoch": 3.3296943231441047,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003917742113597918,
      "loss": 5.0273,
      "mean_token_accuracy": 0.19204406142234803,
      "num_tokens": 73493128.0,
      "step": 33550
    },
    {
      "entropy": 5.640534448623657,
      "epoch": 3.330190551806272,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003917439239065208,
      "loss": 4.9472,
      "mean_token_accuracy": 0.18954356014728546,
      "num_tokens": 73503549.0,
      "step": 33555
    },
    {
      "entropy": 5.6212077140808105,
      "epoch": 3.3306867804684397,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039171363355825216,
      "loss": 4.9336,
      "mean_token_accuracy": 0.19285167008638382,
      "num_tokens": 73513850.0,
      "step": 33560
    },
    {
      "entropy": 5.684678316116333,
      "epoch": 3.3311830091306076,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039168334031573695,
      "loss": 4.9858,
      "mean_token_accuracy": 0.19109046310186387,
      "num_tokens": 73524400.0,
      "step": 33565
    },
    {
      "entropy": 5.706504678726196,
      "epoch": 3.331679237792775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003916530441797262,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18780893683433533,
      "num_tokens": 73535004.0,
      "step": 33570
    },
    {
      "entropy": 5.700225019454956,
      "epoch": 3.3321754664549426,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039162274515097147,
      "loss": 5.0687,
      "mean_token_accuracy": 0.18212442398071288,
      "num_tokens": 73546058.0,
      "step": 33575
    },
    {
      "entropy": 5.651583433151245,
      "epoch": 3.33267169511711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003915924432302239,
      "loss": 4.9706,
      "mean_token_accuracy": 0.18613498359918595,
      "num_tokens": 73557553.0,
      "step": 33580
    },
    {
      "entropy": 5.650988960266114,
      "epoch": 3.3331679237792775,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003915621384182349,
      "loss": 5.0103,
      "mean_token_accuracy": 0.19615473300218583,
      "num_tokens": 73569074.0,
      "step": 33585
    },
    {
      "entropy": 5.600303840637207,
      "epoch": 3.333664152441445,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003915318307157561,
      "loss": 4.8852,
      "mean_token_accuracy": 0.1981668069958687,
      "num_tokens": 73579354.0,
      "step": 33590
    },
    {
      "entropy": 5.571279525756836,
      "epoch": 3.3341603811036125,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039150152012353896,
      "loss": 4.9684,
      "mean_token_accuracy": 0.1901121884584427,
      "num_tokens": 73590380.0,
      "step": 33595
    },
    {
      "entropy": 5.6300976276397705,
      "epoch": 3.33465660976578,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039147120664233525,
      "loss": 5.0285,
      "mean_token_accuracy": 0.19025148302316666,
      "num_tokens": 73602607.0,
      "step": 33600
    },
    {
      "entropy": 5.616990423202514,
      "epoch": 3.3351528384279474,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039144089027289653,
      "loss": 4.929,
      "mean_token_accuracy": 0.19296140670776368,
      "num_tokens": 73612972.0,
      "step": 33605
    },
    {
      "entropy": 5.703885841369629,
      "epoch": 3.335649067090115,
      "grad_norm": 1.125,
      "learning_rate": 0.00039141057101597465,
      "loss": 5.0518,
      "mean_token_accuracy": 0.18669265508651733,
      "num_tokens": 73625267.0,
      "step": 33610
    },
    {
      "entropy": 5.65266375541687,
      "epoch": 3.336145295752283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003913802488723215,
      "loss": 5.0279,
      "mean_token_accuracy": 0.18362773656845094,
      "num_tokens": 73637554.0,
      "step": 33615
    },
    {
      "entropy": 5.638762092590332,
      "epoch": 3.3366415244144503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003913499238426889,
      "loss": 4.9874,
      "mean_token_accuracy": 0.19278548061847686,
      "num_tokens": 73648580.0,
      "step": 33620
    },
    {
      "entropy": 5.661049175262451,
      "epoch": 3.3371377530766178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039131959592782884,
      "loss": 5.0002,
      "mean_token_accuracy": 0.18959093987941741,
      "num_tokens": 73659066.0,
      "step": 33625
    },
    {
      "entropy": 5.6444220542907715,
      "epoch": 3.3376339817387852,
      "grad_norm": 1.125,
      "learning_rate": 0.00039128926512849356,
      "loss": 4.9875,
      "mean_token_accuracy": 0.18443348109722138,
      "num_tokens": 73671221.0,
      "step": 33630
    },
    {
      "entropy": 5.6299604892730715,
      "epoch": 3.3381302104009527,
      "grad_norm": 1.1875,
      "learning_rate": 0.000391258931445435,
      "loss": 4.9318,
      "mean_token_accuracy": 0.1883592963218689,
      "num_tokens": 73681085.0,
      "step": 33635
    },
    {
      "entropy": 5.6406089782714846,
      "epoch": 3.33862643906312,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003912285948794055,
      "loss": 5.045,
      "mean_token_accuracy": 0.19646768867969513,
      "num_tokens": 73693212.0,
      "step": 33640
    },
    {
      "entropy": 5.654138660430908,
      "epoch": 3.3391226677252877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003911982554311573,
      "loss": 5.0233,
      "mean_token_accuracy": 0.18406944423913957,
      "num_tokens": 73704313.0,
      "step": 33645
    },
    {
      "entropy": 5.651717901229858,
      "epoch": 3.3396188963874556,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039116791310144275,
      "loss": 4.9842,
      "mean_token_accuracy": 0.18924235850572585,
      "num_tokens": 73715271.0,
      "step": 33650
    },
    {
      "entropy": 5.650301122665406,
      "epoch": 3.340115125049623,
      "grad_norm": 1.125,
      "learning_rate": 0.0003911375678910142,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19383232593536376,
      "num_tokens": 73726309.0,
      "step": 33655
    },
    {
      "entropy": 5.683779716491699,
      "epoch": 3.3406113537117905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039110721980062434,
      "loss": 5.008,
      "mean_token_accuracy": 0.18423021286726,
      "num_tokens": 73737037.0,
      "step": 33660
    },
    {
      "entropy": 5.606402015686035,
      "epoch": 3.341107582373958,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039107686883102554,
      "loss": 4.9955,
      "mean_token_accuracy": 0.1887106791138649,
      "num_tokens": 73748206.0,
      "step": 33665
    },
    {
      "entropy": 5.712091398239136,
      "epoch": 3.3416038110361255,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039104651498297053,
      "loss": 5.1256,
      "mean_token_accuracy": 0.18622902333736419,
      "num_tokens": 73760034.0,
      "step": 33670
    },
    {
      "entropy": 5.673376560211182,
      "epoch": 3.342100039698293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000391016158257212,
      "loss": 4.9829,
      "mean_token_accuracy": 0.1842476397752762,
      "num_tokens": 73772050.0,
      "step": 33675
    },
    {
      "entropy": 5.612176609039307,
      "epoch": 3.3425962683604604,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003909857986545028,
      "loss": 4.9848,
      "mean_token_accuracy": 0.18693505674600602,
      "num_tokens": 73782577.0,
      "step": 33680
    },
    {
      "entropy": 5.613306140899658,
      "epoch": 3.343092497022628,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039095543617559566,
      "loss": 4.969,
      "mean_token_accuracy": 0.18553595691919328,
      "num_tokens": 73792627.0,
      "step": 33685
    },
    {
      "entropy": 5.668084287643433,
      "epoch": 3.3435887256847954,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003909250708212436,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18175472915172577,
      "num_tokens": 73803910.0,
      "step": 33690
    },
    {
      "entropy": 5.67586579322815,
      "epoch": 3.344084954346963,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039089470259219956,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18453291654586793,
      "num_tokens": 73815224.0,
      "step": 33695
    },
    {
      "entropy": 5.66652774810791,
      "epoch": 3.3445811830091308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039086433148921665,
      "loss": 5.0561,
      "mean_token_accuracy": 0.18269995152950286,
      "num_tokens": 73826448.0,
      "step": 33700
    },
    {
      "entropy": 5.658696126937866,
      "epoch": 3.3450774116712982,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003908339575130479,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19241290986537934,
      "num_tokens": 73837282.0,
      "step": 33705
    },
    {
      "entropy": 5.696055746078491,
      "epoch": 3.3455736403334657,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003908035806644467,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18583523333072663,
      "num_tokens": 73847490.0,
      "step": 33710
    },
    {
      "entropy": 5.685686445236206,
      "epoch": 3.346069868995633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039077320094416617,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18846254497766496,
      "num_tokens": 73858941.0,
      "step": 33715
    },
    {
      "entropy": 5.725353765487671,
      "epoch": 3.3465660976578007,
      "grad_norm": 1.25,
      "learning_rate": 0.0003907428183529598,
      "loss": 5.132,
      "mean_token_accuracy": 0.18386162519454957,
      "num_tokens": 73870935.0,
      "step": 33720
    },
    {
      "entropy": 5.5626623153686525,
      "epoch": 3.347062326319968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039071243289158085,
      "loss": 4.9278,
      "mean_token_accuracy": 0.19298651367425917,
      "num_tokens": 73882971.0,
      "step": 33725
    },
    {
      "entropy": 5.626025867462158,
      "epoch": 3.3475585549821356,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039068204456078296,
      "loss": 4.9589,
      "mean_token_accuracy": 0.18556301295757294,
      "num_tokens": 73893734.0,
      "step": 33730
    },
    {
      "entropy": 5.668165683746338,
      "epoch": 3.3480547836443035,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039065165336131963,
      "loss": 4.9639,
      "mean_token_accuracy": 0.18377825021743774,
      "num_tokens": 73904255.0,
      "step": 33735
    },
    {
      "entropy": 5.6530139446258545,
      "epoch": 3.348551012306471,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003906212592939445,
      "loss": 4.9162,
      "mean_token_accuracy": 0.19587219953536988,
      "num_tokens": 73915231.0,
      "step": 33740
    },
    {
      "entropy": 5.642616701126099,
      "epoch": 3.3490472409686385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039059086235941127,
      "loss": 5.0128,
      "mean_token_accuracy": 0.1889890506863594,
      "num_tokens": 73925214.0,
      "step": 33745
    },
    {
      "entropy": 5.660487079620362,
      "epoch": 3.349543469630806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003905604625584737,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18881827145814895,
      "num_tokens": 73936763.0,
      "step": 33750
    },
    {
      "entropy": 5.674617671966553,
      "epoch": 3.3500396982929734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003905300598918857,
      "loss": 4.9794,
      "mean_token_accuracy": 0.1888033151626587,
      "num_tokens": 73947302.0,
      "step": 33755
    },
    {
      "entropy": 5.750268363952637,
      "epoch": 3.350535926955141,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039049965436040125,
      "loss": 5.1189,
      "mean_token_accuracy": 0.18320928812026976,
      "num_tokens": 73961673.0,
      "step": 33760
    },
    {
      "entropy": 5.678797197341919,
      "epoch": 3.3510321556173084,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003904692459647741,
      "loss": 4.9697,
      "mean_token_accuracy": 0.20079125016927718,
      "num_tokens": 73973246.0,
      "step": 33765
    },
    {
      "entropy": 5.727968549728393,
      "epoch": 3.351528384279476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039043883470575856,
      "loss": 5.0128,
      "mean_token_accuracy": 0.19211111962795258,
      "num_tokens": 73984512.0,
      "step": 33770
    },
    {
      "entropy": 5.644961500167847,
      "epoch": 3.3520246129416433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039040842058410863,
      "loss": 4.9308,
      "mean_token_accuracy": 0.19518663883209228,
      "num_tokens": 73995744.0,
      "step": 33775
    },
    {
      "entropy": 5.61945161819458,
      "epoch": 3.352520841603811,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039037800360057864,
      "loss": 4.9799,
      "mean_token_accuracy": 0.18448541909456254,
      "num_tokens": 74006397.0,
      "step": 33780
    },
    {
      "entropy": 5.620029735565185,
      "epoch": 3.3530170702659787,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003903475837559226,
      "loss": 5.0212,
      "mean_token_accuracy": 0.17961591184139253,
      "num_tokens": 74016845.0,
      "step": 33785
    },
    {
      "entropy": 5.675936079025268,
      "epoch": 3.353513298928146,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039031716105089516,
      "loss": 4.9928,
      "mean_token_accuracy": 0.19015650749206542,
      "num_tokens": 74027669.0,
      "step": 33790
    },
    {
      "entropy": 5.628083419799805,
      "epoch": 3.3540095275903137,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003902867354862506,
      "loss": 4.9898,
      "mean_token_accuracy": 0.18917696923017502,
      "num_tokens": 74038800.0,
      "step": 33795
    },
    {
      "entropy": 5.677988147735595,
      "epoch": 3.354505756252481,
      "grad_norm": 1.125,
      "learning_rate": 0.0003902563070627434,
      "loss": 5.1095,
      "mean_token_accuracy": 0.18012060970067978,
      "num_tokens": 74050960.0,
      "step": 33800
    },
    {
      "entropy": 5.6992682933807375,
      "epoch": 3.3550019849146486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039022587578112813,
      "loss": 4.9891,
      "mean_token_accuracy": 0.1858360916376114,
      "num_tokens": 74062311.0,
      "step": 33805
    },
    {
      "entropy": 5.632759284973145,
      "epoch": 3.355498213576816,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039019544164215956,
      "loss": 4.9592,
      "mean_token_accuracy": 0.18645552098751067,
      "num_tokens": 74073046.0,
      "step": 33810
    },
    {
      "entropy": 5.616465473175049,
      "epoch": 3.3559944422389836,
      "grad_norm": 1.125,
      "learning_rate": 0.00039016500464659213,
      "loss": 4.9549,
      "mean_token_accuracy": 0.1917595237493515,
      "num_tokens": 74083244.0,
      "step": 33815
    },
    {
      "entropy": 5.658607625961304,
      "epoch": 3.3564906709011515,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003901345647951807,
      "loss": 4.993,
      "mean_token_accuracy": 0.19296015202999114,
      "num_tokens": 74093229.0,
      "step": 33820
    },
    {
      "entropy": 5.644766616821289,
      "epoch": 3.356986899563319,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039010412208868026,
      "loss": 4.9708,
      "mean_token_accuracy": 0.1859295293688774,
      "num_tokens": 74104646.0,
      "step": 33825
    },
    {
      "entropy": 5.6410833358764645,
      "epoch": 3.3574831282254864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003900736765278455,
      "loss": 5.0121,
      "mean_token_accuracy": 0.18809131383895875,
      "num_tokens": 74115484.0,
      "step": 33830
    },
    {
      "entropy": 5.5995354652404785,
      "epoch": 3.357979356887654,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039004322811343165,
      "loss": 4.9746,
      "mean_token_accuracy": 0.1893061101436615,
      "num_tokens": 74126793.0,
      "step": 33835
    },
    {
      "entropy": 5.611005878448486,
      "epoch": 3.3584755855498214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039001277684619357,
      "loss": 4.9904,
      "mean_token_accuracy": 0.1854339674115181,
      "num_tokens": 74137815.0,
      "step": 33840
    },
    {
      "entropy": 5.632381200790405,
      "epoch": 3.358971814211989,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003899823227268865,
      "loss": 4.9201,
      "mean_token_accuracy": 0.19378723204135895,
      "num_tokens": 74147591.0,
      "step": 33845
    },
    {
      "entropy": 5.612455749511719,
      "epoch": 3.3594680428741563,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038995186575626554,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18213758021593093,
      "num_tokens": 74158952.0,
      "step": 33850
    },
    {
      "entropy": 5.688726282119751,
      "epoch": 3.359964271536324,
      "grad_norm": 1.09375,
      "learning_rate": 0.000389921405935086,
      "loss": 4.9735,
      "mean_token_accuracy": 0.1901037275791168,
      "num_tokens": 74170041.0,
      "step": 33855
    },
    {
      "entropy": 5.672771692276001,
      "epoch": 3.3604605001984913,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038989094326410327,
      "loss": 4.9916,
      "mean_token_accuracy": 0.1903808519244194,
      "num_tokens": 74180825.0,
      "step": 33860
    },
    {
      "entropy": 5.5931077003479,
      "epoch": 3.3609567288606588,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003898604777440727,
      "loss": 4.9221,
      "mean_token_accuracy": 0.18856669068336487,
      "num_tokens": 74191645.0,
      "step": 33865
    },
    {
      "entropy": 5.628138160705566,
      "epoch": 3.3614529575228267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003898300093757499,
      "loss": 4.9961,
      "mean_token_accuracy": 0.1862305745482445,
      "num_tokens": 74202621.0,
      "step": 33870
    },
    {
      "entropy": 5.606231689453125,
      "epoch": 3.361949186184994,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003897995381598901,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2024626240134239,
      "num_tokens": 74213972.0,
      "step": 33875
    },
    {
      "entropy": 5.675998783111572,
      "epoch": 3.3624454148471616,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003897690640972492,
      "loss": 5.0084,
      "mean_token_accuracy": 0.19029929637908935,
      "num_tokens": 74226086.0,
      "step": 33880
    },
    {
      "entropy": 5.62421498298645,
      "epoch": 3.362941643509329,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038973858718858295,
      "loss": 5.0306,
      "mean_token_accuracy": 0.18552486896514891,
      "num_tokens": 74239404.0,
      "step": 33885
    },
    {
      "entropy": 5.617890071868897,
      "epoch": 3.3634378721714966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003897081074346468,
      "loss": 4.9035,
      "mean_token_accuracy": 0.19248816668987273,
      "num_tokens": 74250672.0,
      "step": 33890
    },
    {
      "entropy": 5.6795703887939455,
      "epoch": 3.363934100833664,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038967762483619684,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19263024181127547,
      "num_tokens": 74260587.0,
      "step": 33895
    },
    {
      "entropy": 5.65102071762085,
      "epoch": 3.3644303294958315,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003896471393939889,
      "loss": 5.0398,
      "mean_token_accuracy": 0.18504211008548738,
      "num_tokens": 74270315.0,
      "step": 33900
    },
    {
      "entropy": 5.616307592391967,
      "epoch": 3.364926558157999,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003896166511087789,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18811099082231522,
      "num_tokens": 74281380.0,
      "step": 33905
    },
    {
      "entropy": 5.694399118423462,
      "epoch": 3.365422786820167,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000389586159981323,
      "loss": 4.9573,
      "mean_token_accuracy": 0.18846535682678223,
      "num_tokens": 74292425.0,
      "step": 33910
    },
    {
      "entropy": 5.663718605041504,
      "epoch": 3.3659190154823344,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003895556660123771,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18650489449501037,
      "num_tokens": 74303455.0,
      "step": 33915
    },
    {
      "entropy": 5.691957473754883,
      "epoch": 3.366415244144502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038952516920269764,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17620498389005662,
      "num_tokens": 74314397.0,
      "step": 33920
    },
    {
      "entropy": 5.731981706619263,
      "epoch": 3.3669114728066694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003894946695530408,
      "loss": 5.0312,
      "mean_token_accuracy": 0.18557399213314058,
      "num_tokens": 74326479.0,
      "step": 33925
    },
    {
      "entropy": 5.769417476654053,
      "epoch": 3.367407701468837,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038946416706416275,
      "loss": 5.1042,
      "mean_token_accuracy": 0.18261929750442504,
      "num_tokens": 74339100.0,
      "step": 33930
    },
    {
      "entropy": 5.6695887565612795,
      "epoch": 3.3679039301310043,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003894336617368201,
      "loss": 4.9683,
      "mean_token_accuracy": 0.18880395591259003,
      "num_tokens": 74349601.0,
      "step": 33935
    },
    {
      "entropy": 5.618640279769897,
      "epoch": 3.368400158793172,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038940315357176926,
      "loss": 4.9678,
      "mean_token_accuracy": 0.1891314923763275,
      "num_tokens": 74361074.0,
      "step": 33940
    },
    {
      "entropy": 5.696162319183349,
      "epoch": 3.3688963874553393,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038937264256976657,
      "loss": 5.0012,
      "mean_token_accuracy": 0.1867850363254547,
      "num_tokens": 74372411.0,
      "step": 33945
    },
    {
      "entropy": 5.649953985214234,
      "epoch": 3.3693926161175067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003893421287315689,
      "loss": 4.9805,
      "mean_token_accuracy": 0.18835909962654113,
      "num_tokens": 74382456.0,
      "step": 33950
    },
    {
      "entropy": 5.641812133789062,
      "epoch": 3.3698888447796747,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003893116120579328,
      "loss": 4.9337,
      "mean_token_accuracy": 0.19197751432657242,
      "num_tokens": 74393435.0,
      "step": 33955
    },
    {
      "entropy": 5.683027458190918,
      "epoch": 3.370385073441842,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003892810925496151,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18141921907663344,
      "num_tokens": 74403267.0,
      "step": 33960
    },
    {
      "entropy": 5.6851499557495115,
      "epoch": 3.3708813021040096,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003892505702073725,
      "loss": 5.0081,
      "mean_token_accuracy": 0.18935341536998748,
      "num_tokens": 74415120.0,
      "step": 33965
    },
    {
      "entropy": 5.594009923934936,
      "epoch": 3.371377530766177,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038922004503196195,
      "loss": 4.9394,
      "mean_token_accuracy": 0.19510009735822678,
      "num_tokens": 74426215.0,
      "step": 33970
    },
    {
      "entropy": 5.651070547103882,
      "epoch": 3.3718737594283446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003891895170241405,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18233417272567748,
      "num_tokens": 74437693.0,
      "step": 33975
    },
    {
      "entropy": 5.676677179336548,
      "epoch": 3.372369988090512,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038915898618466487,
      "loss": 5.018,
      "mean_token_accuracy": 0.18124495148658754,
      "num_tokens": 74447861.0,
      "step": 33980
    },
    {
      "entropy": 5.6929549217224125,
      "epoch": 3.3728662167526795,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038912845251429256,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18767764121294023,
      "num_tokens": 74458927.0,
      "step": 33985
    },
    {
      "entropy": 5.721896982192993,
      "epoch": 3.373362445414847,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038909791601378054,
      "loss": 5.0478,
      "mean_token_accuracy": 0.18200832158327102,
      "num_tokens": 74468813.0,
      "step": 33990
    },
    {
      "entropy": 5.651466226577758,
      "epoch": 3.373858674077015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038906737668388595,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17183421701192855,
      "num_tokens": 74480775.0,
      "step": 33995
    },
    {
      "entropy": 5.637767887115478,
      "epoch": 3.3743549027391824,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038903683452536634,
      "loss": 4.9222,
      "mean_token_accuracy": 0.1920903891324997,
      "num_tokens": 74490343.0,
      "step": 34000
    },
    {
      "entropy": 5.650766515731812,
      "epoch": 3.37485113140135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003890062895389789,
      "loss": 4.9177,
      "mean_token_accuracy": 0.1924394428730011,
      "num_tokens": 74499995.0,
      "step": 34005
    },
    {
      "entropy": 5.6281966209411625,
      "epoch": 3.3753473600635173,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003889757417254811,
      "loss": 4.9658,
      "mean_token_accuracy": 0.18677852749824525,
      "num_tokens": 74511338.0,
      "step": 34010
    },
    {
      "entropy": 5.687041902542115,
      "epoch": 3.375843588725685,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038894519108563047,
      "loss": 5.0268,
      "mean_token_accuracy": 0.1821264609694481,
      "num_tokens": 74521865.0,
      "step": 34015
    },
    {
      "entropy": 5.621453046798706,
      "epoch": 3.3763398173878523,
      "grad_norm": 1.125,
      "learning_rate": 0.00038891463762018476,
      "loss": 4.9703,
      "mean_token_accuracy": 0.1931063190102577,
      "num_tokens": 74532518.0,
      "step": 34020
    },
    {
      "entropy": 5.642768907546997,
      "epoch": 3.3768360460500197,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038888408132990146,
      "loss": 4.9896,
      "mean_token_accuracy": 0.1891770839691162,
      "num_tokens": 74543425.0,
      "step": 34025
    },
    {
      "entropy": 5.701641464233399,
      "epoch": 3.377332274712187,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038885352221553827,
      "loss": 5.0341,
      "mean_token_accuracy": 0.17679747939109802,
      "num_tokens": 74555021.0,
      "step": 34030
    },
    {
      "entropy": 5.656613826751709,
      "epoch": 3.3778285033743547,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003888229602778532,
      "loss": 5.0311,
      "mean_token_accuracy": 0.18461243361234664,
      "num_tokens": 74565396.0,
      "step": 34035
    },
    {
      "entropy": 5.684913158416748,
      "epoch": 3.3783247320365226,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038879239551760386,
      "loss": 5.033,
      "mean_token_accuracy": 0.181166310608387,
      "num_tokens": 74577379.0,
      "step": 34040
    },
    {
      "entropy": 5.568873739242553,
      "epoch": 3.37882096069869,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003887618279355483,
      "loss": 4.9257,
      "mean_token_accuracy": 0.1922122672200203,
      "num_tokens": 74587449.0,
      "step": 34045
    },
    {
      "entropy": 5.588452529907227,
      "epoch": 3.3793171893608576,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038873125753244466,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18992105275392532,
      "num_tokens": 74598426.0,
      "step": 34050
    },
    {
      "entropy": 5.6213696002960205,
      "epoch": 3.379813418023025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003887006843090508,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18609188944101335,
      "num_tokens": 74609316.0,
      "step": 34055
    },
    {
      "entropy": 5.641353797912598,
      "epoch": 3.3803096466851925,
      "grad_norm": 1.125,
      "learning_rate": 0.000388670108266125,
      "loss": 4.9647,
      "mean_token_accuracy": 0.1918978810310364,
      "num_tokens": 74619451.0,
      "step": 34060
    },
    {
      "entropy": 5.639559698104859,
      "epoch": 3.38080587534736,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003886395294044255,
      "loss": 4.9188,
      "mean_token_accuracy": 0.19967617094516754,
      "num_tokens": 74629830.0,
      "step": 34065
    },
    {
      "entropy": 5.599289083480835,
      "epoch": 3.3813021040095275,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003886089477247105,
      "loss": 4.9732,
      "mean_token_accuracy": 0.18888098299503325,
      "num_tokens": 74640890.0,
      "step": 34070
    },
    {
      "entropy": 5.687490940093994,
      "epoch": 3.381798332671695,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003885783632277384,
      "loss": 5.067,
      "mean_token_accuracy": 0.18291052877902986,
      "num_tokens": 74652031.0,
      "step": 34075
    },
    {
      "entropy": 5.6579625606536865,
      "epoch": 3.382294561333863,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003885477759142676,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18034628033638,
      "num_tokens": 74662269.0,
      "step": 34080
    },
    {
      "entropy": 5.621004581451416,
      "epoch": 3.3827907899960303,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003885171857850567,
      "loss": 5.0096,
      "mean_token_accuracy": 0.186972214281559,
      "num_tokens": 74673799.0,
      "step": 34085
    },
    {
      "entropy": 5.663876152038574,
      "epoch": 3.383287018658198,
      "grad_norm": 1.25,
      "learning_rate": 0.0003884865928408642,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18882728070020677,
      "num_tokens": 74685130.0,
      "step": 34090
    },
    {
      "entropy": 5.677091312408447,
      "epoch": 3.3837832473203653,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003884559970824486,
      "loss": 5.0798,
      "mean_token_accuracy": 0.181556236743927,
      "num_tokens": 74695762.0,
      "step": 34095
    },
    {
      "entropy": 5.647458600997925,
      "epoch": 3.3842794759825328,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003884253985105689,
      "loss": 5.0094,
      "mean_token_accuracy": 0.1838410019874573,
      "num_tokens": 74706436.0,
      "step": 34100
    },
    {
      "entropy": 5.668571662902832,
      "epoch": 3.3847757046447002,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003883947971259837,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18461159765720367,
      "num_tokens": 74717787.0,
      "step": 34105
    },
    {
      "entropy": 5.626921033859253,
      "epoch": 3.3852719333068677,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003883641929294519,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18417952358722686,
      "num_tokens": 74728922.0,
      "step": 34110
    },
    {
      "entropy": 5.643360328674317,
      "epoch": 3.385768161969035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003883335859217323,
      "loss": 4.9543,
      "mean_token_accuracy": 0.18412410318851472,
      "num_tokens": 74739737.0,
      "step": 34115
    },
    {
      "entropy": 5.649766969680786,
      "epoch": 3.3862643906312027,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000388302976103584,
      "loss": 5.008,
      "mean_token_accuracy": 0.1859864503145218,
      "num_tokens": 74749969.0,
      "step": 34120
    },
    {
      "entropy": 5.657258462905884,
      "epoch": 3.3867606192933706,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003882723634757661,
      "loss": 4.9468,
      "mean_token_accuracy": 0.191874723136425,
      "num_tokens": 74760735.0,
      "step": 34125
    },
    {
      "entropy": 5.624790954589844,
      "epoch": 3.387256847955538,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003882417480390376,
      "loss": 4.9129,
      "mean_token_accuracy": 0.19677374362945557,
      "num_tokens": 74771769.0,
      "step": 34130
    },
    {
      "entropy": 5.6331418514251705,
      "epoch": 3.3877530766177055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038821112979415786,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18895514011383058,
      "num_tokens": 74780967.0,
      "step": 34135
    },
    {
      "entropy": 5.6484130859375,
      "epoch": 3.388249305279873,
      "grad_norm": 1.015625,
      "learning_rate": 0.000388180508741886,
      "loss": 4.96,
      "mean_token_accuracy": 0.1896696627140045,
      "num_tokens": 74793514.0,
      "step": 34140
    },
    {
      "entropy": 5.652983283996582,
      "epoch": 3.3887455339420405,
      "grad_norm": 1.34375,
      "learning_rate": 0.00038814988488298136,
      "loss": 4.958,
      "mean_token_accuracy": 0.18594396114349365,
      "num_tokens": 74803970.0,
      "step": 34145
    },
    {
      "entropy": 5.655226802825927,
      "epoch": 3.389241762604208,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003881192582182034,
      "loss": 4.9675,
      "mean_token_accuracy": 0.1855667933821678,
      "num_tokens": 74814900.0,
      "step": 34150
    },
    {
      "entropy": 5.607626295089721,
      "epoch": 3.3897379912663754,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038808862874831164,
      "loss": 4.9547,
      "mean_token_accuracy": 0.18341802060604095,
      "num_tokens": 74825749.0,
      "step": 34155
    },
    {
      "entropy": 5.628345584869384,
      "epoch": 3.390234219928543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003880579964740655,
      "loss": 4.9604,
      "mean_token_accuracy": 0.19067326486110686,
      "num_tokens": 74837176.0,
      "step": 34160
    },
    {
      "entropy": 5.6539734363555905,
      "epoch": 3.390730448590711,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038802736139622473,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18912942707538605,
      "num_tokens": 74848566.0,
      "step": 34165
    },
    {
      "entropy": 5.685419368743896,
      "epoch": 3.3912266772528783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038799672351554895,
      "loss": 5.0194,
      "mean_token_accuracy": 0.18692058175802231,
      "num_tokens": 74859938.0,
      "step": 34170
    },
    {
      "entropy": 5.633371734619141,
      "epoch": 3.3917229059150458,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038796608283279786,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18378352820873262,
      "num_tokens": 74873716.0,
      "step": 34175
    },
    {
      "entropy": 5.720241641998291,
      "epoch": 3.3922191345772132,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003879354393487314,
      "loss": 5.0658,
      "mean_token_accuracy": 0.18613173216581344,
      "num_tokens": 74884188.0,
      "step": 34180
    },
    {
      "entropy": 5.719536304473877,
      "epoch": 3.3927153632393807,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038790479306410934,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17659202963113785,
      "num_tokens": 74894526.0,
      "step": 34185
    },
    {
      "entropy": 5.645160055160522,
      "epoch": 3.393211591901548,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003878741439796918,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18753945231437683,
      "num_tokens": 74906586.0,
      "step": 34190
    },
    {
      "entropy": 5.667746067047119,
      "epoch": 3.3937078205637157,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003878434920962386,
      "loss": 4.9948,
      "mean_token_accuracy": 0.18527491092681886,
      "num_tokens": 74919001.0,
      "step": 34195
    },
    {
      "entropy": 5.628152179718017,
      "epoch": 3.394204049225883,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038781283741451,
      "loss": 4.9145,
      "mean_token_accuracy": 0.19075217545032502,
      "num_tokens": 74929634.0,
      "step": 34200
    },
    {
      "entropy": 5.637691926956177,
      "epoch": 3.3947002778880506,
      "grad_norm": 1.125,
      "learning_rate": 0.00038778217993526616,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18639865815639495,
      "num_tokens": 74939559.0,
      "step": 34205
    },
    {
      "entropy": 5.676739835739136,
      "epoch": 3.395196506550218,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003877515196592672,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18993424475193024,
      "num_tokens": 74950320.0,
      "step": 34210
    },
    {
      "entropy": 5.644649410247803,
      "epoch": 3.395692735212386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038772085658727353,
      "loss": 5.0101,
      "mean_token_accuracy": 0.18600819557905196,
      "num_tokens": 74961904.0,
      "step": 34215
    },
    {
      "entropy": 5.624983453750611,
      "epoch": 3.3961889638745535,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038769019072004555,
      "loss": 4.9569,
      "mean_token_accuracy": 0.19181010574102403,
      "num_tokens": 74973100.0,
      "step": 34220
    },
    {
      "entropy": 5.64586443901062,
      "epoch": 3.396685192536721,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003876595220583436,
      "loss": 5.01,
      "mean_token_accuracy": 0.1897437423467636,
      "num_tokens": 74984150.0,
      "step": 34225
    },
    {
      "entropy": 5.703281831741333,
      "epoch": 3.3971814211988884,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003876288506029283,
      "loss": 5.0464,
      "mean_token_accuracy": 0.18706174492835997,
      "num_tokens": 74994403.0,
      "step": 34230
    },
    {
      "entropy": 5.72701997756958,
      "epoch": 3.397677649861056,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003875981763545602,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17858375757932662,
      "num_tokens": 75005227.0,
      "step": 34235
    },
    {
      "entropy": 5.791402292251587,
      "epoch": 3.3981738785232234,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038756749931399993,
      "loss": 5.0944,
      "mean_token_accuracy": 0.17984263896942138,
      "num_tokens": 75015932.0,
      "step": 34240
    },
    {
      "entropy": 5.664501619338989,
      "epoch": 3.398670107185391,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003875368194820083,
      "loss": 5.0576,
      "mean_token_accuracy": 0.1881216749548912,
      "num_tokens": 75027664.0,
      "step": 34245
    },
    {
      "entropy": 5.688663673400879,
      "epoch": 3.399166335847559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000387506136859346,
      "loss": 4.965,
      "mean_token_accuracy": 0.18898870497941972,
      "num_tokens": 75037987.0,
      "step": 34250
    },
    {
      "entropy": 5.686731004714966,
      "epoch": 3.3996625645097263,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003874754514467739,
      "loss": 5.0275,
      "mean_token_accuracy": 0.18593038767576217,
      "num_tokens": 75049227.0,
      "step": 34255
    },
    {
      "entropy": 5.6496350288391115,
      "epoch": 3.4001587931718937,
      "grad_norm": 1.078125,
      "learning_rate": 0.000387444763245053,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18729568421840667,
      "num_tokens": 75060213.0,
      "step": 34260
    },
    {
      "entropy": 5.670573329925537,
      "epoch": 3.400655021834061,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003874140722549443,
      "loss": 4.9424,
      "mean_token_accuracy": 0.18990874141454697,
      "num_tokens": 75071466.0,
      "step": 34265
    },
    {
      "entropy": 5.685924386978149,
      "epoch": 3.4011512504962287,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003873833784772088,
      "loss": 5.1161,
      "mean_token_accuracy": 0.176593554019928,
      "num_tokens": 75082966.0,
      "step": 34270
    },
    {
      "entropy": 5.660399055480957,
      "epoch": 3.401647479158396,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003873526819126077,
      "loss": 4.9533,
      "mean_token_accuracy": 0.19592699855566026,
      "num_tokens": 75093860.0,
      "step": 34275
    },
    {
      "entropy": 5.696930074691773,
      "epoch": 3.4021437078205636,
      "grad_norm": 1.25,
      "learning_rate": 0.0003873219825619022,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1861377999186516,
      "num_tokens": 75104468.0,
      "step": 34280
    },
    {
      "entropy": 5.612629985809326,
      "epoch": 3.402639936482731,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003872912804258536,
      "loss": 4.9765,
      "mean_token_accuracy": 0.18724024593830108,
      "num_tokens": 75117938.0,
      "step": 34285
    },
    {
      "entropy": 5.658967971801758,
      "epoch": 3.4031361651448986,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038726057550522307,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18194178938865663,
      "num_tokens": 75128354.0,
      "step": 34290
    },
    {
      "entropy": 5.707356739044189,
      "epoch": 3.403632393807066,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038722986780077227,
      "loss": 4.9949,
      "mean_token_accuracy": 0.18775756806135177,
      "num_tokens": 75139145.0,
      "step": 34295
    },
    {
      "entropy": 5.668583583831787,
      "epoch": 3.404128622469234,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038719915731326264,
      "loss": 4.937,
      "mean_token_accuracy": 0.1918017566204071,
      "num_tokens": 75148887.0,
      "step": 34300
    },
    {
      "entropy": 5.652385711669922,
      "epoch": 3.4046248511314015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038716844404345556,
      "loss": 5.057,
      "mean_token_accuracy": 0.1853220507502556,
      "num_tokens": 75160455.0,
      "step": 34305
    },
    {
      "entropy": 5.606260681152344,
      "epoch": 3.405121079793569,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038713772799211284,
      "loss": 4.9203,
      "mean_token_accuracy": 0.1878638207912445,
      "num_tokens": 75172100.0,
      "step": 34310
    },
    {
      "entropy": 5.684635877609253,
      "epoch": 3.4056173084557364,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003871070091599961,
      "loss": 4.9162,
      "mean_token_accuracy": 0.189907069504261,
      "num_tokens": 75183573.0,
      "step": 34315
    },
    {
      "entropy": 5.65213189125061,
      "epoch": 3.406113537117904,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038707628754786706,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18727974444627762,
      "num_tokens": 75194231.0,
      "step": 34320
    },
    {
      "entropy": 5.6293059349060055,
      "epoch": 3.4066097657800714,
      "grad_norm": 1.125,
      "learning_rate": 0.00038704556315648757,
      "loss": 5.0209,
      "mean_token_accuracy": 0.18908983767032622,
      "num_tokens": 75206256.0,
      "step": 34325
    },
    {
      "entropy": 5.637465715408325,
      "epoch": 3.407105994442239,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003870148359866196,
      "loss": 4.9571,
      "mean_token_accuracy": 0.18592972308397293,
      "num_tokens": 75216967.0,
      "step": 34330
    },
    {
      "entropy": 5.601105690002441,
      "epoch": 3.4076022231044067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000386984106039025,
      "loss": 4.908,
      "mean_token_accuracy": 0.1907800927758217,
      "num_tokens": 75226998.0,
      "step": 34335
    },
    {
      "entropy": 5.669775915145874,
      "epoch": 3.408098451766574,
      "grad_norm": 1.125,
      "learning_rate": 0.000386953373314466,
      "loss": 5.0774,
      "mean_token_accuracy": 0.18460873067378997,
      "num_tokens": 75238788.0,
      "step": 34340
    },
    {
      "entropy": 5.653550338745117,
      "epoch": 3.4085946804287417,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038692263781370446,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18984477669000627,
      "num_tokens": 75249820.0,
      "step": 34345
    },
    {
      "entropy": 5.669407320022583,
      "epoch": 3.409090909090909,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038689189953750275,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18437726944684982,
      "num_tokens": 75260425.0,
      "step": 34350
    },
    {
      "entropy": 5.616374921798706,
      "epoch": 3.4095871377530766,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038686115848662303,
      "loss": 4.9454,
      "mean_token_accuracy": 0.1958703428506851,
      "num_tokens": 75269880.0,
      "step": 34355
    },
    {
      "entropy": 5.645887231826782,
      "epoch": 3.410083366415244,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003868304146618276,
      "loss": 4.9859,
      "mean_token_accuracy": 0.18846474289894105,
      "num_tokens": 75279960.0,
      "step": 34360
    },
    {
      "entropy": 5.596679639816284,
      "epoch": 3.4105795950774116,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038679966806387886,
      "loss": 4.8924,
      "mean_token_accuracy": 0.19766199588775635,
      "num_tokens": 75290693.0,
      "step": 34365
    },
    {
      "entropy": 5.567077112197876,
      "epoch": 3.411075823739579,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003867689186935393,
      "loss": 4.9545,
      "mean_token_accuracy": 0.1900473490357399,
      "num_tokens": 75302093.0,
      "step": 34370
    },
    {
      "entropy": 5.651590347290039,
      "epoch": 3.4115720524017465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003867381665515712,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18493617177009583,
      "num_tokens": 75313767.0,
      "step": 34375
    },
    {
      "entropy": 5.663585472106933,
      "epoch": 3.412068281063914,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003867074116387376,
      "loss": 4.9579,
      "mean_token_accuracy": 0.1902636021375656,
      "num_tokens": 75324188.0,
      "step": 34380
    },
    {
      "entropy": 5.597812652587891,
      "epoch": 3.412564509726082,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038667665395580083,
      "loss": 4.9119,
      "mean_token_accuracy": 0.1907627448439598,
      "num_tokens": 75334410.0,
      "step": 34385
    },
    {
      "entropy": 5.639297819137573,
      "epoch": 3.4130607383882494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003866458935035236,
      "loss": 5.0374,
      "mean_token_accuracy": 0.1804431825876236,
      "num_tokens": 75344754.0,
      "step": 34390
    },
    {
      "entropy": 5.717687559127808,
      "epoch": 3.413556967050417,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038661513028266886,
      "loss": 5.0027,
      "mean_token_accuracy": 0.18808865696191787,
      "num_tokens": 75355898.0,
      "step": 34395
    },
    {
      "entropy": 5.677997303009033,
      "epoch": 3.4140531957125844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003865843642939994,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18254223167896272,
      "num_tokens": 75365945.0,
      "step": 34400
    },
    {
      "entropy": 5.649655294418335,
      "epoch": 3.414549424374752,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038655359553827807,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18920951932668686,
      "num_tokens": 75376382.0,
      "step": 34405
    },
    {
      "entropy": 5.5984845638275145,
      "epoch": 3.4150456530369193,
      "grad_norm": 1.078125,
      "learning_rate": 0.000386522824016268,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18394317477941513,
      "num_tokens": 75386987.0,
      "step": 34410
    },
    {
      "entropy": 5.621092510223389,
      "epoch": 3.415541881699087,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038649204972873225,
      "loss": 4.9329,
      "mean_token_accuracy": 0.1910938188433647,
      "num_tokens": 75397583.0,
      "step": 34415
    },
    {
      "entropy": 5.655617570877075,
      "epoch": 3.4160381103612547,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003864612726764339,
      "loss": 4.9248,
      "mean_token_accuracy": 0.1891719400882721,
      "num_tokens": 75408205.0,
      "step": 34420
    },
    {
      "entropy": 5.638412857055664,
      "epoch": 3.416534339023422,
      "grad_norm": 1.125,
      "learning_rate": 0.0003864304928601361,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18555794507265091,
      "num_tokens": 75418950.0,
      "step": 34425
    },
    {
      "entropy": 5.655980443954467,
      "epoch": 3.4170305676855897,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003863997102806023,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18838605731725694,
      "num_tokens": 75429377.0,
      "step": 34430
    },
    {
      "entropy": 5.643809366226196,
      "epoch": 3.417526796347757,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003863689249385956,
      "loss": 4.9742,
      "mean_token_accuracy": 0.18868017494678496,
      "num_tokens": 75441211.0,
      "step": 34435
    },
    {
      "entropy": 5.668684673309326,
      "epoch": 3.4180230250099246,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038633813683487954,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18526477962732316,
      "num_tokens": 75450943.0,
      "step": 34440
    },
    {
      "entropy": 5.70779504776001,
      "epoch": 3.418519253672092,
      "grad_norm": 1.125,
      "learning_rate": 0.0003863073459702177,
      "loss": 4.9899,
      "mean_token_accuracy": 0.18744012117385864,
      "num_tokens": 75461309.0,
      "step": 34445
    },
    {
      "entropy": 5.605449676513672,
      "epoch": 3.4190154823342596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038627655234537345,
      "loss": 4.9544,
      "mean_token_accuracy": 0.1926722526550293,
      "num_tokens": 75472552.0,
      "step": 34450
    },
    {
      "entropy": 5.706601142883301,
      "epoch": 3.419511710996427,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038624575596111056,
      "loss": 5.108,
      "mean_token_accuracy": 0.18748119324445725,
      "num_tokens": 75483314.0,
      "step": 34455
    },
    {
      "entropy": 5.625531911849976,
      "epoch": 3.4200079396585945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038621495681819256,
      "loss": 4.9736,
      "mean_token_accuracy": 0.19063244611024857,
      "num_tokens": 75493463.0,
      "step": 34460
    },
    {
      "entropy": 5.646122741699219,
      "epoch": 3.420504168320762,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003861841549173833,
      "loss": 4.9813,
      "mean_token_accuracy": 0.19549905210733415,
      "num_tokens": 75504406.0,
      "step": 34465
    },
    {
      "entropy": 5.666406059265137,
      "epoch": 3.42100039698293,
      "grad_norm": 1.125,
      "learning_rate": 0.0003861533502594466,
      "loss": 4.9694,
      "mean_token_accuracy": 0.18520499020814896,
      "num_tokens": 75515141.0,
      "step": 34470
    },
    {
      "entropy": 5.712864398956299,
      "epoch": 3.4214966256450974,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003861225428451463,
      "loss": 4.9865,
      "mean_token_accuracy": 0.19027609527111053,
      "num_tokens": 75525965.0,
      "step": 34475
    },
    {
      "entropy": 5.574916934967041,
      "epoch": 3.421992854307265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003860917326752464,
      "loss": 4.9414,
      "mean_token_accuracy": 0.1910489708185196,
      "num_tokens": 75536885.0,
      "step": 34480
    },
    {
      "entropy": 5.643745279312133,
      "epoch": 3.4224890829694323,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038606091975051097,
      "loss": 4.9376,
      "mean_token_accuracy": 0.18903007209300995,
      "num_tokens": 75548845.0,
      "step": 34485
    },
    {
      "entropy": 5.630414724349976,
      "epoch": 3.4229853116316,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038603010407170393,
      "loss": 4.9522,
      "mean_token_accuracy": 0.19231366068124772,
      "num_tokens": 75559002.0,
      "step": 34490
    },
    {
      "entropy": 5.6377979755401615,
      "epoch": 3.4234815402937673,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003859992856395896,
      "loss": 4.9534,
      "mean_token_accuracy": 0.1845873087644577,
      "num_tokens": 75569270.0,
      "step": 34495
    },
    {
      "entropy": 5.659086990356445,
      "epoch": 3.4239777689559348,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003859684644549322,
      "loss": 4.9747,
      "mean_token_accuracy": 0.19040183275938033,
      "num_tokens": 75579109.0,
      "step": 34500
    },
    {
      "entropy": 5.683581447601318,
      "epoch": 3.4244739976181022,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003859376405184959,
      "loss": 5.0198,
      "mean_token_accuracy": 0.1782880425453186,
      "num_tokens": 75589384.0,
      "step": 34505
    },
    {
      "entropy": 5.736395454406738,
      "epoch": 3.42497022628027,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003859068138310453,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18756965696811675,
      "num_tokens": 75600851.0,
      "step": 34510
    },
    {
      "entropy": 5.663330745697022,
      "epoch": 3.4254664549424376,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038587598439334466,
      "loss": 4.9368,
      "mean_token_accuracy": 0.18632653057575227,
      "num_tokens": 75611329.0,
      "step": 34515
    },
    {
      "entropy": 5.586145305633545,
      "epoch": 3.425962683604605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038584515220615843,
      "loss": 4.9912,
      "mean_token_accuracy": 0.1852847456932068,
      "num_tokens": 75621579.0,
      "step": 34520
    },
    {
      "entropy": 5.63627610206604,
      "epoch": 3.4264589122667726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038581431727025136,
      "loss": 4.9515,
      "mean_token_accuracy": 0.19010138511657715,
      "num_tokens": 75632366.0,
      "step": 34525
    },
    {
      "entropy": 5.596221923828125,
      "epoch": 3.42695514092894,
      "grad_norm": 1.265625,
      "learning_rate": 0.000385783479586388,
      "loss": 4.8886,
      "mean_token_accuracy": 0.20333942174911498,
      "num_tokens": 75642572.0,
      "step": 34530
    },
    {
      "entropy": 5.6609046459198,
      "epoch": 3.4274513695911075,
      "grad_norm": 1.125,
      "learning_rate": 0.0003857526391553331,
      "loss": 4.9452,
      "mean_token_accuracy": 0.18873332738876342,
      "num_tokens": 75653523.0,
      "step": 34535
    },
    {
      "entropy": 5.657945966720581,
      "epoch": 3.427947598253275,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038572179597785134,
      "loss": 5.0035,
      "mean_token_accuracy": 0.1906817227602005,
      "num_tokens": 75663065.0,
      "step": 34540
    },
    {
      "entropy": 5.753428316116333,
      "epoch": 3.4284438269154425,
      "grad_norm": 1.359375,
      "learning_rate": 0.00038569095005470765,
      "loss": 5.1315,
      "mean_token_accuracy": 0.18741156458854674,
      "num_tokens": 75674063.0,
      "step": 34545
    },
    {
      "entropy": 5.72303991317749,
      "epoch": 3.42894005557761,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038566010138666693,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18236401975154876,
      "num_tokens": 75685572.0,
      "step": 34550
    },
    {
      "entropy": 5.62032790184021,
      "epoch": 3.429436284239778,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038562924997449414,
      "loss": 4.9743,
      "mean_token_accuracy": 0.187100288271904,
      "num_tokens": 75696190.0,
      "step": 34555
    },
    {
      "entropy": 5.6859382629394535,
      "epoch": 3.4299325129019453,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003855983958189543,
      "loss": 5.042,
      "mean_token_accuracy": 0.18134492337703706,
      "num_tokens": 75707099.0,
      "step": 34560
    },
    {
      "entropy": 5.672867631912231,
      "epoch": 3.430428741564113,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003855675389208126,
      "loss": 4.9606,
      "mean_token_accuracy": 0.19321242421865464,
      "num_tokens": 75716751.0,
      "step": 34565
    },
    {
      "entropy": 5.645115184783935,
      "epoch": 3.4309249702262803,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003855366792808341,
      "loss": 4.9374,
      "mean_token_accuracy": 0.19297754168510436,
      "num_tokens": 75727091.0,
      "step": 34570
    },
    {
      "entropy": 5.617162132263184,
      "epoch": 3.4314211988884478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038550581689978427,
      "loss": 4.9231,
      "mean_token_accuracy": 0.192873252928257,
      "num_tokens": 75737863.0,
      "step": 34575
    },
    {
      "entropy": 5.620417165756225,
      "epoch": 3.4319174275506152,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038547495177842817,
      "loss": 5.0016,
      "mean_token_accuracy": 0.18886780440807344,
      "num_tokens": 75749773.0,
      "step": 34580
    },
    {
      "entropy": 5.688537311553955,
      "epoch": 3.4324136562127827,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003854440839175314,
      "loss": 4.9327,
      "mean_token_accuracy": 0.19040210247039796,
      "num_tokens": 75760553.0,
      "step": 34585
    },
    {
      "entropy": 5.68053092956543,
      "epoch": 3.43290988487495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038541321331785934,
      "loss": 4.9943,
      "mean_token_accuracy": 0.1855522111058235,
      "num_tokens": 75771082.0,
      "step": 34590
    },
    {
      "entropy": 5.625995111465454,
      "epoch": 3.433406113537118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038538233998017745,
      "loss": 5.0773,
      "mean_token_accuracy": 0.18479882329702377,
      "num_tokens": 75782530.0,
      "step": 34595
    },
    {
      "entropy": 5.731579256057739,
      "epoch": 3.4339023421992856,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038535146390525143,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18881592601537706,
      "num_tokens": 75793584.0,
      "step": 34600
    },
    {
      "entropy": 5.662628698348999,
      "epoch": 3.434398570861453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038532058509384694,
      "loss": 4.9728,
      "mean_token_accuracy": 0.19330462217330932,
      "num_tokens": 75804569.0,
      "step": 34605
    },
    {
      "entropy": 5.5803107738494875,
      "epoch": 3.4348947995236205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003852897035467296,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18781760185956956,
      "num_tokens": 75815697.0,
      "step": 34610
    },
    {
      "entropy": 5.596248817443848,
      "epoch": 3.435391028185788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038525881926466525,
      "loss": 4.9455,
      "mean_token_accuracy": 0.19260303676128387,
      "num_tokens": 75826787.0,
      "step": 34615
    },
    {
      "entropy": 5.739540719985962,
      "epoch": 3.4358872568479555,
      "grad_norm": 1.125,
      "learning_rate": 0.00038522793224841977,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18449333757162095,
      "num_tokens": 75837992.0,
      "step": 34620
    },
    {
      "entropy": 5.728651666641236,
      "epoch": 3.436383485510123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038519704249875917,
      "loss": 4.9971,
      "mean_token_accuracy": 0.19130730032920837,
      "num_tokens": 75849324.0,
      "step": 34625
    },
    {
      "entropy": 5.662909460067749,
      "epoch": 3.4368797141722904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003851661500164493,
      "loss": 4.9746,
      "mean_token_accuracy": 0.19196005165576935,
      "num_tokens": 75860721.0,
      "step": 34630
    },
    {
      "entropy": 5.641267156600952,
      "epoch": 3.437375942834458,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003851352548022563,
      "loss": 4.9762,
      "mean_token_accuracy": 0.18511761277914046,
      "num_tokens": 75870381.0,
      "step": 34635
    },
    {
      "entropy": 5.6582118511199955,
      "epoch": 3.437872171496626,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003851043568569463,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18306845277547837,
      "num_tokens": 75881378.0,
      "step": 34640
    },
    {
      "entropy": 5.6907751083374025,
      "epoch": 3.4383684001587933,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003850734561812855,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18952385783195497,
      "num_tokens": 75891079.0,
      "step": 34645
    },
    {
      "entropy": 5.620957756042481,
      "epoch": 3.438864628820961,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038504255277604016,
      "loss": 4.9312,
      "mean_token_accuracy": 0.19214481562376023,
      "num_tokens": 75899893.0,
      "step": 34650
    },
    {
      "entropy": 5.645739316940308,
      "epoch": 3.4393608574831283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003850116466419766,
      "loss": 5.0164,
      "mean_token_accuracy": 0.18463145494461058,
      "num_tokens": 75911291.0,
      "step": 34655
    },
    {
      "entropy": 5.659209108352661,
      "epoch": 3.4398570861452957,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003849807377798614,
      "loss": 4.9143,
      "mean_token_accuracy": 0.19127961993217468,
      "num_tokens": 75921791.0,
      "step": 34660
    },
    {
      "entropy": 5.615454578399659,
      "epoch": 3.440353314807463,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038494982619046076,
      "loss": 4.9058,
      "mean_token_accuracy": 0.19035658091306687,
      "num_tokens": 75931771.0,
      "step": 34665
    },
    {
      "entropy": 5.634121084213257,
      "epoch": 3.4408495434696307,
      "grad_norm": 1.25,
      "learning_rate": 0.0003849189118745414,
      "loss": 4.9857,
      "mean_token_accuracy": 0.193137064576149,
      "num_tokens": 75941457.0,
      "step": 34670
    },
    {
      "entropy": 5.575797367095947,
      "epoch": 3.441345772131798,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003848879948328699,
      "loss": 4.9098,
      "mean_token_accuracy": 0.1963341638445854,
      "num_tokens": 75953838.0,
      "step": 34675
    },
    {
      "entropy": 5.592445325851441,
      "epoch": 3.441842000793966,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003848570750662129,
      "loss": 4.972,
      "mean_token_accuracy": 0.19155174493789673,
      "num_tokens": 75967167.0,
      "step": 34680
    },
    {
      "entropy": 5.698031091690064,
      "epoch": 3.4423382294561335,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003848261525753371,
      "loss": 5.0303,
      "mean_token_accuracy": 0.1829117938876152,
      "num_tokens": 75978678.0,
      "step": 34685
    },
    {
      "entropy": 5.7198113918304445,
      "epoch": 3.442834458118301,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003847952273610094,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18343799114227294,
      "num_tokens": 75989940.0,
      "step": 34690
    },
    {
      "entropy": 5.683634948730469,
      "epoch": 3.4433306867804685,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003847642994239968,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19022559225559235,
      "num_tokens": 76000400.0,
      "step": 34695
    },
    {
      "entropy": 5.682715368270874,
      "epoch": 3.443826915442636,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038473336876506585,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18681044280529022,
      "num_tokens": 76011415.0,
      "step": 34700
    },
    {
      "entropy": 5.6962010860443115,
      "epoch": 3.4443231441048034,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038470243538498393,
      "loss": 5.0052,
      "mean_token_accuracy": 0.18569254726171494,
      "num_tokens": 76023440.0,
      "step": 34705
    },
    {
      "entropy": 5.655298089981079,
      "epoch": 3.444819372766971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038467149928451803,
      "loss": 5.0118,
      "mean_token_accuracy": 0.18782948702573776,
      "num_tokens": 76036795.0,
      "step": 34710
    },
    {
      "entropy": 5.636894035339355,
      "epoch": 3.4453156014291384,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038464056046443523,
      "loss": 4.8953,
      "mean_token_accuracy": 0.1985381007194519,
      "num_tokens": 76047333.0,
      "step": 34715
    },
    {
      "entropy": 5.629690313339234,
      "epoch": 3.445811830091306,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003846096189255027,
      "loss": 5.0176,
      "mean_token_accuracy": 0.18673235177993774,
      "num_tokens": 76058026.0,
      "step": 34720
    },
    {
      "entropy": 5.694230461120606,
      "epoch": 3.446308058753474,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038457867466848796,
      "loss": 4.9909,
      "mean_token_accuracy": 0.18953143507242204,
      "num_tokens": 76068283.0,
      "step": 34725
    },
    {
      "entropy": 5.64088921546936,
      "epoch": 3.4468042874156413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003845477276941582,
      "loss": 4.9464,
      "mean_token_accuracy": 0.18156130760908126,
      "num_tokens": 76078821.0,
      "step": 34730
    },
    {
      "entropy": 5.628732490539551,
      "epoch": 3.4473005160778087,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003845167780032807,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18559216111898422,
      "num_tokens": 76089681.0,
      "step": 34735
    },
    {
      "entropy": 5.671344661712647,
      "epoch": 3.447796744739976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038448582559662324,
      "loss": 5.0357,
      "mean_token_accuracy": 0.18719395101070405,
      "num_tokens": 76103165.0,
      "step": 34740
    },
    {
      "entropy": 5.64943118095398,
      "epoch": 3.4482929734021437,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038445487047495313,
      "loss": 5.0445,
      "mean_token_accuracy": 0.19072683602571489,
      "num_tokens": 76114252.0,
      "step": 34745
    },
    {
      "entropy": 5.680829477310181,
      "epoch": 3.448789202064311,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003844239126390382,
      "loss": 4.9467,
      "mean_token_accuracy": 0.1954288050532341,
      "num_tokens": 76123707.0,
      "step": 34750
    },
    {
      "entropy": 5.59231915473938,
      "epoch": 3.4492854307264786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038439295208964595,
      "loss": 4.9038,
      "mean_token_accuracy": 0.19201588481664658,
      "num_tokens": 76134732.0,
      "step": 34755
    },
    {
      "entropy": 5.610567855834961,
      "epoch": 3.449781659388646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003843619888275442,
      "loss": 5.0096,
      "mean_token_accuracy": 0.1926530659198761,
      "num_tokens": 76146187.0,
      "step": 34760
    },
    {
      "entropy": 5.664962244033814,
      "epoch": 3.450277888050814,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038433102285350074,
      "loss": 5.0513,
      "mean_token_accuracy": 0.1833432599902153,
      "num_tokens": 76157405.0,
      "step": 34765
    },
    {
      "entropy": 5.655008268356323,
      "epoch": 3.4507741167129815,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038430005416828354,
      "loss": 4.8847,
      "mean_token_accuracy": 0.1911861479282379,
      "num_tokens": 76168275.0,
      "step": 34770
    },
    {
      "entropy": 5.612475824356079,
      "epoch": 3.451270345375149,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038426908277266054,
      "loss": 4.9701,
      "mean_token_accuracy": 0.19136992543935777,
      "num_tokens": 76178882.0,
      "step": 34775
    },
    {
      "entropy": 5.656029891967774,
      "epoch": 3.4517665740373165,
      "grad_norm": 1.125,
      "learning_rate": 0.00038423810866739975,
      "loss": 4.9787,
      "mean_token_accuracy": 0.18942078948020935,
      "num_tokens": 76189937.0,
      "step": 34780
    },
    {
      "entropy": 5.67535514831543,
      "epoch": 3.452262802699484,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038420713185326926,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18208543211221695,
      "num_tokens": 76200297.0,
      "step": 34785
    },
    {
      "entropy": 5.6500402927398685,
      "epoch": 3.4527590313616514,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003841761523310372,
      "loss": 4.9829,
      "mean_token_accuracy": 0.19130569696426392,
      "num_tokens": 76211107.0,
      "step": 34790
    },
    {
      "entropy": 5.678364706039429,
      "epoch": 3.453255260023819,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003841451701014718,
      "loss": 4.9571,
      "mean_token_accuracy": 0.1917690709233284,
      "num_tokens": 76222198.0,
      "step": 34795
    },
    {
      "entropy": 5.686587476730347,
      "epoch": 3.4537514886859864,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038411418516534135,
      "loss": 5.03,
      "mean_token_accuracy": 0.18926768749952316,
      "num_tokens": 76234547.0,
      "step": 34800
    },
    {
      "entropy": 5.630616140365601,
      "epoch": 3.454247717348154,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038408319752341425,
      "loss": 4.9311,
      "mean_token_accuracy": 0.1863040879368782,
      "num_tokens": 76246039.0,
      "step": 34805
    },
    {
      "entropy": 5.672507286071777,
      "epoch": 3.4547439460103213,
      "grad_norm": 1.078125,
      "learning_rate": 0.000384052207176459,
      "loss": 5.0753,
      "mean_token_accuracy": 0.1878259927034378,
      "num_tokens": 76258386.0,
      "step": 34810
    },
    {
      "entropy": 5.630110549926758,
      "epoch": 3.4552401746724892,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003840212141252438,
      "loss": 4.9778,
      "mean_token_accuracy": 0.19308388382196426,
      "num_tokens": 76270034.0,
      "step": 34815
    },
    {
      "entropy": 5.737409830093384,
      "epoch": 3.4557364033346567,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038399021837053754,
      "loss": 5.0603,
      "mean_token_accuracy": 0.1814594656229019,
      "num_tokens": 76279869.0,
      "step": 34820
    },
    {
      "entropy": 5.5985430717468265,
      "epoch": 3.456232631996824,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003839592199131088,
      "loss": 4.9058,
      "mean_token_accuracy": 0.19441583305597304,
      "num_tokens": 76289195.0,
      "step": 34825
    },
    {
      "entropy": 5.614547204971314,
      "epoch": 3.4567288606589917,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038392821875372603,
      "loss": 4.9244,
      "mean_token_accuracy": 0.19734974652528764,
      "num_tokens": 76300082.0,
      "step": 34830
    },
    {
      "entropy": 5.643660211563111,
      "epoch": 3.457225089321159,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003838972148931582,
      "loss": 4.8962,
      "mean_token_accuracy": 0.19128705859184264,
      "num_tokens": 76311415.0,
      "step": 34835
    },
    {
      "entropy": 5.653623104095459,
      "epoch": 3.4577213179833266,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003838662083321741,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18241975903511048,
      "num_tokens": 76321719.0,
      "step": 34840
    },
    {
      "entropy": 5.599065923690796,
      "epoch": 3.458217546645494,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038383519907154263,
      "loss": 4.9112,
      "mean_token_accuracy": 0.18673837929964066,
      "num_tokens": 76332852.0,
      "step": 34845
    },
    {
      "entropy": 5.659399938583374,
      "epoch": 3.458713775307662,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038380418711203275,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18223463445901872,
      "num_tokens": 76345047.0,
      "step": 34850
    },
    {
      "entropy": 5.662752294540406,
      "epoch": 3.4592100039698295,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038377317245441345,
      "loss": 5.0055,
      "mean_token_accuracy": 0.19196300953626633,
      "num_tokens": 76357950.0,
      "step": 34855
    },
    {
      "entropy": 5.680264616012574,
      "epoch": 3.459706232631997,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038374215509945386,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18905241191387176,
      "num_tokens": 76368042.0,
      "step": 34860
    },
    {
      "entropy": 5.711025190353394,
      "epoch": 3.4602024612941644,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003837111350479232,
      "loss": 5.0008,
      "mean_token_accuracy": 0.1874355599284172,
      "num_tokens": 76379448.0,
      "step": 34865
    },
    {
      "entropy": 5.762305068969726,
      "epoch": 3.460698689956332,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038368011230059055,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1862009808421135,
      "num_tokens": 76391040.0,
      "step": 34870
    },
    {
      "entropy": 5.6910929679870605,
      "epoch": 3.4611949186184994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038364908685822543,
      "loss": 4.9478,
      "mean_token_accuracy": 0.19932027906179428,
      "num_tokens": 76402345.0,
      "step": 34875
    },
    {
      "entropy": 5.654553318023682,
      "epoch": 3.461691147280667,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038361805872159693,
      "loss": 4.9643,
      "mean_token_accuracy": 0.1880474105477333,
      "num_tokens": 76414072.0,
      "step": 34880
    },
    {
      "entropy": 5.61743197441101,
      "epoch": 3.4621873759428343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003835870278914747,
      "loss": 4.9603,
      "mean_token_accuracy": 0.19455018788576126,
      "num_tokens": 76424207.0,
      "step": 34885
    },
    {
      "entropy": 5.70203857421875,
      "epoch": 3.462683604605002,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003835559943686282,
      "loss": 5.0737,
      "mean_token_accuracy": 0.18433639854192735,
      "num_tokens": 76434683.0,
      "step": 34890
    },
    {
      "entropy": 5.700143766403198,
      "epoch": 3.4631798332671693,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003835249581538269,
      "loss": 5.0136,
      "mean_token_accuracy": 0.18995137363672257,
      "num_tokens": 76445687.0,
      "step": 34895
    },
    {
      "entropy": 5.7290223121643065,
      "epoch": 3.463676061929337,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038349391924784055,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18631723672151565,
      "num_tokens": 76455809.0,
      "step": 34900
    },
    {
      "entropy": 5.664031744003296,
      "epoch": 3.4641722905915047,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038346287765143866,
      "loss": 5.032,
      "mean_token_accuracy": 0.18185776174068452,
      "num_tokens": 76465584.0,
      "step": 34905
    },
    {
      "entropy": 5.6880247592926025,
      "epoch": 3.464668519253672,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038343183336539123,
      "loss": 4.9845,
      "mean_token_accuracy": 0.19716692864894866,
      "num_tokens": 76476353.0,
      "step": 34910
    },
    {
      "entropy": 5.703866910934448,
      "epoch": 3.4651647479158396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000383400786390468,
      "loss": 5.0158,
      "mean_token_accuracy": 0.1872249960899353,
      "num_tokens": 76487038.0,
      "step": 34915
    },
    {
      "entropy": 5.654179096221924,
      "epoch": 3.465660976578007,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038336973672743876,
      "loss": 4.9594,
      "mean_token_accuracy": 0.185035540163517,
      "num_tokens": 76497803.0,
      "step": 34920
    },
    {
      "entropy": 5.6730669975280765,
      "epoch": 3.4661572052401746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038333868437707367,
      "loss": 4.9704,
      "mean_token_accuracy": 0.1924333840608597,
      "num_tokens": 76509621.0,
      "step": 34925
    },
    {
      "entropy": 5.616334247589111,
      "epoch": 3.466653433902342,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003833076293401425,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18621351271867753,
      "num_tokens": 76520213.0,
      "step": 34930
    },
    {
      "entropy": 5.727531671524048,
      "epoch": 3.46714966256451,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003832765716174156,
      "loss": 5.0848,
      "mean_token_accuracy": 0.17799449265003203,
      "num_tokens": 76530591.0,
      "step": 34935
    },
    {
      "entropy": 5.610144710540771,
      "epoch": 3.4676458912266774,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003832455112096631,
      "loss": 4.9296,
      "mean_token_accuracy": 0.18928580582141877,
      "num_tokens": 76541400.0,
      "step": 34940
    },
    {
      "entropy": 5.6556459903717045,
      "epoch": 3.468142119888845,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003832144481176551,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18639102578163147,
      "num_tokens": 76551983.0,
      "step": 34945
    },
    {
      "entropy": 5.6564857006073,
      "epoch": 3.4686383485510124,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038318338234216194,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18568406850099564,
      "num_tokens": 76563918.0,
      "step": 34950
    },
    {
      "entropy": 5.636260938644409,
      "epoch": 3.46913457721318,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038315231388395403,
      "loss": 5.0146,
      "mean_token_accuracy": 0.1865580976009369,
      "num_tokens": 76575339.0,
      "step": 34955
    },
    {
      "entropy": 5.696403932571411,
      "epoch": 3.4696308058753473,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003831212427438018,
      "loss": 4.932,
      "mean_token_accuracy": 0.1981727138161659,
      "num_tokens": 76585817.0,
      "step": 34960
    },
    {
      "entropy": 5.635676670074463,
      "epoch": 3.470127034537515,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003830901689224757,
      "loss": 4.9339,
      "mean_token_accuracy": 0.19274300932884217,
      "num_tokens": 76596818.0,
      "step": 34965
    },
    {
      "entropy": 5.642588901519775,
      "epoch": 3.4706232631996823,
      "grad_norm": 1.125,
      "learning_rate": 0.00038305909242074645,
      "loss": 4.966,
      "mean_token_accuracy": 0.1925506830215454,
      "num_tokens": 76608205.0,
      "step": 34970
    },
    {
      "entropy": 5.621747732162476,
      "epoch": 3.4711194918618498,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003830280132393844,
      "loss": 4.9002,
      "mean_token_accuracy": 0.19251504838466643,
      "num_tokens": 76618803.0,
      "step": 34975
    },
    {
      "entropy": 5.716272687911987,
      "epoch": 3.4716157205240172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038299693137916046,
      "loss": 5.0166,
      "mean_token_accuracy": 0.19106344729661942,
      "num_tokens": 76630197.0,
      "step": 34980
    },
    {
      "entropy": 5.672535943984985,
      "epoch": 3.472111949186185,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038296584684084536,
      "loss": 5.0713,
      "mean_token_accuracy": 0.18580393493175507,
      "num_tokens": 76641544.0,
      "step": 34985
    },
    {
      "entropy": 5.705133295059204,
      "epoch": 3.4726081778483526,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003829347596252099,
      "loss": 5.055,
      "mean_token_accuracy": 0.18247182816267013,
      "num_tokens": 76653746.0,
      "step": 34990
    },
    {
      "entropy": 5.665880966186523,
      "epoch": 3.47310440651052,
      "grad_norm": 1.0,
      "learning_rate": 0.00038290366973302493,
      "loss": 4.9246,
      "mean_token_accuracy": 0.19475600570440293,
      "num_tokens": 76665887.0,
      "step": 34995
    },
    {
      "entropy": 5.684539556503296,
      "epoch": 3.4736006351726876,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038287257716506154,
      "loss": 5.0136,
      "mean_token_accuracy": 0.18317438811063766,
      "num_tokens": 76676748.0,
      "step": 35000
    },
    {
      "entropy": 5.685671329498291,
      "epoch": 3.474096863834855,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038284148192209064,
      "loss": 5.056,
      "mean_token_accuracy": 0.18294815719127655,
      "num_tokens": 76689392.0,
      "step": 35005
    },
    {
      "entropy": 5.585976552963257,
      "epoch": 3.4745930924970225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038281038400488326,
      "loss": 4.9129,
      "mean_token_accuracy": 0.1885753780603409,
      "num_tokens": 76700342.0,
      "step": 35010
    },
    {
      "entropy": 5.720550918579102,
      "epoch": 3.47508932115919,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038277928341421086,
      "loss": 5.0596,
      "mean_token_accuracy": 0.18639056831598283,
      "num_tokens": 76710899.0,
      "step": 35015
    },
    {
      "entropy": 5.664621639251709,
      "epoch": 3.475585549821358,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003827481801508444,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18744442164897918,
      "num_tokens": 76722198.0,
      "step": 35020
    },
    {
      "entropy": 5.760924673080444,
      "epoch": 3.4760817784835254,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038271707421555517,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17639027386903763,
      "num_tokens": 76733288.0,
      "step": 35025
    },
    {
      "entropy": 5.641990995407104,
      "epoch": 3.476578007145693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003826859656091147,
      "loss": 4.9418,
      "mean_token_accuracy": 0.19596228152513503,
      "num_tokens": 76744204.0,
      "step": 35030
    },
    {
      "entropy": 5.638422441482544,
      "epoch": 3.4770742358078603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003826548543322943,
      "loss": 4.9704,
      "mean_token_accuracy": 0.18874475210905076,
      "num_tokens": 76753833.0,
      "step": 35035
    },
    {
      "entropy": 5.6490222930908205,
      "epoch": 3.477570464470028,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003826237403858654,
      "loss": 5.0628,
      "mean_token_accuracy": 0.1855490878224373,
      "num_tokens": 76765207.0,
      "step": 35040
    },
    {
      "entropy": 5.660787200927734,
      "epoch": 3.4780666931321953,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003825926237705998,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18793827742338182,
      "num_tokens": 76776559.0,
      "step": 35045
    },
    {
      "entropy": 5.68814058303833,
      "epoch": 3.4785629217943628,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000382561504487269,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18293771892786026,
      "num_tokens": 76788152.0,
      "step": 35050
    },
    {
      "entropy": 5.649702310562134,
      "epoch": 3.4790591504565302,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038253038253664446,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1848691627383232,
      "num_tokens": 76800825.0,
      "step": 35055
    },
    {
      "entropy": 5.662819051742554,
      "epoch": 3.4795553791186977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038249925791949827,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18653478473424911,
      "num_tokens": 76812509.0,
      "step": 35060
    },
    {
      "entropy": 5.7764444828033445,
      "epoch": 3.480051607780865,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003824681306366021,
      "loss": 5.1217,
      "mean_token_accuracy": 0.18294283747673035,
      "num_tokens": 76822771.0,
      "step": 35065
    },
    {
      "entropy": 5.678350496292114,
      "epoch": 3.480547836443033,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038243700068872784,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1857887625694275,
      "num_tokens": 76834449.0,
      "step": 35070
    },
    {
      "entropy": 5.6763591289520265,
      "epoch": 3.4810440651052006,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038240586807664745,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18417733460664748,
      "num_tokens": 76844415.0,
      "step": 35075
    },
    {
      "entropy": 5.656747055053711,
      "epoch": 3.481540293767368,
      "grad_norm": 1.140625,
      "learning_rate": 0.000382374732801133,
      "loss": 4.958,
      "mean_token_accuracy": 0.19533224105834962,
      "num_tokens": 76854547.0,
      "step": 35080
    },
    {
      "entropy": 5.615177726745605,
      "epoch": 3.4820365224295355,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038234359486295663,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1831950694322586,
      "num_tokens": 76865120.0,
      "step": 35085
    },
    {
      "entropy": 5.60591950416565,
      "epoch": 3.482532751091703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003823124542628904,
      "loss": 4.8692,
      "mean_token_accuracy": 0.20030972957611085,
      "num_tokens": 76877431.0,
      "step": 35090
    },
    {
      "entropy": 5.625038194656372,
      "epoch": 3.4830289797538705,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038228131100170637,
      "loss": 4.9466,
      "mean_token_accuracy": 0.19453664273023605,
      "num_tokens": 76888476.0,
      "step": 35095
    },
    {
      "entropy": 5.617580413818359,
      "epoch": 3.483525208416038,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038225016508017717,
      "loss": 4.9361,
      "mean_token_accuracy": 0.18898107260465621,
      "num_tokens": 76899971.0,
      "step": 35100
    },
    {
      "entropy": 5.610208225250244,
      "epoch": 3.484021437078206,
      "grad_norm": 1.234375,
      "learning_rate": 0.000382219016499075,
      "loss": 4.9666,
      "mean_token_accuracy": 0.1925977110862732,
      "num_tokens": 76910983.0,
      "step": 35105
    },
    {
      "entropy": 5.701343297958374,
      "epoch": 3.4845176657403734,
      "grad_norm": 1.125,
      "learning_rate": 0.00038218786525917207,
      "loss": 5.0239,
      "mean_token_accuracy": 0.18727829456329345,
      "num_tokens": 76921706.0,
      "step": 35110
    },
    {
      "entropy": 5.7685235977172855,
      "epoch": 3.485013894402541,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003821567113612412,
      "loss": 5.1481,
      "mean_token_accuracy": 0.18038144260644912,
      "num_tokens": 76932226.0,
      "step": 35115
    },
    {
      "entropy": 5.6482545852661135,
      "epoch": 3.4855101230647083,
      "grad_norm": 1.125,
      "learning_rate": 0.00038212555480605475,
      "loss": 4.9294,
      "mean_token_accuracy": 0.18925298303365706,
      "num_tokens": 76943880.0,
      "step": 35120
    },
    {
      "entropy": 5.66555438041687,
      "epoch": 3.486006351726876,
      "grad_norm": 1.25,
      "learning_rate": 0.00038209439559438535,
      "loss": 5.0154,
      "mean_token_accuracy": 0.19491072744131088,
      "num_tokens": 76955482.0,
      "step": 35125
    },
    {
      "entropy": 5.6157684326171875,
      "epoch": 3.4865025803890433,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003820632337270057,
      "loss": 4.9877,
      "mean_token_accuracy": 0.19210824370384216,
      "num_tokens": 76964946.0,
      "step": 35130
    },
    {
      "entropy": 5.6502197742462155,
      "epoch": 3.4869988090512107,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038203206920468867,
      "loss": 4.9828,
      "mean_token_accuracy": 0.18497546315193175,
      "num_tokens": 76974462.0,
      "step": 35135
    },
    {
      "entropy": 5.679144954681396,
      "epoch": 3.487495037713378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038200090202820686,
      "loss": 5.0001,
      "mean_token_accuracy": 0.1815027818083763,
      "num_tokens": 76985167.0,
      "step": 35140
    },
    {
      "entropy": 5.669968843460083,
      "epoch": 3.4879912663755457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038196973219833323,
      "loss": 5.0773,
      "mean_token_accuracy": 0.1773666337132454,
      "num_tokens": 76996510.0,
      "step": 35145
    },
    {
      "entropy": 5.642748308181763,
      "epoch": 3.488487495037713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038193855971584086,
      "loss": 4.9368,
      "mean_token_accuracy": 0.1943150356411934,
      "num_tokens": 77007261.0,
      "step": 35150
    },
    {
      "entropy": 5.67383599281311,
      "epoch": 3.488983723699881,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038190738458150256,
      "loss": 4.9718,
      "mean_token_accuracy": 0.1857769951224327,
      "num_tokens": 77018700.0,
      "step": 35155
    },
    {
      "entropy": 5.673634767532349,
      "epoch": 3.4894799523620486,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038187620679609144,
      "loss": 4.9658,
      "mean_token_accuracy": 0.1960369035601616,
      "num_tokens": 77029633.0,
      "step": 35160
    },
    {
      "entropy": 5.705674362182617,
      "epoch": 3.489976181024216,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038184502636038074,
      "loss": 5.129,
      "mean_token_accuracy": 0.18634118884801865,
      "num_tokens": 77043520.0,
      "step": 35165
    },
    {
      "entropy": 5.673147296905517,
      "epoch": 3.4904724096863835,
      "grad_norm": 1.125,
      "learning_rate": 0.0003818138432751436,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18183874040842057,
      "num_tokens": 77054903.0,
      "step": 35170
    },
    {
      "entropy": 5.687815999984741,
      "epoch": 3.490968638348551,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003817826575411533,
      "loss": 4.9633,
      "mean_token_accuracy": 0.1884274512529373,
      "num_tokens": 77066575.0,
      "step": 35175
    },
    {
      "entropy": 5.724591779708862,
      "epoch": 3.4914648670107185,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038175146915918323,
      "loss": 5.0332,
      "mean_token_accuracy": 0.18473239094018937,
      "num_tokens": 77077189.0,
      "step": 35180
    },
    {
      "entropy": 5.675984191894531,
      "epoch": 3.491961095672886,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038172027813000676,
      "loss": 5.0434,
      "mean_token_accuracy": 0.18411727398633956,
      "num_tokens": 77088512.0,
      "step": 35185
    },
    {
      "entropy": 5.695505142211914,
      "epoch": 3.4924573243350534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038168908445439725,
      "loss": 5.1053,
      "mean_token_accuracy": 0.18508398979902269,
      "num_tokens": 77099801.0,
      "step": 35190
    },
    {
      "entropy": 5.729029178619385,
      "epoch": 3.4929535529972213,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038165788813312844,
      "loss": 5.0909,
      "mean_token_accuracy": 0.18258563429117203,
      "num_tokens": 77111265.0,
      "step": 35195
    },
    {
      "entropy": 5.593093156814575,
      "epoch": 3.493449781659389,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038162668916697385,
      "loss": 4.8268,
      "mean_token_accuracy": 0.19720659404993057,
      "num_tokens": 77121809.0,
      "step": 35200
    },
    {
      "entropy": 5.6883961200714115,
      "epoch": 3.4939460103215563,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038159548755670707,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1834118440747261,
      "num_tokens": 77132401.0,
      "step": 35205
    },
    {
      "entropy": 5.652073669433594,
      "epoch": 3.4944422389837237,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038156428330310193,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18878669440746307,
      "num_tokens": 77144058.0,
      "step": 35210
    },
    {
      "entropy": 5.665723037719727,
      "epoch": 3.4949384676458912,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003815330764069322,
      "loss": 4.9162,
      "mean_token_accuracy": 0.195392644405365,
      "num_tokens": 77156130.0,
      "step": 35215
    },
    {
      "entropy": 5.670557260513306,
      "epoch": 3.4954346963080587,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003815018668689717,
      "loss": 4.975,
      "mean_token_accuracy": 0.18573682010173798,
      "num_tokens": 77166597.0,
      "step": 35220
    },
    {
      "entropy": 5.662858867645264,
      "epoch": 3.495930924970226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038147065468999435,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18745617121458052,
      "num_tokens": 77177015.0,
      "step": 35225
    },
    {
      "entropy": 5.666221046447754,
      "epoch": 3.4964271536323936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038143943987077426,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18627085089683532,
      "num_tokens": 77188686.0,
      "step": 35230
    },
    {
      "entropy": 5.664900016784668,
      "epoch": 3.496923382294561,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003814082224120854,
      "loss": 5.027,
      "mean_token_accuracy": 0.18121143579483032,
      "num_tokens": 77198806.0,
      "step": 35235
    },
    {
      "entropy": 5.675192165374756,
      "epoch": 3.497419610956729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003813770023147019,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18584085404872894,
      "num_tokens": 77210515.0,
      "step": 35240
    },
    {
      "entropy": 5.64503664970398,
      "epoch": 3.4979158396188965,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038134577957939795,
      "loss": 4.9236,
      "mean_token_accuracy": 0.18834201544523238,
      "num_tokens": 77221661.0,
      "step": 35245
    },
    {
      "entropy": 5.643261575698853,
      "epoch": 3.498412068281064,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038131455420694786,
      "loss": 4.95,
      "mean_token_accuracy": 0.1930951863527298,
      "num_tokens": 77232512.0,
      "step": 35250
    },
    {
      "entropy": 5.68066668510437,
      "epoch": 3.4989082969432315,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003812833261981259,
      "loss": 5.0485,
      "mean_token_accuracy": 0.18616009056568145,
      "num_tokens": 77242821.0,
      "step": 35255
    },
    {
      "entropy": 5.663513040542602,
      "epoch": 3.499404525605399,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038125209555370644,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18886077851057054,
      "num_tokens": 77253644.0,
      "step": 35260
    },
    {
      "entropy": 5.678481006622315,
      "epoch": 3.4999007542675664,
      "grad_norm": 1.109375,
      "learning_rate": 0.000381220862274464,
      "loss": 4.9634,
      "mean_token_accuracy": 0.18874461948871613,
      "num_tokens": 77264505.0,
      "step": 35265
    },
    {
      "entropy": 5.671102714538574,
      "epoch": 3.500396982929734,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038118962636117304,
      "loss": 4.9603,
      "mean_token_accuracy": 0.1905036300420761,
      "num_tokens": 77275956.0,
      "step": 35270
    },
    {
      "entropy": 5.688012170791626,
      "epoch": 3.500893211591902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038115838781460816,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18553890585899352,
      "num_tokens": 77286766.0,
      "step": 35275
    },
    {
      "entropy": 5.6668964385986325,
      "epoch": 3.5013894402540693,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000381127146635544,
      "loss": 5.063,
      "mean_token_accuracy": 0.18404539078474044,
      "num_tokens": 77297619.0,
      "step": 35280
    },
    {
      "entropy": 5.741467905044556,
      "epoch": 3.5018856689162368,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003810959028247554,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1781470075249672,
      "num_tokens": 77309362.0,
      "step": 35285
    },
    {
      "entropy": 5.649155187606811,
      "epoch": 3.5023818975784042,
      "grad_norm": 1.125,
      "learning_rate": 0.0003810646563830169,
      "loss": 4.9273,
      "mean_token_accuracy": 0.19048328846693038,
      "num_tokens": 77320385.0,
      "step": 35290
    },
    {
      "entropy": 5.612674713134766,
      "epoch": 3.5028781262405717,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038103340731110354,
      "loss": 4.9459,
      "mean_token_accuracy": 0.18826918303966522,
      "num_tokens": 77330584.0,
      "step": 35295
    },
    {
      "entropy": 5.6254791736602785,
      "epoch": 3.503374354902739,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038100215560979016,
      "loss": 4.9755,
      "mean_token_accuracy": 0.19376160055398942,
      "num_tokens": 77340867.0,
      "step": 35300
    },
    {
      "entropy": 5.623582887649536,
      "epoch": 3.5038705835649067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038097090127985165,
      "loss": 4.9296,
      "mean_token_accuracy": 0.18687749654054642,
      "num_tokens": 77351570.0,
      "step": 35305
    },
    {
      "entropy": 5.61955885887146,
      "epoch": 3.504366812227074,
      "grad_norm": 1.125,
      "learning_rate": 0.00038093964432206316,
      "loss": 4.9155,
      "mean_token_accuracy": 0.19252712577581405,
      "num_tokens": 77361433.0,
      "step": 35310
    },
    {
      "entropy": 5.600323438644409,
      "epoch": 3.5048630408892416,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003809083847371999,
      "loss": 4.9717,
      "mean_token_accuracy": 0.19608704447746278,
      "num_tokens": 77373503.0,
      "step": 35315
    },
    {
      "entropy": 5.686743307113647,
      "epoch": 3.505359269551409,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003808771225260368,
      "loss": 5.0887,
      "mean_token_accuracy": 0.18411872982978822,
      "num_tokens": 77385590.0,
      "step": 35320
    },
    {
      "entropy": 5.646589088439941,
      "epoch": 3.5058554982135766,
      "grad_norm": 1.375,
      "learning_rate": 0.00038084585768934917,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18798412531614303,
      "num_tokens": 77396328.0,
      "step": 35325
    },
    {
      "entropy": 5.738536739349366,
      "epoch": 3.5063517268757445,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038081459022791243,
      "loss": 5.0826,
      "mean_token_accuracy": 0.18176025301218032,
      "num_tokens": 77407736.0,
      "step": 35330
    },
    {
      "entropy": 5.654345512390137,
      "epoch": 3.506847955537912,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038078332014250185,
      "loss": 4.9138,
      "mean_token_accuracy": 0.19224769324064256,
      "num_tokens": 77419194.0,
      "step": 35335
    },
    {
      "entropy": 5.672034740447998,
      "epoch": 3.5073441842000794,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003807520474338928,
      "loss": 4.9463,
      "mean_token_accuracy": 0.19079162180423737,
      "num_tokens": 77430153.0,
      "step": 35340
    },
    {
      "entropy": 5.623704481124878,
      "epoch": 3.507840412862247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003807207721028609,
      "loss": 4.975,
      "mean_token_accuracy": 0.19079503864049913,
      "num_tokens": 77442021.0,
      "step": 35345
    },
    {
      "entropy": 5.672588443756103,
      "epoch": 3.5083366415244144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003806894941501817,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1849016383290291,
      "num_tokens": 77453789.0,
      "step": 35350
    },
    {
      "entropy": 5.715796566009521,
      "epoch": 3.508832870186582,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003806582135766307,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18777197301387788,
      "num_tokens": 77465368.0,
      "step": 35355
    },
    {
      "entropy": 5.626656866073608,
      "epoch": 3.5093290988487498,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038062693038298367,
      "loss": 4.9735,
      "mean_token_accuracy": 0.18311306834220886,
      "num_tokens": 77476667.0,
      "step": 35360
    },
    {
      "entropy": 5.68265814781189,
      "epoch": 3.5098253275109172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003805956445700165,
      "loss": 4.9776,
      "mean_token_accuracy": 0.1906988337635994,
      "num_tokens": 77486216.0,
      "step": 35365
    },
    {
      "entropy": 5.700672292709351,
      "epoch": 3.5103215561730847,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003805643561385048,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17911529988050462,
      "num_tokens": 77496529.0,
      "step": 35370
    },
    {
      "entropy": 5.685584449768067,
      "epoch": 3.510817784835252,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038053306508922455,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17919226586818696,
      "num_tokens": 77507412.0,
      "step": 35375
    },
    {
      "entropy": 5.692639112472534,
      "epoch": 3.5113140134974197,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038050177142295166,
      "loss": 5.0415,
      "mean_token_accuracy": 0.18380393087863922,
      "num_tokens": 77519124.0,
      "step": 35380
    },
    {
      "entropy": 5.653730010986328,
      "epoch": 3.511810242159587,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038047047514046223,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18566142469644548,
      "num_tokens": 77528815.0,
      "step": 35385
    },
    {
      "entropy": 5.615491104125977,
      "epoch": 3.5123064708217546,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003804391762425322,
      "loss": 4.9572,
      "mean_token_accuracy": 0.19170129150152207,
      "num_tokens": 77538999.0,
      "step": 35390
    },
    {
      "entropy": 5.6812121868133545,
      "epoch": 3.512802699483922,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038040787472993793,
      "loss": 4.9602,
      "mean_token_accuracy": 0.19275042563676834,
      "num_tokens": 77549392.0,
      "step": 35395
    },
    {
      "entropy": 5.665266513824463,
      "epoch": 3.5132989281460896,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003803765706034554,
      "loss": 4.9684,
      "mean_token_accuracy": 0.19399045556783676,
      "num_tokens": 77561901.0,
      "step": 35400
    },
    {
      "entropy": 5.678743505477906,
      "epoch": 3.513795156808257,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038034526386386097,
      "loss": 5.0436,
      "mean_token_accuracy": 0.1864382803440094,
      "num_tokens": 77572324.0,
      "step": 35405
    },
    {
      "entropy": 5.680277347564697,
      "epoch": 3.5142913854704245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038031395451193105,
      "loss": 5.0535,
      "mean_token_accuracy": 0.18753542751073837,
      "num_tokens": 77584022.0,
      "step": 35410
    },
    {
      "entropy": 5.645429468154907,
      "epoch": 3.5147876141325924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038028264254844204,
      "loss": 5.0127,
      "mean_token_accuracy": 0.19215748459100723,
      "num_tokens": 77595974.0,
      "step": 35415
    },
    {
      "entropy": 5.687426614761352,
      "epoch": 3.51528384279476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038025132797417024,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18722105920314788,
      "num_tokens": 77608092.0,
      "step": 35420
    },
    {
      "entropy": 5.676814270019531,
      "epoch": 3.5157800714569274,
      "grad_norm": 1.125,
      "learning_rate": 0.0003802200107898924,
      "loss": 4.9531,
      "mean_token_accuracy": 0.1901462122797966,
      "num_tokens": 77620763.0,
      "step": 35425
    },
    {
      "entropy": 5.641657781600952,
      "epoch": 3.516276300119095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038018869099638506,
      "loss": 4.9689,
      "mean_token_accuracy": 0.18251008540391922,
      "num_tokens": 77631082.0,
      "step": 35430
    },
    {
      "entropy": 5.6267482280731205,
      "epoch": 3.5167725287812623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003801573685944247,
      "loss": 4.9553,
      "mean_token_accuracy": 0.18716302663087844,
      "num_tokens": 77642322.0,
      "step": 35435
    },
    {
      "entropy": 5.6801306247711185,
      "epoch": 3.51726875744343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038012604358478823,
      "loss": 4.9884,
      "mean_token_accuracy": 0.18726988434791564,
      "num_tokens": 77655596.0,
      "step": 35440
    },
    {
      "entropy": 5.642403221130371,
      "epoch": 3.5177649861055977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038009471596825244,
      "loss": 5.0243,
      "mean_token_accuracy": 0.1863602340221405,
      "num_tokens": 77666955.0,
      "step": 35445
    },
    {
      "entropy": 5.647271966934204,
      "epoch": 3.518261214767765,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003800633857455942,
      "loss": 4.9409,
      "mean_token_accuracy": 0.19510759860277177,
      "num_tokens": 77677736.0,
      "step": 35450
    },
    {
      "entropy": 5.619517087936401,
      "epoch": 3.5187574434299327,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003800320529175903,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18905575722455978,
      "num_tokens": 77689086.0,
      "step": 35455
    },
    {
      "entropy": 5.660412931442261,
      "epoch": 3.5192536720921,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038000071748501787,
      "loss": 5.0065,
      "mean_token_accuracy": 0.18439997136592864,
      "num_tokens": 77700827.0,
      "step": 35460
    },
    {
      "entropy": 5.633200740814209,
      "epoch": 3.5197499007542676,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037996937944865383,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18681274354457855,
      "num_tokens": 77710490.0,
      "step": 35465
    },
    {
      "entropy": 5.720481824874878,
      "epoch": 3.520246129416435,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003799380388092754,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18289458006620407,
      "num_tokens": 77721125.0,
      "step": 35470
    },
    {
      "entropy": 5.720034742355347,
      "epoch": 3.5207423580786026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003799066955676598,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18990846574306489,
      "num_tokens": 77731684.0,
      "step": 35475
    },
    {
      "entropy": 5.718450546264648,
      "epoch": 3.52123858674077,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003798753497245841,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19343065172433854,
      "num_tokens": 77741010.0,
      "step": 35480
    },
    {
      "entropy": 5.627637577056885,
      "epoch": 3.5217348154029375,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037984400128082576,
      "loss": 4.9671,
      "mean_token_accuracy": 0.19198305159807205,
      "num_tokens": 77751451.0,
      "step": 35485
    },
    {
      "entropy": 5.613250684738159,
      "epoch": 3.522231044065105,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037981265023716217,
      "loss": 4.9494,
      "mean_token_accuracy": 0.18835102021694183,
      "num_tokens": 77762152.0,
      "step": 35490
    },
    {
      "entropy": 5.711161088943482,
      "epoch": 3.5227272727272725,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037978129659437063,
      "loss": 5.0508,
      "mean_token_accuracy": 0.181399866938591,
      "num_tokens": 77773383.0,
      "step": 35495
    },
    {
      "entropy": 5.686159372329712,
      "epoch": 3.5232235013894404,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003797499403532287,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18750853538513185,
      "num_tokens": 77784773.0,
      "step": 35500
    },
    {
      "entropy": 5.682394742965698,
      "epoch": 3.523719730051608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037971858151451403,
      "loss": 5.051,
      "mean_token_accuracy": 0.18342022001743316,
      "num_tokens": 77796878.0,
      "step": 35505
    },
    {
      "entropy": 5.66561861038208,
      "epoch": 3.5242159587137754,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037968722007900415,
      "loss": 5.0191,
      "mean_token_accuracy": 0.1841773957014084,
      "num_tokens": 77808588.0,
      "step": 35510
    },
    {
      "entropy": 5.592322254180909,
      "epoch": 3.524712187375943,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037965585604747684,
      "loss": 4.8871,
      "mean_token_accuracy": 0.20421095043420792,
      "num_tokens": 77818702.0,
      "step": 35515
    },
    {
      "entropy": 5.650558423995972,
      "epoch": 3.5252084160381103,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003796244894207097,
      "loss": 4.9693,
      "mean_token_accuracy": 0.19090209454298018,
      "num_tokens": 77829673.0,
      "step": 35520
    },
    {
      "entropy": 5.701386451721191,
      "epoch": 3.525704644700278,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037959312019948075,
      "loss": 4.991,
      "mean_token_accuracy": 0.18615727126598358,
      "num_tokens": 77840350.0,
      "step": 35525
    },
    {
      "entropy": 5.618369102478027,
      "epoch": 3.5262008733624453,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003795617483845678,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18820865005254744,
      "num_tokens": 77853228.0,
      "step": 35530
    },
    {
      "entropy": 5.653105020523071,
      "epoch": 3.526697102024613,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037953037397674876,
      "loss": 4.9781,
      "mean_token_accuracy": 0.19583080261945723,
      "num_tokens": 77864564.0,
      "step": 35535
    },
    {
      "entropy": 5.689950847625733,
      "epoch": 3.5271933306867806,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003794989969768017,
      "loss": 4.986,
      "mean_token_accuracy": 0.18952905386686325,
      "num_tokens": 77875449.0,
      "step": 35540
    },
    {
      "entropy": 5.7288624286651615,
      "epoch": 3.527689559348948,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003794676173855048,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18539163023233413,
      "num_tokens": 77885809.0,
      "step": 35545
    },
    {
      "entropy": 5.611064386367798,
      "epoch": 3.5281857880111156,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037943623520363595,
      "loss": 4.9639,
      "mean_token_accuracy": 0.19123158603906631,
      "num_tokens": 77896414.0,
      "step": 35550
    },
    {
      "entropy": 5.56978554725647,
      "epoch": 3.528682016673283,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003794048504319736,
      "loss": 4.9445,
      "mean_token_accuracy": 0.18992667496204377,
      "num_tokens": 77907784.0,
      "step": 35555
    },
    {
      "entropy": 5.616452360153199,
      "epoch": 3.5291782453354505,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037937346307129584,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18442287147045136,
      "num_tokens": 77919137.0,
      "step": 35560
    },
    {
      "entropy": 5.689523649215698,
      "epoch": 3.529674473997618,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037934207312238116,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1878311038017273,
      "num_tokens": 77929607.0,
      "step": 35565
    },
    {
      "entropy": 5.641639566421508,
      "epoch": 3.5301707026597855,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037931068058600785,
      "loss": 4.9337,
      "mean_token_accuracy": 0.19365223944187165,
      "num_tokens": 77939477.0,
      "step": 35570
    },
    {
      "entropy": 5.669222497940064,
      "epoch": 3.530666931321953,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037927928546295454,
      "loss": 4.9506,
      "mean_token_accuracy": 0.1916508123278618,
      "num_tokens": 77950228.0,
      "step": 35575
    },
    {
      "entropy": 5.674670839309693,
      "epoch": 3.5311631599841204,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003792478877539995,
      "loss": 4.9551,
      "mean_token_accuracy": 0.19389386624097824,
      "num_tokens": 77961635.0,
      "step": 35580
    },
    {
      "entropy": 5.630092191696167,
      "epoch": 3.5316593886462884,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003792164874599216,
      "loss": 4.9185,
      "mean_token_accuracy": 0.1972474619746208,
      "num_tokens": 77972008.0,
      "step": 35585
    },
    {
      "entropy": 5.647489356994629,
      "epoch": 3.532155617308456,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003791850845814994,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19230015128850936,
      "num_tokens": 77982097.0,
      "step": 35590
    },
    {
      "entropy": 5.682254314422607,
      "epoch": 3.5326518459706233,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037915367911951146,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18115246295928955,
      "num_tokens": 77994284.0,
      "step": 35595
    },
    {
      "entropy": 5.65321569442749,
      "epoch": 3.533148074632791,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037912227107473684,
      "loss": 4.9895,
      "mean_token_accuracy": 0.19199251383543015,
      "num_tokens": 78005327.0,
      "step": 35600
    },
    {
      "entropy": 5.685219955444336,
      "epoch": 3.5336443032949583,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037909086044795427,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18865454494953154,
      "num_tokens": 78016820.0,
      "step": 35605
    },
    {
      "entropy": 5.745446681976318,
      "epoch": 3.5341405319571257,
      "grad_norm": 1.125,
      "learning_rate": 0.00037905944723994257,
      "loss": 5.0113,
      "mean_token_accuracy": 0.1896643176674843,
      "num_tokens": 78027584.0,
      "step": 35610
    },
    {
      "entropy": 5.744402885437012,
      "epoch": 3.534636760619293,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037902803145148085,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17969150096178055,
      "num_tokens": 78038479.0,
      "step": 35615
    },
    {
      "entropy": 5.693875360488891,
      "epoch": 3.535132989281461,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003789966130833481,
      "loss": 5.0725,
      "mean_token_accuracy": 0.1831950157880783,
      "num_tokens": 78051196.0,
      "step": 35620
    },
    {
      "entropy": 5.691845321655274,
      "epoch": 3.5356292179436286,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003789651921363234,
      "loss": 4.9893,
      "mean_token_accuracy": 0.1848585158586502,
      "num_tokens": 78061878.0,
      "step": 35625
    },
    {
      "entropy": 5.663714790344239,
      "epoch": 3.536125446605796,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037893376861118606,
      "loss": 4.9478,
      "mean_token_accuracy": 0.1886444941163063,
      "num_tokens": 78072641.0,
      "step": 35630
    },
    {
      "entropy": 5.736423110961914,
      "epoch": 3.5366216752679636,
      "grad_norm": 1.125,
      "learning_rate": 0.00037890234250871515,
      "loss": 5.02,
      "mean_token_accuracy": 0.19319268614053725,
      "num_tokens": 78083358.0,
      "step": 35635
    },
    {
      "entropy": 5.639775800704956,
      "epoch": 3.537117903930131,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037887091382969,
      "loss": 5.0193,
      "mean_token_accuracy": 0.1857089102268219,
      "num_tokens": 78094761.0,
      "step": 35640
    },
    {
      "entropy": 5.625664091110229,
      "epoch": 3.5376141325922985,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003788394825748901,
      "loss": 4.9206,
      "mean_token_accuracy": 0.18897290378808976,
      "num_tokens": 78104959.0,
      "step": 35645
    },
    {
      "entropy": 5.6914787769317625,
      "epoch": 3.538110361254466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037880804874509467,
      "loss": 5.0183,
      "mean_token_accuracy": 0.18895953595638276,
      "num_tokens": 78115579.0,
      "step": 35650
    },
    {
      "entropy": 5.730977725982666,
      "epoch": 3.5386065899166335,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003787766123410834,
      "loss": 5.08,
      "mean_token_accuracy": 0.18203415274620055,
      "num_tokens": 78126597.0,
      "step": 35655
    },
    {
      "entropy": 5.742032241821289,
      "epoch": 3.539102818578801,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003787451733636357,
      "loss": 5.169,
      "mean_token_accuracy": 0.17743874043226243,
      "num_tokens": 78138454.0,
      "step": 35660
    },
    {
      "entropy": 5.659061050415039,
      "epoch": 3.5395990472409684,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037871373181353135,
      "loss": 4.9681,
      "mean_token_accuracy": 0.1932522878050804,
      "num_tokens": 78149887.0,
      "step": 35665
    },
    {
      "entropy": 5.644956398010254,
      "epoch": 3.5400952759031363,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037868228769154987,
      "loss": 4.9629,
      "mean_token_accuracy": 0.18709340393543245,
      "num_tokens": 78160847.0,
      "step": 35670
    },
    {
      "entropy": 5.6296820640563965,
      "epoch": 3.540591504565304,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003786508409984711,
      "loss": 4.9315,
      "mean_token_accuracy": 0.18946865051984788,
      "num_tokens": 78171683.0,
      "step": 35675
    },
    {
      "entropy": 5.665595531463623,
      "epoch": 3.5410877332274713,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037861939173507484,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19375549256801605,
      "num_tokens": 78181713.0,
      "step": 35680
    },
    {
      "entropy": 5.650196695327759,
      "epoch": 3.5415839618896388,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037858793990214087,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1862523749470711,
      "num_tokens": 78193113.0,
      "step": 35685
    },
    {
      "entropy": 5.71964020729065,
      "epoch": 3.5420801905518062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037855648550044925,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18800357431173326,
      "num_tokens": 78204703.0,
      "step": 35690
    },
    {
      "entropy": 5.694424247741699,
      "epoch": 3.5425764192139737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003785250285307799,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1897359162569046,
      "num_tokens": 78215780.0,
      "step": 35695
    },
    {
      "entropy": 5.624383497238159,
      "epoch": 3.543072647876141,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037849356899391295,
      "loss": 4.8588,
      "mean_token_accuracy": 0.2029893085360527,
      "num_tokens": 78227460.0,
      "step": 35700
    },
    {
      "entropy": 5.6344140529632565,
      "epoch": 3.543568876538309,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003784621068906285,
      "loss": 4.9912,
      "mean_token_accuracy": 0.19934451431035996,
      "num_tokens": 78238268.0,
      "step": 35705
    },
    {
      "entropy": 5.591061687469482,
      "epoch": 3.5440651052004766,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037843064222170677,
      "loss": 4.9078,
      "mean_token_accuracy": 0.1997681051492691,
      "num_tokens": 78249404.0,
      "step": 35710
    },
    {
      "entropy": 5.621793127059936,
      "epoch": 3.544561333862644,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037839917498792787,
      "loss": 5.0174,
      "mean_token_accuracy": 0.1848747655749321,
      "num_tokens": 78261327.0,
      "step": 35715
    },
    {
      "entropy": 5.700201892852784,
      "epoch": 3.5450575625248115,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037836770519007236,
      "loss": 5.0296,
      "mean_token_accuracy": 0.19032408446073532,
      "num_tokens": 78273033.0,
      "step": 35720
    },
    {
      "entropy": 5.635148429870606,
      "epoch": 3.545553791186979,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003783362328289205,
      "loss": 4.887,
      "mean_token_accuracy": 0.19249800145626067,
      "num_tokens": 78283438.0,
      "step": 35725
    },
    {
      "entropy": 5.688898372650146,
      "epoch": 3.5460500198491465,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003783047579052526,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18740324676036835,
      "num_tokens": 78294876.0,
      "step": 35730
    },
    {
      "entropy": 5.6255439758300785,
      "epoch": 3.546546248511314,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037827328041984944,
      "loss": 4.9615,
      "mean_token_accuracy": 0.18833053410053252,
      "num_tokens": 78303967.0,
      "step": 35735
    },
    {
      "entropy": 5.647879934310913,
      "epoch": 3.5470424771734814,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003782418003734914,
      "loss": 4.9809,
      "mean_token_accuracy": 0.1868677854537964,
      "num_tokens": 78314425.0,
      "step": 35740
    },
    {
      "entropy": 5.77304892539978,
      "epoch": 3.547538705835649,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003782103177669592,
      "loss": 5.0903,
      "mean_token_accuracy": 0.18450748771429062,
      "num_tokens": 78324954.0,
      "step": 35745
    },
    {
      "entropy": 5.699878931045532,
      "epoch": 3.5480349344978164,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037817883260103354,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18243056237697602,
      "num_tokens": 78334781.0,
      "step": 35750
    },
    {
      "entropy": 5.629838943481445,
      "epoch": 3.548531163159984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037814734487649516,
      "loss": 4.9217,
      "mean_token_accuracy": 0.19078635573387145,
      "num_tokens": 78347130.0,
      "step": 35755
    },
    {
      "entropy": 5.678336715698242,
      "epoch": 3.5490273918221518,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003781158545941248,
      "loss": 5.041,
      "mean_token_accuracy": 0.18766190558671952,
      "num_tokens": 78357859.0,
      "step": 35760
    },
    {
      "entropy": 5.635000276565552,
      "epoch": 3.5495236204843192,
      "grad_norm": 1.25,
      "learning_rate": 0.00037808436175470344,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18586394637823106,
      "num_tokens": 78368523.0,
      "step": 35765
    },
    {
      "entropy": 5.7097475051879885,
      "epoch": 3.5500198491464867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003780528663590122,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17593529224395751,
      "num_tokens": 78379262.0,
      "step": 35770
    },
    {
      "entropy": 5.662974882125854,
      "epoch": 3.550516077808654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003780213684078318,
      "loss": 4.9287,
      "mean_token_accuracy": 0.19603077322244644,
      "num_tokens": 78389477.0,
      "step": 35775
    },
    {
      "entropy": 5.604390001296997,
      "epoch": 3.5510123064708217,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037798986790194347,
      "loss": 4.9111,
      "mean_token_accuracy": 0.19689782112836837,
      "num_tokens": 78400606.0,
      "step": 35780
    },
    {
      "entropy": 5.645483493804932,
      "epoch": 3.551508535132989,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037795836484212837,
      "loss": 4.959,
      "mean_token_accuracy": 0.1873690277338028,
      "num_tokens": 78411190.0,
      "step": 35785
    },
    {
      "entropy": 5.654002714157104,
      "epoch": 3.552004763795157,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037792685922916763,
      "loss": 4.9966,
      "mean_token_accuracy": 0.1869819939136505,
      "num_tokens": 78421694.0,
      "step": 35790
    },
    {
      "entropy": 5.6575335502624515,
      "epoch": 3.5525009924573245,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037789535106384266,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18892253339290618,
      "num_tokens": 78432414.0,
      "step": 35795
    },
    {
      "entropy": 5.652266788482666,
      "epoch": 3.552997221119492,
      "grad_norm": 1.125,
      "learning_rate": 0.0003778638403469346,
      "loss": 4.93,
      "mean_token_accuracy": 0.1937077447772026,
      "num_tokens": 78443899.0,
      "step": 35800
    },
    {
      "entropy": 5.643211841583252,
      "epoch": 3.5534934497816595,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037783232707922507,
      "loss": 4.9182,
      "mean_token_accuracy": 0.18945260494947433,
      "num_tokens": 78454644.0,
      "step": 35805
    },
    {
      "entropy": 5.625439310073853,
      "epoch": 3.553989678443827,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037780081126149523,
      "loss": 4.9559,
      "mean_token_accuracy": 0.19119685441255568,
      "num_tokens": 78465645.0,
      "step": 35810
    },
    {
      "entropy": 5.611953926086426,
      "epoch": 3.5544859071059944,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037776929289452696,
      "loss": 4.9934,
      "mean_token_accuracy": 0.184048655629158,
      "num_tokens": 78476411.0,
      "step": 35815
    },
    {
      "entropy": 5.679567527770996,
      "epoch": 3.554982135768162,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003777377719791016,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18738163411617278,
      "num_tokens": 78488562.0,
      "step": 35820
    },
    {
      "entropy": 5.672986173629761,
      "epoch": 3.5554783644303294,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037770624851600086,
      "loss": 4.9653,
      "mean_token_accuracy": 0.19375281631946564,
      "num_tokens": 78499455.0,
      "step": 35825
    },
    {
      "entropy": 5.834922409057617,
      "epoch": 3.555974593092497,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003776747225060065,
      "loss": 5.1381,
      "mean_token_accuracy": 0.18019738793373108,
      "num_tokens": 78510541.0,
      "step": 35830
    },
    {
      "entropy": 5.62293815612793,
      "epoch": 3.5564708217546643,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003776431939499003,
      "loss": 4.9501,
      "mean_token_accuracy": 0.19447430670261384,
      "num_tokens": 78521754.0,
      "step": 35835
    },
    {
      "entropy": 5.624145555496216,
      "epoch": 3.556967050416832,
      "grad_norm": 1.25,
      "learning_rate": 0.000377611662848464,
      "loss": 4.9685,
      "mean_token_accuracy": 0.19345417022705078,
      "num_tokens": 78531462.0,
      "step": 35840
    },
    {
      "entropy": 5.6812457084655765,
      "epoch": 3.5574632790789997,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003775801292024796,
      "loss": 5.0195,
      "mean_token_accuracy": 0.18800682425498963,
      "num_tokens": 78542465.0,
      "step": 35845
    },
    {
      "entropy": 5.671347427368164,
      "epoch": 3.557959507741167,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037754859301272905,
      "loss": 4.9372,
      "mean_token_accuracy": 0.1876778483390808,
      "num_tokens": 78553841.0,
      "step": 35850
    },
    {
      "entropy": 5.59931697845459,
      "epoch": 3.5584557364033347,
      "grad_norm": 1.125,
      "learning_rate": 0.00037751705427999436,
      "loss": 4.9215,
      "mean_token_accuracy": 0.18828086256980897,
      "num_tokens": 78563445.0,
      "step": 35855
    },
    {
      "entropy": 5.6482484340667725,
      "epoch": 3.558951965065502,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003774855130050576,
      "loss": 5.0794,
      "mean_token_accuracy": 0.18655998706817628,
      "num_tokens": 78575200.0,
      "step": 35860
    },
    {
      "entropy": 5.623643255233764,
      "epoch": 3.5594481937276696,
      "grad_norm": 1.328125,
      "learning_rate": 0.00037745396918870095,
      "loss": 4.9056,
      "mean_token_accuracy": 0.19536938220262529,
      "num_tokens": 78585866.0,
      "step": 35865
    },
    {
      "entropy": 5.702214860916138,
      "epoch": 3.559944422389837,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003774224228317067,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18677064925432205,
      "num_tokens": 78595716.0,
      "step": 35870
    },
    {
      "entropy": 5.635422325134277,
      "epoch": 3.560440651052005,
      "grad_norm": 1.1875,
      "learning_rate": 0.000377390873934857,
      "loss": 4.9771,
      "mean_token_accuracy": 0.19376278817653655,
      "num_tokens": 78607136.0,
      "step": 35875
    },
    {
      "entropy": 5.619132804870605,
      "epoch": 3.5609368797141725,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003773593224989343,
      "loss": 4.9492,
      "mean_token_accuracy": 0.19041824340820312,
      "num_tokens": 78617627.0,
      "step": 35880
    },
    {
      "entropy": 5.634245347976685,
      "epoch": 3.56143310837634,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003773277685247209,
      "loss": 4.9872,
      "mean_token_accuracy": 0.1865844964981079,
      "num_tokens": 78629152.0,
      "step": 35885
    },
    {
      "entropy": 5.674570751190186,
      "epoch": 3.5619293370385074,
      "grad_norm": 1.25,
      "learning_rate": 0.00037729621201299933,
      "loss": 5.007,
      "mean_token_accuracy": 0.1895175099372864,
      "num_tokens": 78640576.0,
      "step": 35890
    },
    {
      "entropy": 5.663583278656006,
      "epoch": 3.562425565700675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003772646529645522,
      "loss": 4.9531,
      "mean_token_accuracy": 0.18823898136615752,
      "num_tokens": 78651752.0,
      "step": 35895
    },
    {
      "entropy": 5.772834730148316,
      "epoch": 3.5629217943628424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003772330913801621,
      "loss": 5.0737,
      "mean_token_accuracy": 0.18291983008384705,
      "num_tokens": 78662928.0,
      "step": 35900
    },
    {
      "entropy": 5.641568565368653,
      "epoch": 3.56341802302501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003772015272606115,
      "loss": 5.0298,
      "mean_token_accuracy": 0.19366795867681502,
      "num_tokens": 78674862.0,
      "step": 35905
    },
    {
      "entropy": 5.604776096343994,
      "epoch": 3.5639142516871773,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003771699606066833,
      "loss": 4.9028,
      "mean_token_accuracy": 0.19449348002672195,
      "num_tokens": 78685599.0,
      "step": 35910
    },
    {
      "entropy": 5.656731986999512,
      "epoch": 3.564410480349345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003771383914191602,
      "loss": 4.9326,
      "mean_token_accuracy": 0.18845946490764617,
      "num_tokens": 78696318.0,
      "step": 35915
    },
    {
      "entropy": 5.733932113647461,
      "epoch": 3.5649067090115123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003771068196988251,
      "loss": 5.0278,
      "mean_token_accuracy": 0.18472936749458313,
      "num_tokens": 78708847.0,
      "step": 35920
    },
    {
      "entropy": 5.63212742805481,
      "epoch": 3.5654029376736798,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037707524544646094,
      "loss": 4.9332,
      "mean_token_accuracy": 0.19115409404039382,
      "num_tokens": 78720476.0,
      "step": 35925
    },
    {
      "entropy": 5.684889411926269,
      "epoch": 3.5658991663358477,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003770436686628506,
      "loss": 5.0429,
      "mean_token_accuracy": 0.18470740765333177,
      "num_tokens": 78731832.0,
      "step": 35930
    },
    {
      "entropy": 5.675644731521606,
      "epoch": 3.566395394998015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037701208934877716,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18577982783317565,
      "num_tokens": 78743192.0,
      "step": 35935
    },
    {
      "entropy": 5.67233190536499,
      "epoch": 3.5668916236601826,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037698050750502375,
      "loss": 4.9265,
      "mean_token_accuracy": 0.1877664163708687,
      "num_tokens": 78753951.0,
      "step": 35940
    },
    {
      "entropy": 5.7299675941467285,
      "epoch": 3.56738785232235,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003769489231323735,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18510354608297347,
      "num_tokens": 78765752.0,
      "step": 35945
    },
    {
      "entropy": 5.652947092056275,
      "epoch": 3.5678840809845176,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003769173362316096,
      "loss": 4.9669,
      "mean_token_accuracy": 0.18859882205724715,
      "num_tokens": 78777137.0,
      "step": 35950
    },
    {
      "entropy": 5.644150447845459,
      "epoch": 3.568380309646685,
      "grad_norm": 1.25,
      "learning_rate": 0.00037688574680351544,
      "loss": 4.998,
      "mean_token_accuracy": 0.19437896311283112,
      "num_tokens": 78787420.0,
      "step": 35955
    },
    {
      "entropy": 5.664277601242065,
      "epoch": 3.568876538308853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037685415484887426,
      "loss": 5.0002,
      "mean_token_accuracy": 0.1844146028161049,
      "num_tokens": 78797604.0,
      "step": 35960
    },
    {
      "entropy": 5.655544710159302,
      "epoch": 3.5693727669710205,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037682256036846947,
      "loss": 4.9493,
      "mean_token_accuracy": 0.19275809377431868,
      "num_tokens": 78808865.0,
      "step": 35965
    },
    {
      "entropy": 5.620528602600098,
      "epoch": 3.569868995633188,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037679096336308473,
      "loss": 4.9347,
      "mean_token_accuracy": 0.1889899879693985,
      "num_tokens": 78820223.0,
      "step": 35970
    },
    {
      "entropy": 5.691919374465942,
      "epoch": 3.5703652242953554,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037675936383350326,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1819260150194168,
      "num_tokens": 78832830.0,
      "step": 35975
    },
    {
      "entropy": 5.6840253353118895,
      "epoch": 3.570861452957523,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003767277617805089,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18948336988687514,
      "num_tokens": 78843059.0,
      "step": 35980
    },
    {
      "entropy": 5.66449670791626,
      "epoch": 3.5713576816196904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003766961572048853,
      "loss": 5.0147,
      "mean_token_accuracy": 0.19170699119567872,
      "num_tokens": 78854699.0,
      "step": 35985
    },
    {
      "entropy": 5.647118759155274,
      "epoch": 3.571853910281858,
      "grad_norm": 1.125,
      "learning_rate": 0.0003766645501074161,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18620632886886596,
      "num_tokens": 78866521.0,
      "step": 35990
    },
    {
      "entropy": 5.626260709762573,
      "epoch": 3.5723501389440253,
      "grad_norm": 1.125,
      "learning_rate": 0.00037663294048888513,
      "loss": 4.9742,
      "mean_token_accuracy": 0.19014747738838195,
      "num_tokens": 78877025.0,
      "step": 35995
    },
    {
      "entropy": 5.593442106246949,
      "epoch": 3.572846367606193,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037660132835007625,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2009861409664154,
      "num_tokens": 78888088.0,
      "step": 36000
    },
    {
      "epoch": 3.572846367606193,
      "eval_entropy": 5.3887415648243095,
      "eval_loss": 5.117626667022705,
      "eval_mean_token_accuracy": 0.1881009769230111,
      "eval_num_tokens": 78888088.0,
      "eval_runtime": 20.512,
      "eval_samples_per_second": 1641.426,
      "eval_steps_per_second": 205.197,
      "step": 36000
    },
    {
      "entropy": 5.630583381652832,
      "epoch": 3.5733425962683603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003765697136917734,
      "loss": 4.9592,
      "mean_token_accuracy": 0.1845878094434738,
      "num_tokens": 78898750.0,
      "step": 36005
    },
    {
      "entropy": 5.644330835342407,
      "epoch": 3.5738388249305277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003765380965147604,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18313964307308198,
      "num_tokens": 78911212.0,
      "step": 36010
    },
    {
      "entropy": 5.6820672988891605,
      "epoch": 3.5743350535926957,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037650647681982145,
      "loss": 5.0221,
      "mean_token_accuracy": 0.18790335357189178,
      "num_tokens": 78921970.0,
      "step": 36015
    },
    {
      "entropy": 5.657589721679687,
      "epoch": 3.574831282254863,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037647485460774065,
      "loss": 4.9641,
      "mean_token_accuracy": 0.18782323300838472,
      "num_tokens": 78931997.0,
      "step": 36020
    },
    {
      "entropy": 5.6440629959106445,
      "epoch": 3.5753275109170306,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003764432298793021,
      "loss": 4.9169,
      "mean_token_accuracy": 0.19003953784704208,
      "num_tokens": 78942504.0,
      "step": 36025
    },
    {
      "entropy": 5.6321272373199465,
      "epoch": 3.575823739579198,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037641160263529003,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18612733483314514,
      "num_tokens": 78953697.0,
      "step": 36030
    },
    {
      "entropy": 5.619674825668335,
      "epoch": 3.5763199682413656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037637997287648877,
      "loss": 4.9264,
      "mean_token_accuracy": 0.1969086319208145,
      "num_tokens": 78964751.0,
      "step": 36035
    },
    {
      "entropy": 5.623809909820556,
      "epoch": 3.576816196903533,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037634834060368256,
      "loss": 4.933,
      "mean_token_accuracy": 0.19512505680322648,
      "num_tokens": 78975910.0,
      "step": 36040
    },
    {
      "entropy": 5.613362789154053,
      "epoch": 3.577312425565701,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003763167058176559,
      "loss": 4.9513,
      "mean_token_accuracy": 0.18947090357542037,
      "num_tokens": 78985843.0,
      "step": 36045
    },
    {
      "entropy": 5.687096500396729,
      "epoch": 3.5778086542278684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003762850685191933,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18511902838945388,
      "num_tokens": 78997117.0,
      "step": 36050
    },
    {
      "entropy": 5.644630193710327,
      "epoch": 3.578304882890036,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003762534287090792,
      "loss": 4.9884,
      "mean_token_accuracy": 0.19241684675216675,
      "num_tokens": 79007944.0,
      "step": 36055
    },
    {
      "entropy": 5.661894273757935,
      "epoch": 3.5788011115522034,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003762217863880983,
      "loss": 5.0102,
      "mean_token_accuracy": 0.19124542623758317,
      "num_tokens": 79019136.0,
      "step": 36060
    },
    {
      "entropy": 5.719294881820678,
      "epoch": 3.579297340214371,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003761901415570351,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18442236185073851,
      "num_tokens": 79028906.0,
      "step": 36065
    },
    {
      "entropy": 5.705501651763916,
      "epoch": 3.5797935688765383,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037615849421667466,
      "loss": 4.9718,
      "mean_token_accuracy": 0.1881726562976837,
      "num_tokens": 79040249.0,
      "step": 36070
    },
    {
      "entropy": 5.695062160491943,
      "epoch": 3.580289797538706,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003761268443678014,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1837122067809105,
      "num_tokens": 79051278.0,
      "step": 36075
    },
    {
      "entropy": 5.740511274337768,
      "epoch": 3.5807860262008733,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003760951920112003,
      "loss": 5.0408,
      "mean_token_accuracy": 0.18282549381256102,
      "num_tokens": 79062131.0,
      "step": 36080
    },
    {
      "entropy": 5.640097951889038,
      "epoch": 3.5812822548630407,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037606353714765626,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18639382421970369,
      "num_tokens": 79073765.0,
      "step": 36085
    },
    {
      "entropy": 5.653460359573364,
      "epoch": 3.5817784835252082,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003760318797779543,
      "loss": 4.981,
      "mean_token_accuracy": 0.18916058391332627,
      "num_tokens": 79085125.0,
      "step": 36090
    },
    {
      "entropy": 5.656866455078125,
      "epoch": 3.5822747121873757,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003760002199028794,
      "loss": 4.9536,
      "mean_token_accuracy": 0.18832300305366517,
      "num_tokens": 79096198.0,
      "step": 36095
    },
    {
      "entropy": 5.6102324485778805,
      "epoch": 3.5827709408495436,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003759685575232168,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20185589790344238,
      "num_tokens": 79107117.0,
      "step": 36100
    },
    {
      "entropy": 5.6179399490356445,
      "epoch": 3.583267169511711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003759368926397514,
      "loss": 4.9244,
      "mean_token_accuracy": 0.1898450717329979,
      "num_tokens": 79117176.0,
      "step": 36105
    },
    {
      "entropy": 5.636882162094116,
      "epoch": 3.5837633981738786,
      "grad_norm": 1.125,
      "learning_rate": 0.0003759052252532686,
      "loss": 4.9583,
      "mean_token_accuracy": 0.19337248653173447,
      "num_tokens": 79128426.0,
      "step": 36110
    },
    {
      "entropy": 5.651574802398682,
      "epoch": 3.584259626836046,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003758735553645537,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18767762631177903,
      "num_tokens": 79139026.0,
      "step": 36115
    },
    {
      "entropy": 5.643036413192749,
      "epoch": 3.5847558554982135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000375841882974392,
      "loss": 4.9273,
      "mean_token_accuracy": 0.19171142876148223,
      "num_tokens": 79150188.0,
      "step": 36120
    },
    {
      "entropy": 5.615115022659301,
      "epoch": 3.585252084160381,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003758102080835688,
      "loss": 4.9246,
      "mean_token_accuracy": 0.19283712953329085,
      "num_tokens": 79161202.0,
      "step": 36125
    },
    {
      "entropy": 5.736236667633056,
      "epoch": 3.5857483128225485,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003757785306928697,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1886037528514862,
      "num_tokens": 79171647.0,
      "step": 36130
    },
    {
      "entropy": 5.715788793563843,
      "epoch": 3.5862445414847164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003757468508030803,
      "loss": 5.0321,
      "mean_token_accuracy": 0.18672584146261215,
      "num_tokens": 79182281.0,
      "step": 36135
    },
    {
      "entropy": 5.66501932144165,
      "epoch": 3.586740770146884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003757151684149859,
      "loss": 4.9919,
      "mean_token_accuracy": 0.1877460926771164,
      "num_tokens": 79192890.0,
      "step": 36140
    },
    {
      "entropy": 5.624875688552857,
      "epoch": 3.5872369988090513,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003756834835293725,
      "loss": 4.975,
      "mean_token_accuracy": 0.19079286158084868,
      "num_tokens": 79203371.0,
      "step": 36145
    },
    {
      "entropy": 5.675127840042114,
      "epoch": 3.587733227471219,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003756517961470256,
      "loss": 4.9783,
      "mean_token_accuracy": 0.19287658631801605,
      "num_tokens": 79214367.0,
      "step": 36150
    },
    {
      "entropy": 5.662697172164917,
      "epoch": 3.5882294561333863,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003756201062687311,
      "loss": 4.9451,
      "mean_token_accuracy": 0.19200853556394576,
      "num_tokens": 79225072.0,
      "step": 36155
    },
    {
      "entropy": 5.6130943775177,
      "epoch": 3.5887256847955538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037558841389527466,
      "loss": 4.9175,
      "mean_token_accuracy": 0.19611262679100036,
      "num_tokens": 79235852.0,
      "step": 36160
    },
    {
      "entropy": 5.6981425285339355,
      "epoch": 3.5892219134577212,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037555671902744244,
      "loss": 5.0537,
      "mean_token_accuracy": 0.1811954766511917,
      "num_tokens": 79246022.0,
      "step": 36165
    },
    {
      "entropy": 5.73314642906189,
      "epoch": 3.5897181421198887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037552502166602014,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18732177913188935,
      "num_tokens": 79256954.0,
      "step": 36170
    },
    {
      "entropy": 5.70257534980774,
      "epoch": 3.590214370782056,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037549332181179403,
      "loss": 5.0399,
      "mean_token_accuracy": 0.180110664665699,
      "num_tokens": 79266782.0,
      "step": 36175
    },
    {
      "entropy": 5.617231225967407,
      "epoch": 3.5907105994442237,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037546161946555004,
      "loss": 4.9098,
      "mean_token_accuracy": 0.19592326134443283,
      "num_tokens": 79276813.0,
      "step": 36180
    },
    {
      "entropy": 5.672365760803222,
      "epoch": 3.5912068281063916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037542991462807433,
      "loss": 5.0234,
      "mean_token_accuracy": 0.19390855282545089,
      "num_tokens": 79288184.0,
      "step": 36185
    },
    {
      "entropy": 5.738161373138428,
      "epoch": 3.591703056768559,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037539820730015316,
      "loss": 5.0329,
      "mean_token_accuracy": 0.18231531083583832,
      "num_tokens": 79298986.0,
      "step": 36190
    },
    {
      "entropy": 5.652725887298584,
      "epoch": 3.5921992854307265,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037536649748257285,
      "loss": 4.984,
      "mean_token_accuracy": 0.18395398557186127,
      "num_tokens": 79308548.0,
      "step": 36195
    },
    {
      "entropy": 5.672579145431518,
      "epoch": 3.592695514092894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037533478517611955,
      "loss": 4.9686,
      "mean_token_accuracy": 0.19251926988363266,
      "num_tokens": 79321197.0,
      "step": 36200
    },
    {
      "entropy": 5.599051809310913,
      "epoch": 3.5931917427550615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003753030703815799,
      "loss": 4.9214,
      "mean_token_accuracy": 0.19578320831060408,
      "num_tokens": 79331967.0,
      "step": 36205
    },
    {
      "entropy": 5.631637907028198,
      "epoch": 3.593687971417229,
      "grad_norm": 1.25,
      "learning_rate": 0.0003752713530997402,
      "loss": 4.9547,
      "mean_token_accuracy": 0.19170645922422408,
      "num_tokens": 79344850.0,
      "step": 36210
    },
    {
      "entropy": 5.641121959686279,
      "epoch": 3.5941842000793964,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037523963333138693,
      "loss": 4.9743,
      "mean_token_accuracy": 0.1954112783074379,
      "num_tokens": 79354820.0,
      "step": 36215
    },
    {
      "entropy": 5.63588695526123,
      "epoch": 3.5946804287415643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003752079110773068,
      "loss": 4.9349,
      "mean_token_accuracy": 0.20096879154443742,
      "num_tokens": 79365389.0,
      "step": 36220
    },
    {
      "entropy": 5.698690176010132,
      "epoch": 3.595176657403732,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003751761863382865,
      "loss": 5.0412,
      "mean_token_accuracy": 0.1868120864033699,
      "num_tokens": 79377086.0,
      "step": 36225
    },
    {
      "entropy": 5.645179224014282,
      "epoch": 3.5956728860658993,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037514445911511257,
      "loss": 4.9511,
      "mean_token_accuracy": 0.19063282012939453,
      "num_tokens": 79388206.0,
      "step": 36230
    },
    {
      "entropy": 5.615921306610107,
      "epoch": 3.5961691147280668,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003751127294085719,
      "loss": 4.9042,
      "mean_token_accuracy": 0.19608356207609176,
      "num_tokens": 79400591.0,
      "step": 36235
    },
    {
      "entropy": 5.6457733631134035,
      "epoch": 3.5966653433902342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037508099721945113,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18741749078035355,
      "num_tokens": 79411569.0,
      "step": 36240
    },
    {
      "entropy": 5.639098882675171,
      "epoch": 3.5971615720524017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003750492625485374,
      "loss": 4.963,
      "mean_token_accuracy": 0.19054219573736192,
      "num_tokens": 79422711.0,
      "step": 36245
    },
    {
      "entropy": 5.6855627536773685,
      "epoch": 3.597657800714569,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003750175253966175,
      "loss": 4.9554,
      "mean_token_accuracy": 0.18691862225532532,
      "num_tokens": 79434156.0,
      "step": 36250
    },
    {
      "entropy": 5.664129829406738,
      "epoch": 3.5981540293767367,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003749857857644785,
      "loss": 4.9666,
      "mean_token_accuracy": 0.19153162240982055,
      "num_tokens": 79445087.0,
      "step": 36255
    },
    {
      "entropy": 5.646773862838745,
      "epoch": 3.598650258038904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003749540436529074,
      "loss": 4.9419,
      "mean_token_accuracy": 0.19484038800001144,
      "num_tokens": 79455649.0,
      "step": 36260
    },
    {
      "entropy": 5.5843604564666744,
      "epoch": 3.5991464867010716,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037492229906269147,
      "loss": 4.9321,
      "mean_token_accuracy": 0.20210970044136048,
      "num_tokens": 79468691.0,
      "step": 36265
    },
    {
      "entropy": 5.663576126098633,
      "epoch": 3.5996427153632395,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037489055199461783,
      "loss": 4.9859,
      "mean_token_accuracy": 0.18583015501499175,
      "num_tokens": 79479547.0,
      "step": 36270
    },
    {
      "entropy": 5.644497489929199,
      "epoch": 3.600138944025407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003748588024494737,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18703556805849075,
      "num_tokens": 79490454.0,
      "step": 36275
    },
    {
      "entropy": 5.642437553405761,
      "epoch": 3.6006351726875745,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037482705042804655,
      "loss": 4.9708,
      "mean_token_accuracy": 0.1929944783449173,
      "num_tokens": 79500461.0,
      "step": 36280
    },
    {
      "entropy": 5.633067941665649,
      "epoch": 3.601131401349742,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037479529593112355,
      "loss": 4.9921,
      "mean_token_accuracy": 0.1874341130256653,
      "num_tokens": 79512063.0,
      "step": 36285
    },
    {
      "entropy": 5.621622514724732,
      "epoch": 3.6016276300119094,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003747635389594923,
      "loss": 4.9416,
      "mean_token_accuracy": 0.19304032772779464,
      "num_tokens": 79523831.0,
      "step": 36290
    },
    {
      "entropy": 5.647227191925049,
      "epoch": 3.602123858674077,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037473177951394027,
      "loss": 5.025,
      "mean_token_accuracy": 0.1945912778377533,
      "num_tokens": 79535103.0,
      "step": 36295
    },
    {
      "entropy": 5.6409097671508786,
      "epoch": 3.6026200873362444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037470001759525495,
      "loss": 4.9114,
      "mean_token_accuracy": 0.1944926992058754,
      "num_tokens": 79546928.0,
      "step": 36300
    },
    {
      "entropy": 5.694170141220093,
      "epoch": 3.6031163159984123,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037466825320422407,
      "loss": 4.9406,
      "mean_token_accuracy": 0.19032416492700577,
      "num_tokens": 79557903.0,
      "step": 36305
    },
    {
      "entropy": 5.719187307357788,
      "epoch": 3.60361254466058,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037463648634163534,
      "loss": 5.0237,
      "mean_token_accuracy": 0.18454048782587051,
      "num_tokens": 79568269.0,
      "step": 36310
    },
    {
      "entropy": 5.653478193283081,
      "epoch": 3.6041087733227473,
      "grad_norm": 1.125,
      "learning_rate": 0.0003746047170082764,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18802666813135147,
      "num_tokens": 79579118.0,
      "step": 36315
    },
    {
      "entropy": 5.6512007236480715,
      "epoch": 3.6046050019849147,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037457294520493504,
      "loss": 5.019,
      "mean_token_accuracy": 0.1956293612718582,
      "num_tokens": 79590738.0,
      "step": 36320
    },
    {
      "entropy": 5.618615961074829,
      "epoch": 3.605101230647082,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037454117093239925,
      "loss": 4.9224,
      "mean_token_accuracy": 0.19773118644952775,
      "num_tokens": 79601367.0,
      "step": 36325
    },
    {
      "entropy": 5.6513991355896,
      "epoch": 3.6055974593092497,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003745093941914569,
      "loss": 5.0115,
      "mean_token_accuracy": 0.18861310929059982,
      "num_tokens": 79611361.0,
      "step": 36330
    },
    {
      "entropy": 5.692750263214111,
      "epoch": 3.606093687971417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037447761498289595,
      "loss": 5.0714,
      "mean_token_accuracy": 0.18681968748569489,
      "num_tokens": 79622520.0,
      "step": 36335
    },
    {
      "entropy": 5.661763429641724,
      "epoch": 3.6065899166335846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003744458333075046,
      "loss": 4.963,
      "mean_token_accuracy": 0.18976689875125885,
      "num_tokens": 79632268.0,
      "step": 36340
    },
    {
      "entropy": 5.740756177902222,
      "epoch": 3.607086145295752,
      "grad_norm": 1.125,
      "learning_rate": 0.0003744140491660708,
      "loss": 5.0395,
      "mean_token_accuracy": 0.19060586094856263,
      "num_tokens": 79643438.0,
      "step": 36345
    },
    {
      "entropy": 5.658565711975098,
      "epoch": 3.6075823739579196,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003743822625593829,
      "loss": 5.0418,
      "mean_token_accuracy": 0.18682287484407425,
      "num_tokens": 79654101.0,
      "step": 36350
    },
    {
      "entropy": 5.565096139907837,
      "epoch": 3.608078602620087,
      "grad_norm": 1.125,
      "learning_rate": 0.000374350473488229,
      "loss": 4.873,
      "mean_token_accuracy": 0.19422168880701066,
      "num_tokens": 79666129.0,
      "step": 36355
    },
    {
      "entropy": 5.666993904113769,
      "epoch": 3.608574831282255,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037431868195339735,
      "loss": 5.0075,
      "mean_token_accuracy": 0.18669512122869492,
      "num_tokens": 79677003.0,
      "step": 36360
    },
    {
      "entropy": 5.611107110977173,
      "epoch": 3.6090710599444225,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003742868879556764,
      "loss": 4.8967,
      "mean_token_accuracy": 0.19787642806768418,
      "num_tokens": 79687040.0,
      "step": 36365
    },
    {
      "entropy": 5.650229024887085,
      "epoch": 3.60956728860659,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037425509149585466,
      "loss": 4.938,
      "mean_token_accuracy": 0.19800885021686554,
      "num_tokens": 79698340.0,
      "step": 36370
    },
    {
      "entropy": 5.655611324310303,
      "epoch": 3.6100635172687574,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003742232925747205,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1892659604549408,
      "num_tokens": 79710230.0,
      "step": 36375
    },
    {
      "entropy": 5.655012893676758,
      "epoch": 3.610559745930925,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003741914911930625,
      "loss": 4.9713,
      "mean_token_accuracy": 0.191680608689785,
      "num_tokens": 79722083.0,
      "step": 36380
    },
    {
      "entropy": 5.712177038192749,
      "epoch": 3.6110559745930924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003741596873516693,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18538496196269988,
      "num_tokens": 79733979.0,
      "step": 36385
    },
    {
      "entropy": 5.603067255020141,
      "epoch": 3.6115522032552603,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003741278810513294,
      "loss": 4.9027,
      "mean_token_accuracy": 0.19546038657426834,
      "num_tokens": 79746577.0,
      "step": 36390
    },
    {
      "entropy": 5.640611934661865,
      "epoch": 3.6120484319174277,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00037409607229283175,
      "loss": 4.9728,
      "mean_token_accuracy": 0.19206819534301758,
      "num_tokens": 79757124.0,
      "step": 36395
    },
    {
      "entropy": 5.642796039581299,
      "epoch": 3.6125446605795952,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000374064261076965,
      "loss": 4.9068,
      "mean_token_accuracy": 0.19602862894535064,
      "num_tokens": 79766314.0,
      "step": 36400
    },
    {
      "entropy": 5.724717617034912,
      "epoch": 3.6130408892417627,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000374032447404518,
      "loss": 5.0633,
      "mean_token_accuracy": 0.17961380928754805,
      "num_tokens": 79777777.0,
      "step": 36405
    },
    {
      "entropy": 5.628776931762696,
      "epoch": 3.61353711790393,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003740006312762799,
      "loss": 4.9201,
      "mean_token_accuracy": 0.19340633302927018,
      "num_tokens": 79788125.0,
      "step": 36410
    },
    {
      "entropy": 5.661110591888428,
      "epoch": 3.6140333465660976,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037396881269303924,
      "loss": 4.9493,
      "mean_token_accuracy": 0.1906571865081787,
      "num_tokens": 79800007.0,
      "step": 36415
    },
    {
      "entropy": 5.676184749603271,
      "epoch": 3.614529575228265,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037393699165558534,
      "loss": 5.0131,
      "mean_token_accuracy": 0.19342387467622757,
      "num_tokens": 79812051.0,
      "step": 36420
    },
    {
      "entropy": 5.635017442703247,
      "epoch": 3.6150258038904326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003739051681647073,
      "loss": 4.9658,
      "mean_token_accuracy": 0.1894103467464447,
      "num_tokens": 79823613.0,
      "step": 36425
    },
    {
      "entropy": 5.626215028762817,
      "epoch": 3.6155220325526,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037387334222119424,
      "loss": 4.8889,
      "mean_token_accuracy": 0.19716309159994125,
      "num_tokens": 79834015.0,
      "step": 36430
    },
    {
      "entropy": 5.565779733657837,
      "epoch": 3.6160182612147675,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037384151382583526,
      "loss": 4.8989,
      "mean_token_accuracy": 0.18988738507032393,
      "num_tokens": 79844210.0,
      "step": 36435
    },
    {
      "entropy": 5.651032733917236,
      "epoch": 3.616514489876935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003738096829794197,
      "loss": 4.9573,
      "mean_token_accuracy": 0.18996265530586243,
      "num_tokens": 79854918.0,
      "step": 36440
    },
    {
      "entropy": 5.651662063598633,
      "epoch": 3.617010718539103,
      "grad_norm": 1.15625,
      "learning_rate": 0.000373777849682737,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18128376305103303,
      "num_tokens": 79866526.0,
      "step": 36445
    },
    {
      "entropy": 5.6193805694580075,
      "epoch": 3.6175069472012704,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003737460139365764,
      "loss": 4.9917,
      "mean_token_accuracy": 0.18502202033996581,
      "num_tokens": 79876596.0,
      "step": 36450
    },
    {
      "entropy": 5.693127059936524,
      "epoch": 3.618003175863438,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003737141757417275,
      "loss": 4.992,
      "mean_token_accuracy": 0.18932974487543106,
      "num_tokens": 79887430.0,
      "step": 36455
    },
    {
      "entropy": 5.695441770553589,
      "epoch": 3.6184994045256054,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003736823350989797,
      "loss": 4.9852,
      "mean_token_accuracy": 0.1987675592303276,
      "num_tokens": 79899497.0,
      "step": 36460
    },
    {
      "entropy": 5.671443128585816,
      "epoch": 3.618995633187773,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037365049200912274,
      "loss": 4.9374,
      "mean_token_accuracy": 0.19294947683811187,
      "num_tokens": 79909199.0,
      "step": 36465
    },
    {
      "entropy": 5.7151846408844,
      "epoch": 3.6194918618499403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037361864647294596,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1877562075853348,
      "num_tokens": 79920369.0,
      "step": 36470
    },
    {
      "entropy": 5.610972738265991,
      "epoch": 3.6199880905121082,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003735867984912394,
      "loss": 4.9729,
      "mean_token_accuracy": 0.1863068386912346,
      "num_tokens": 79930983.0,
      "step": 36475
    },
    {
      "entropy": 5.628896284103393,
      "epoch": 3.6204843191742757,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037355494806479264,
      "loss": 4.9223,
      "mean_token_accuracy": 0.19660317599773408,
      "num_tokens": 79942006.0,
      "step": 36480
    },
    {
      "entropy": 5.627692651748657,
      "epoch": 3.620980547836443,
      "grad_norm": 1.125,
      "learning_rate": 0.0003735230951943955,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18393238484859467,
      "num_tokens": 79951956.0,
      "step": 36485
    },
    {
      "entropy": 5.648015832901001,
      "epoch": 3.6214767764986107,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037349123988083795,
      "loss": 4.9769,
      "mean_token_accuracy": 0.19148577600717545,
      "num_tokens": 79962084.0,
      "step": 36490
    },
    {
      "entropy": 5.659411668777466,
      "epoch": 3.621973005160778,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037345938212490987,
      "loss": 4.935,
      "mean_token_accuracy": 0.19023015052080156,
      "num_tokens": 79971514.0,
      "step": 36495
    },
    {
      "entropy": 5.6921639919281,
      "epoch": 3.6224692338229456,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037342752192740127,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1797749236226082,
      "num_tokens": 79983017.0,
      "step": 36500
    },
    {
      "entropy": 5.699198532104492,
      "epoch": 3.622965462485113,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003733956592891022,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18422981798648835,
      "num_tokens": 79994137.0,
      "step": 36505
    },
    {
      "entropy": 5.754218053817749,
      "epoch": 3.6234616911472806,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037336379421080274,
      "loss": 5.0547,
      "mean_token_accuracy": 0.18410540670156478,
      "num_tokens": 80005384.0,
      "step": 36510
    },
    {
      "entropy": 5.6926741123199465,
      "epoch": 3.623957919809448,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003733319266932932,
      "loss": 5.046,
      "mean_token_accuracy": 0.1871706262230873,
      "num_tokens": 80016608.0,
      "step": 36515
    },
    {
      "entropy": 5.610333442687988,
      "epoch": 3.6244541484716155,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037330005673736385,
      "loss": 4.9287,
      "mean_token_accuracy": 0.19504816085100174,
      "num_tokens": 80027746.0,
      "step": 36520
    },
    {
      "entropy": 5.661565494537354,
      "epoch": 3.624950377133783,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003732681843438048,
      "loss": 4.9244,
      "mean_token_accuracy": 0.1905094176530838,
      "num_tokens": 80038449.0,
      "step": 36525
    },
    {
      "entropy": 5.6724122047424315,
      "epoch": 3.625446605795951,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037323630951340654,
      "loss": 4.9487,
      "mean_token_accuracy": 0.18758575171232222,
      "num_tokens": 80049079.0,
      "step": 36530
    },
    {
      "entropy": 5.7537071228027346,
      "epoch": 3.6259428344581184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037320443224695955,
      "loss": 5.1202,
      "mean_token_accuracy": 0.17444924712181092,
      "num_tokens": 80061130.0,
      "step": 36535
    },
    {
      "entropy": 5.709438896179199,
      "epoch": 3.626439063120286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003731725525452542,
      "loss": 4.9756,
      "mean_token_accuracy": 0.1895027533173561,
      "num_tokens": 80072714.0,
      "step": 36540
    },
    {
      "entropy": 5.694350957870483,
      "epoch": 3.6269352917824533,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003731406704090811,
      "loss": 4.977,
      "mean_token_accuracy": 0.18940132409334182,
      "num_tokens": 80083506.0,
      "step": 36545
    },
    {
      "entropy": 5.694737958908081,
      "epoch": 3.627431520444621,
      "grad_norm": 1.125,
      "learning_rate": 0.00037310878583923085,
      "loss": 4.9952,
      "mean_token_accuracy": 0.18967297673225403,
      "num_tokens": 80094866.0,
      "step": 36550
    },
    {
      "entropy": 5.637590980529785,
      "epoch": 3.6279277491067883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003730768988364941,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18858034014701844,
      "num_tokens": 80105857.0,
      "step": 36555
    },
    {
      "entropy": 5.591903924942017,
      "epoch": 3.628423977768956,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037304500940166165,
      "loss": 4.9241,
      "mean_token_accuracy": 0.19446289539337158,
      "num_tokens": 80116634.0,
      "step": 36560
    },
    {
      "entropy": 5.7064855098724365,
      "epoch": 3.6289202064311237,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037301311753552407,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1878653883934021,
      "num_tokens": 80126619.0,
      "step": 36565
    },
    {
      "entropy": 5.643278312683106,
      "epoch": 3.629416435093291,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003729812232388726,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1847036898136139,
      "num_tokens": 80137090.0,
      "step": 36570
    },
    {
      "entropy": 5.698978424072266,
      "epoch": 3.6299126637554586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003729493265124978,
      "loss": 5.0148,
      "mean_token_accuracy": 0.18756673783063887,
      "num_tokens": 80148039.0,
      "step": 36575
    },
    {
      "entropy": 5.6149462223052975,
      "epoch": 3.630408892417626,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037291742735719077,
      "loss": 4.8952,
      "mean_token_accuracy": 0.1918693885207176,
      "num_tokens": 80158674.0,
      "step": 36580
    },
    {
      "entropy": 5.714664554595947,
      "epoch": 3.6309051210797936,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003728855257737425,
      "loss": 5.0987,
      "mean_token_accuracy": 0.18877894282341004,
      "num_tokens": 80170371.0,
      "step": 36585
    },
    {
      "entropy": 5.746000289916992,
      "epoch": 3.631401349741961,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003728536217629442,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18854381889104843,
      "num_tokens": 80182817.0,
      "step": 36590
    },
    {
      "entropy": 5.652479600906372,
      "epoch": 3.6318975784041285,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037282171532558697,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1850455954670906,
      "num_tokens": 80194093.0,
      "step": 36595
    },
    {
      "entropy": 5.717727422714233,
      "epoch": 3.632393807066296,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037278980646246186,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18528606742620468,
      "num_tokens": 80204573.0,
      "step": 36600
    },
    {
      "entropy": 5.60080189704895,
      "epoch": 3.6328900357284635,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037275789517436035,
      "loss": 4.9124,
      "mean_token_accuracy": 0.19209342151880265,
      "num_tokens": 80216233.0,
      "step": 36605
    },
    {
      "entropy": 5.633810091018677,
      "epoch": 3.633386264390631,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037272598146207363,
      "loss": 4.9415,
      "mean_token_accuracy": 0.19344754368066788,
      "num_tokens": 80226412.0,
      "step": 36610
    },
    {
      "entropy": 5.570567846298218,
      "epoch": 3.633882493052799,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037269406532639324,
      "loss": 4.8856,
      "mean_token_accuracy": 0.19604823738336563,
      "num_tokens": 80238256.0,
      "step": 36615
    },
    {
      "entropy": 5.655985641479492,
      "epoch": 3.6343787217149663,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003726621467681105,
      "loss": 5.002,
      "mean_token_accuracy": 0.19570130854845047,
      "num_tokens": 80250143.0,
      "step": 36620
    },
    {
      "entropy": 5.573909616470337,
      "epoch": 3.634874950377134,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003726302257880169,
      "loss": 4.9126,
      "mean_token_accuracy": 0.1991301506757736,
      "num_tokens": 80261915.0,
      "step": 36625
    },
    {
      "entropy": 5.643295764923096,
      "epoch": 3.6353711790393013,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003725983023869041,
      "loss": 4.9997,
      "mean_token_accuracy": 0.19218621253967286,
      "num_tokens": 80274269.0,
      "step": 36630
    },
    {
      "entropy": 5.639516973495484,
      "epoch": 3.6358674077014688,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003725663765655638,
      "loss": 4.9136,
      "mean_token_accuracy": 0.19100714176893235,
      "num_tokens": 80285903.0,
      "step": 36635
    },
    {
      "entropy": 5.609486722946167,
      "epoch": 3.6363636363636362,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003725344483247875,
      "loss": 4.9522,
      "mean_token_accuracy": 0.18829403519630433,
      "num_tokens": 80298146.0,
      "step": 36640
    },
    {
      "entropy": 5.615925264358521,
      "epoch": 3.636859865025804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037250251766536715,
      "loss": 4.9377,
      "mean_token_accuracy": 0.19245900213718414,
      "num_tokens": 80309000.0,
      "step": 36645
    },
    {
      "entropy": 5.736731338500976,
      "epoch": 3.6373560936879716,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003724705845880944,
      "loss": 5.0698,
      "mean_token_accuracy": 0.18096752762794494,
      "num_tokens": 80318998.0,
      "step": 36650
    },
    {
      "entropy": 5.636800718307495,
      "epoch": 3.637852322350139,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003724386490937611,
      "loss": 4.9016,
      "mean_token_accuracy": 0.19518256336450576,
      "num_tokens": 80330413.0,
      "step": 36655
    },
    {
      "entropy": 5.656925916671753,
      "epoch": 3.6383485510123066,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003724067111831594,
      "loss": 4.9919,
      "mean_token_accuracy": 0.1925901934504509,
      "num_tokens": 80342543.0,
      "step": 36660
    },
    {
      "entropy": 5.607098054885864,
      "epoch": 3.638844779674474,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003723747708570811,
      "loss": 4.9215,
      "mean_token_accuracy": 0.1924879238009453,
      "num_tokens": 80353674.0,
      "step": 36665
    },
    {
      "entropy": 5.6806446552276615,
      "epoch": 3.6393410083366415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037234282811631826,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1869166299700737,
      "num_tokens": 80365565.0,
      "step": 36670
    },
    {
      "entropy": 5.670293521881104,
      "epoch": 3.639837236998809,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003723108829616631,
      "loss": 5.0136,
      "mean_token_accuracy": 0.1842315152287483,
      "num_tokens": 80376588.0,
      "step": 36675
    },
    {
      "entropy": 5.607781410217285,
      "epoch": 3.6403334656609765,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003722789353939077,
      "loss": 4.9034,
      "mean_token_accuracy": 0.19661624878644943,
      "num_tokens": 80387385.0,
      "step": 36680
    },
    {
      "entropy": 5.624921989440918,
      "epoch": 3.640829694323144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037224698541384434,
      "loss": 4.9156,
      "mean_token_accuracy": 0.19667692184448243,
      "num_tokens": 80398217.0,
      "step": 36685
    },
    {
      "entropy": 5.717300605773926,
      "epoch": 3.6413259229853114,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037221503302226533,
      "loss": 5.0466,
      "mean_token_accuracy": 0.1810264006257057,
      "num_tokens": 80410102.0,
      "step": 36690
    },
    {
      "entropy": 5.625971698760987,
      "epoch": 3.641822151647479,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003721830782199629,
      "loss": 4.8892,
      "mean_token_accuracy": 0.19249364733695984,
      "num_tokens": 80421362.0,
      "step": 36695
    },
    {
      "entropy": 5.602900600433349,
      "epoch": 3.642318380309647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003721511210077296,
      "loss": 4.9416,
      "mean_token_accuracy": 0.19602277427911757,
      "num_tokens": 80432258.0,
      "step": 36700
    },
    {
      "entropy": 5.646578788757324,
      "epoch": 3.6428146089718143,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037211916138635775,
      "loss": 4.9907,
      "mean_token_accuracy": 0.18650889545679092,
      "num_tokens": 80444173.0,
      "step": 36705
    },
    {
      "entropy": 5.6420838832855225,
      "epoch": 3.643310837633982,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037208719935664003,
      "loss": 4.9055,
      "mean_token_accuracy": 0.19683746248483658,
      "num_tokens": 80453516.0,
      "step": 36710
    },
    {
      "entropy": 5.628501129150391,
      "epoch": 3.6438070662961493,
      "grad_norm": 1.125,
      "learning_rate": 0.000372055234919369,
      "loss": 4.967,
      "mean_token_accuracy": 0.1929011180996895,
      "num_tokens": 80464467.0,
      "step": 36715
    },
    {
      "entropy": 5.675149297714233,
      "epoch": 3.6443032949583167,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037202326807533724,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18208200633525848,
      "num_tokens": 80475871.0,
      "step": 36720
    },
    {
      "entropy": 5.645878696441651,
      "epoch": 3.644799523620484,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003719912988253376,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18695057332515716,
      "num_tokens": 80485636.0,
      "step": 36725
    },
    {
      "entropy": 5.590349626541138,
      "epoch": 3.6452957522826517,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037195932717016254,
      "loss": 4.9677,
      "mean_token_accuracy": 0.19289606511592866,
      "num_tokens": 80496820.0,
      "step": 36730
    },
    {
      "entropy": 5.607950353622437,
      "epoch": 3.6457919809448196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003719273531106053,
      "loss": 4.9357,
      "mean_token_accuracy": 0.18966564238071443,
      "num_tokens": 80508057.0,
      "step": 36735
    },
    {
      "entropy": 5.645865249633789,
      "epoch": 3.646288209606987,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037189537664745846,
      "loss": 4.9527,
      "mean_token_accuracy": 0.1913609802722931,
      "num_tokens": 80517705.0,
      "step": 36740
    },
    {
      "entropy": 5.59333291053772,
      "epoch": 3.6467844382691545,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003718633977815151,
      "loss": 4.9077,
      "mean_token_accuracy": 0.1911364659667015,
      "num_tokens": 80528429.0,
      "step": 36745
    },
    {
      "entropy": 5.658314371109009,
      "epoch": 3.647280666931322,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037183141651356814,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18244606554508208,
      "num_tokens": 80538152.0,
      "step": 36750
    },
    {
      "entropy": 5.610689687728882,
      "epoch": 3.6477768955934895,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003717994328444108,
      "loss": 4.9345,
      "mean_token_accuracy": 0.18901923149824143,
      "num_tokens": 80548561.0,
      "step": 36755
    },
    {
      "entropy": 5.7095232009887695,
      "epoch": 3.648273124255657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037176744677483605,
      "loss": 5.0476,
      "mean_token_accuracy": 0.1844421371817589,
      "num_tokens": 80561170.0,
      "step": 36760
    },
    {
      "entropy": 5.648474550247192,
      "epoch": 3.6487693529178244,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037173545830563714,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19307381957769393,
      "num_tokens": 80572173.0,
      "step": 36765
    },
    {
      "entropy": 5.766360282897949,
      "epoch": 3.649265581579992,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003717034674376073,
      "loss": 5.0066,
      "mean_token_accuracy": 0.1885954797267914,
      "num_tokens": 80583323.0,
      "step": 36770
    },
    {
      "entropy": 5.662705183029175,
      "epoch": 3.6497618102421594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037167147417153983,
      "loss": 4.9237,
      "mean_token_accuracy": 0.19936973303556443,
      "num_tokens": 80592895.0,
      "step": 36775
    },
    {
      "entropy": 5.597684383392334,
      "epoch": 3.650258038904327,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037163947850822816,
      "loss": 4.9633,
      "mean_token_accuracy": 0.18951406627893447,
      "num_tokens": 80604015.0,
      "step": 36780
    },
    {
      "entropy": 5.675048875808716,
      "epoch": 3.650754267566495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037160748044846566,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18763148337602614,
      "num_tokens": 80615701.0,
      "step": 36785
    },
    {
      "entropy": 5.634411334991455,
      "epoch": 3.6512504962286623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003715754799930458,
      "loss": 4.8939,
      "mean_token_accuracy": 0.19313872158527373,
      "num_tokens": 80626177.0,
      "step": 36790
    },
    {
      "entropy": 5.664039421081543,
      "epoch": 3.6517467248908297,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003715434771427621,
      "loss": 4.9514,
      "mean_token_accuracy": 0.19807450026273726,
      "num_tokens": 80636478.0,
      "step": 36795
    },
    {
      "entropy": 5.723472738265992,
      "epoch": 3.652242953552997,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003715114718984083,
      "loss": 5.0642,
      "mean_token_accuracy": 0.1788049429655075,
      "num_tokens": 80647471.0,
      "step": 36800
    },
    {
      "entropy": 5.697486019134521,
      "epoch": 3.6527391822151647,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037147946426077785,
      "loss": 4.9676,
      "mean_token_accuracy": 0.18979230225086213,
      "num_tokens": 80657625.0,
      "step": 36805
    },
    {
      "entropy": 5.563672685623169,
      "epoch": 3.653235410877332,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003714474542306646,
      "loss": 4.9117,
      "mean_token_accuracy": 0.1983661487698555,
      "num_tokens": 80667250.0,
      "step": 36810
    },
    {
      "entropy": 5.666664266586304,
      "epoch": 3.6537316395394996,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037141544180886233,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18366582542657853,
      "num_tokens": 80677643.0,
      "step": 36815
    },
    {
      "entropy": 5.7053735733032225,
      "epoch": 3.6542278682016676,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003713834269961649,
      "loss": 4.9628,
      "mean_token_accuracy": 0.18909056037664412,
      "num_tokens": 80689246.0,
      "step": 36820
    },
    {
      "entropy": 5.678879117965698,
      "epoch": 3.654724096863835,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037135140979336614,
      "loss": 4.9644,
      "mean_token_accuracy": 0.18031488507986068,
      "num_tokens": 80699383.0,
      "step": 36825
    },
    {
      "entropy": 5.645223999023438,
      "epoch": 3.6552203255260025,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037131939020126,
      "loss": 4.9596,
      "mean_token_accuracy": 0.19409577697515487,
      "num_tokens": 80711221.0,
      "step": 36830
    },
    {
      "entropy": 5.643523550033569,
      "epoch": 3.65571655418817,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003712873682206406,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1896125480532646,
      "num_tokens": 80723379.0,
      "step": 36835
    },
    {
      "entropy": 5.704649353027344,
      "epoch": 3.6562127828503375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003712553438523019,
      "loss": 4.9845,
      "mean_token_accuracy": 0.19285110384225845,
      "num_tokens": 80735404.0,
      "step": 36840
    },
    {
      "entropy": 5.686197662353516,
      "epoch": 3.656709011512505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037122331709703803,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18552727103233338,
      "num_tokens": 80747053.0,
      "step": 36845
    },
    {
      "entropy": 5.707088947296143,
      "epoch": 3.6572052401746724,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003711912879556433,
      "loss": 5.0264,
      "mean_token_accuracy": 0.1880284383893013,
      "num_tokens": 80757574.0,
      "step": 36850
    },
    {
      "entropy": 5.596158266067505,
      "epoch": 3.65770146883684,
      "grad_norm": 1.0625,
      "learning_rate": 0.000371159256428912,
      "loss": 4.9456,
      "mean_token_accuracy": 0.19616479575634002,
      "num_tokens": 80767916.0,
      "step": 36855
    },
    {
      "entropy": 5.697897481918335,
      "epoch": 3.6581976974990074,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037112722251763817,
      "loss": 5.1067,
      "mean_token_accuracy": 0.1798612505197525,
      "num_tokens": 80778698.0,
      "step": 36860
    },
    {
      "entropy": 5.698330688476562,
      "epoch": 3.658693926161175,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003710951862226165,
      "loss": 5.0128,
      "mean_token_accuracy": 0.19278208166360855,
      "num_tokens": 80789997.0,
      "step": 36865
    },
    {
      "entropy": 5.6009032249450685,
      "epoch": 3.6591901548233428,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037106314754464125,
      "loss": 4.8978,
      "mean_token_accuracy": 0.19558734744787215,
      "num_tokens": 80800435.0,
      "step": 36870
    },
    {
      "entropy": 5.634718227386474,
      "epoch": 3.6596863834855102,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003710311064845069,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18554605096578597,
      "num_tokens": 80811761.0,
      "step": 36875
    },
    {
      "entropy": 5.658965158462524,
      "epoch": 3.6601826121476777,
      "grad_norm": 1.078125,
      "learning_rate": 0.000370999063043008,
      "loss": 4.9499,
      "mean_token_accuracy": 0.19339633882045745,
      "num_tokens": 80822084.0,
      "step": 36880
    },
    {
      "entropy": 5.6261210441589355,
      "epoch": 3.660678840809845,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003709670172209393,
      "loss": 4.9517,
      "mean_token_accuracy": 0.18880850821733475,
      "num_tokens": 80832863.0,
      "step": 36885
    },
    {
      "entropy": 5.6106963634490965,
      "epoch": 3.6611750694720127,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003709349690190952,
      "loss": 4.9264,
      "mean_token_accuracy": 0.19123917669057847,
      "num_tokens": 80843253.0,
      "step": 36890
    },
    {
      "entropy": 5.679027938842774,
      "epoch": 3.66167129813418,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003709029184382708,
      "loss": 4.9063,
      "mean_token_accuracy": 0.19638219475746155,
      "num_tokens": 80854002.0,
      "step": 36895
    },
    {
      "entropy": 5.596775913238526,
      "epoch": 3.6621675267963476,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003708708654792605,
      "loss": 4.8717,
      "mean_token_accuracy": 0.19648280292749404,
      "num_tokens": 80863484.0,
      "step": 36900
    },
    {
      "entropy": 5.572892141342163,
      "epoch": 3.6626637554585155,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037083881014285944,
      "loss": 4.8921,
      "mean_token_accuracy": 0.19809595197439195,
      "num_tokens": 80874952.0,
      "step": 36905
    },
    {
      "entropy": 5.579440689086914,
      "epoch": 3.663159984120683,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037080675242986234,
      "loss": 4.9482,
      "mean_token_accuracy": 0.18957941085100175,
      "num_tokens": 80886046.0,
      "step": 36910
    },
    {
      "entropy": 5.693275785446167,
      "epoch": 3.6636562127828505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003707746923410642,
      "loss": 5.0465,
      "mean_token_accuracy": 0.18398632556200029,
      "num_tokens": 80898247.0,
      "step": 36915
    },
    {
      "entropy": 5.683212041854858,
      "epoch": 3.664152441445018,
      "grad_norm": 1.125,
      "learning_rate": 0.00037074262987726013,
      "loss": 4.9255,
      "mean_token_accuracy": 0.19553124010562897,
      "num_tokens": 80908086.0,
      "step": 36920
    },
    {
      "entropy": 5.664433860778809,
      "epoch": 3.6646486701071854,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003707105650392451,
      "loss": 4.9163,
      "mean_token_accuracy": 0.19678571373224257,
      "num_tokens": 80918607.0,
      "step": 36925
    },
    {
      "entropy": 5.688886976242065,
      "epoch": 3.665144898769353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003706784978278144,
      "loss": 5.0539,
      "mean_token_accuracy": 0.1879463091492653,
      "num_tokens": 80930480.0,
      "step": 36930
    },
    {
      "entropy": 5.718728399276733,
      "epoch": 3.6656411274315204,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003706464282437629,
      "loss": 4.9913,
      "mean_token_accuracy": 0.1874854952096939,
      "num_tokens": 80941090.0,
      "step": 36935
    },
    {
      "entropy": 5.71385440826416,
      "epoch": 3.666137356093688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003706143562878863,
      "loss": 5.0767,
      "mean_token_accuracy": 0.18188907206058502,
      "num_tokens": 80952307.0,
      "step": 36940
    },
    {
      "entropy": 5.668071651458741,
      "epoch": 3.6666335847558553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003705822819609796,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18235420137643815,
      "num_tokens": 80962397.0,
      "step": 36945
    },
    {
      "entropy": 5.639870595932007,
      "epoch": 3.667129813418023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003705502052638383,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18745334893465043,
      "num_tokens": 80973581.0,
      "step": 36950
    },
    {
      "entropy": 5.636158037185669,
      "epoch": 3.6676260420801907,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037051812619725785,
      "loss": 4.9449,
      "mean_token_accuracy": 0.19609642773866653,
      "num_tokens": 80984915.0,
      "step": 36955
    },
    {
      "entropy": 5.689930486679077,
      "epoch": 3.668122270742358,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003704860447620335,
      "loss": 5.0818,
      "mean_token_accuracy": 0.18255483210086823,
      "num_tokens": 80996776.0,
      "step": 36960
    },
    {
      "entropy": 5.744790744781494,
      "epoch": 3.6686184994045257,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003704539609589612,
      "loss": 5.0441,
      "mean_token_accuracy": 0.186337248980999,
      "num_tokens": 81008334.0,
      "step": 36965
    },
    {
      "entropy": 5.672661304473877,
      "epoch": 3.669114728066693,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003704218747888363,
      "loss": 4.9254,
      "mean_token_accuracy": 0.19127121567726135,
      "num_tokens": 81018338.0,
      "step": 36970
    },
    {
      "entropy": 5.684189558029175,
      "epoch": 3.6696109567288606,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037038978625245447,
      "loss": 5.0017,
      "mean_token_accuracy": 0.1860604852437973,
      "num_tokens": 81028188.0,
      "step": 36975
    },
    {
      "entropy": 5.61025562286377,
      "epoch": 3.670107185391028,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003703576953506115,
      "loss": 4.9572,
      "mean_token_accuracy": 0.18783017247915268,
      "num_tokens": 81038633.0,
      "step": 36980
    },
    {
      "entropy": 5.541227626800537,
      "epoch": 3.6706034140531956,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037032560208410313,
      "loss": 4.9169,
      "mean_token_accuracy": 0.1896779477596283,
      "num_tokens": 81049749.0,
      "step": 36985
    },
    {
      "entropy": 5.6084705829620365,
      "epoch": 3.6710996427153635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037029350645372526,
      "loss": 4.9322,
      "mean_token_accuracy": 0.194322469830513,
      "num_tokens": 81061555.0,
      "step": 36990
    },
    {
      "entropy": 5.683282899856567,
      "epoch": 3.671595871377531,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037026140846027383,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18378870338201522,
      "num_tokens": 81072233.0,
      "step": 36995
    },
    {
      "entropy": 5.662827873229981,
      "epoch": 3.6720921000396984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003702293081045448,
      "loss": 4.947,
      "mean_token_accuracy": 0.18789568841457366,
      "num_tokens": 81083835.0,
      "step": 37000
    },
    {
      "entropy": 5.741730165481568,
      "epoch": 3.672588328701866,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037019720538733394,
      "loss": 5.0704,
      "mean_token_accuracy": 0.18067402988672257,
      "num_tokens": 81094689.0,
      "step": 37005
    },
    {
      "entropy": 5.6266845703125,
      "epoch": 3.6730845573640334,
      "grad_norm": 1.125,
      "learning_rate": 0.0003701651003094376,
      "loss": 4.9758,
      "mean_token_accuracy": 0.19087432473897933,
      "num_tokens": 81105287.0,
      "step": 37010
    },
    {
      "entropy": 5.61785101890564,
      "epoch": 3.673580786026201,
      "grad_norm": 1.25,
      "learning_rate": 0.0003701329928716519,
      "loss": 4.9273,
      "mean_token_accuracy": 0.19372199475765228,
      "num_tokens": 81114859.0,
      "step": 37015
    },
    {
      "entropy": 5.709231281280518,
      "epoch": 3.6740770146883683,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037010088307477286,
      "loss": 5.0274,
      "mean_token_accuracy": 0.19365298002958298,
      "num_tokens": 81126574.0,
      "step": 37020
    },
    {
      "entropy": 5.696955347061158,
      "epoch": 3.674573243350536,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003700687709195969,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18561843484640123,
      "num_tokens": 81136957.0,
      "step": 37025
    },
    {
      "entropy": 5.641443061828613,
      "epoch": 3.6750694720127033,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003700366564069203,
      "loss": 4.9745,
      "mean_token_accuracy": 0.18382325172424316,
      "num_tokens": 81147224.0,
      "step": 37030
    },
    {
      "entropy": 5.620447778701783,
      "epoch": 3.6755657006748708,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037000453953753943,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18522575348615647,
      "num_tokens": 81159107.0,
      "step": 37035
    },
    {
      "entropy": 5.681623077392578,
      "epoch": 3.6760619293370382,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003699724203122506,
      "loss": 5.0162,
      "mean_token_accuracy": 0.18912388235330582,
      "num_tokens": 81169424.0,
      "step": 37040
    },
    {
      "entropy": 5.694075632095337,
      "epoch": 3.676558157999206,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003699402987318504,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18749355375766755,
      "num_tokens": 81180056.0,
      "step": 37045
    },
    {
      "entropy": 5.672845554351807,
      "epoch": 3.6770543866613736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036990817479713543,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18563733696937562,
      "num_tokens": 81191369.0,
      "step": 37050
    },
    {
      "entropy": 5.7061375141143795,
      "epoch": 3.677550615323541,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036987604850890224,
      "loss": 5.0948,
      "mean_token_accuracy": 0.18233601599931717,
      "num_tokens": 81203015.0,
      "step": 37055
    },
    {
      "entropy": 5.709108209609985,
      "epoch": 3.6780468439857086,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036984391986794745,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18535896092653276,
      "num_tokens": 81214219.0,
      "step": 37060
    },
    {
      "entropy": 5.597890996932984,
      "epoch": 3.678543072647876,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036981178887506776,
      "loss": 5.004,
      "mean_token_accuracy": 0.1849323406815529,
      "num_tokens": 81225164.0,
      "step": 37065
    },
    {
      "entropy": 5.695930480957031,
      "epoch": 3.6790393013100435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036977965553106014,
      "loss": 4.9546,
      "mean_token_accuracy": 0.1881064474582672,
      "num_tokens": 81235435.0,
      "step": 37070
    },
    {
      "entropy": 5.696724605560303,
      "epoch": 3.6795355299722114,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003697475198367212,
      "loss": 5.0919,
      "mean_token_accuracy": 0.18596115112304687,
      "num_tokens": 81245044.0,
      "step": 37075
    },
    {
      "entropy": 5.632186651229858,
      "epoch": 3.680031758634379,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000369715381792848,
      "loss": 4.9529,
      "mean_token_accuracy": 0.1903688371181488,
      "num_tokens": 81255651.0,
      "step": 37080
    },
    {
      "entropy": 5.644629335403442,
      "epoch": 3.6805279872965464,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003696832414002374,
      "loss": 4.9227,
      "mean_token_accuracy": 0.1966767653822899,
      "num_tokens": 81265430.0,
      "step": 37085
    },
    {
      "entropy": 5.574047565460205,
      "epoch": 3.681024215958714,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003696510986596864,
      "loss": 4.9048,
      "mean_token_accuracy": 0.19084521532058715,
      "num_tokens": 81275746.0,
      "step": 37090
    },
    {
      "entropy": 5.700339412689209,
      "epoch": 3.6815204446208813,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036961895357199214,
      "loss": 5.092,
      "mean_token_accuracy": 0.18032692074775697,
      "num_tokens": 81286202.0,
      "step": 37095
    },
    {
      "entropy": 5.647034406661987,
      "epoch": 3.682016673283049,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003695868061379518,
      "loss": 4.9924,
      "mean_token_accuracy": 0.1807480663061142,
      "num_tokens": 81297479.0,
      "step": 37100
    },
    {
      "entropy": 5.738335371017456,
      "epoch": 3.6825129019452163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003695546563583623,
      "loss": 5.0252,
      "mean_token_accuracy": 0.1874023973941803,
      "num_tokens": 81310702.0,
      "step": 37105
    },
    {
      "entropy": 5.658018159866333,
      "epoch": 3.6830091306073838,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003695225042340212,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18563095182180406,
      "num_tokens": 81322330.0,
      "step": 37110
    },
    {
      "entropy": 5.652082633972168,
      "epoch": 3.6835053592695512,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003694903497657256,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1921636164188385,
      "num_tokens": 81334308.0,
      "step": 37115
    },
    {
      "entropy": 5.74341983795166,
      "epoch": 3.6840015879317187,
      "grad_norm": 1.1875,
      "learning_rate": 0.000369458192954273,
      "loss": 5.0946,
      "mean_token_accuracy": 0.18166755139827728,
      "num_tokens": 81345467.0,
      "step": 37120
    },
    {
      "entropy": 5.7490232467651365,
      "epoch": 3.684497816593886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003694260338004607,
      "loss": 5.0832,
      "mean_token_accuracy": 0.18630111813545228,
      "num_tokens": 81357106.0,
      "step": 37125
    },
    {
      "entropy": 5.684794092178345,
      "epoch": 3.684994045256054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036939387230508627,
      "loss": 4.929,
      "mean_token_accuracy": 0.19610731899738312,
      "num_tokens": 81367669.0,
      "step": 37130
    },
    {
      "entropy": 5.642946672439575,
      "epoch": 3.6854902739182216,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036936170846894715,
      "loss": 4.9918,
      "mean_token_accuracy": 0.19320001900196077,
      "num_tokens": 81377891.0,
      "step": 37135
    },
    {
      "entropy": 5.696049547195434,
      "epoch": 3.685986502580389,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036932954229284106,
      "loss": 5.021,
      "mean_token_accuracy": 0.18424781113862992,
      "num_tokens": 81387526.0,
      "step": 37140
    },
    {
      "entropy": 5.625890684127808,
      "epoch": 3.6864827312425565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036929737377756563,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2001107856631279,
      "num_tokens": 81398167.0,
      "step": 37145
    },
    {
      "entropy": 5.660914421081543,
      "epoch": 3.686978959904724,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003692652029239184,
      "loss": 4.984,
      "mean_token_accuracy": 0.19221411943435668,
      "num_tokens": 81408740.0,
      "step": 37150
    },
    {
      "entropy": 5.671273040771484,
      "epoch": 3.6874751885668915,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036923302973269727,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1925221264362335,
      "num_tokens": 81419336.0,
      "step": 37155
    },
    {
      "entropy": 5.65339298248291,
      "epoch": 3.6879714172290594,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003692008542047001,
      "loss": 4.9443,
      "mean_token_accuracy": 0.194902902841568,
      "num_tokens": 81429459.0,
      "step": 37160
    },
    {
      "entropy": 5.733927917480469,
      "epoch": 3.688467645891227,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003691686763407248,
      "loss": 5.0638,
      "mean_token_accuracy": 0.1802749365568161,
      "num_tokens": 81440650.0,
      "step": 37165
    },
    {
      "entropy": 5.675605010986328,
      "epoch": 3.6889638745533944,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003691364961415691,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18849564343690872,
      "num_tokens": 81453288.0,
      "step": 37170
    },
    {
      "entropy": 5.721328210830689,
      "epoch": 3.689460103215562,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036910431360803124,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18918945491313935,
      "num_tokens": 81463490.0,
      "step": 37175
    },
    {
      "entropy": 5.657641363143921,
      "epoch": 3.6899563318777293,
      "grad_norm": 1.25,
      "learning_rate": 0.00036907212874090923,
      "loss": 4.966,
      "mean_token_accuracy": 0.1928585261106491,
      "num_tokens": 81474585.0,
      "step": 37180
    },
    {
      "entropy": 5.6827645778656,
      "epoch": 3.690452560539897,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000369039941541001,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1889292523264885,
      "num_tokens": 81485763.0,
      "step": 37185
    },
    {
      "entropy": 5.6537665843963625,
      "epoch": 3.6909487892020643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036900775200910497,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18408945500850676,
      "num_tokens": 81497821.0,
      "step": 37190
    },
    {
      "entropy": 5.734054946899414,
      "epoch": 3.6914450178642317,
      "grad_norm": 1.125,
      "learning_rate": 0.00036897556014601926,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18568665385246277,
      "num_tokens": 81509793.0,
      "step": 37195
    },
    {
      "entropy": 5.679269504547119,
      "epoch": 3.691941246526399,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036894336595254203,
      "loss": 4.964,
      "mean_token_accuracy": 0.19682406038045883,
      "num_tokens": 81521138.0,
      "step": 37200
    },
    {
      "entropy": 5.6283046245574955,
      "epoch": 3.6924374751885667,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036891116942947184,
      "loss": 4.9,
      "mean_token_accuracy": 0.19416973292827605,
      "num_tokens": 81530964.0,
      "step": 37205
    },
    {
      "entropy": 5.644494533538818,
      "epoch": 3.692933703850734,
      "grad_norm": 1.140625,
      "learning_rate": 0.000368878970577607,
      "loss": 4.9414,
      "mean_token_accuracy": 0.18632725328207017,
      "num_tokens": 81540775.0,
      "step": 37210
    },
    {
      "entropy": 5.642445993423462,
      "epoch": 3.693429932512902,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036884676939774597,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18721399903297425,
      "num_tokens": 81551509.0,
      "step": 37215
    },
    {
      "entropy": 5.682804250717163,
      "epoch": 3.6939261611750696,
      "grad_norm": 1.125,
      "learning_rate": 0.00036881456589068725,
      "loss": 5.0537,
      "mean_token_accuracy": 0.1925668850541115,
      "num_tokens": 81563217.0,
      "step": 37220
    },
    {
      "entropy": 5.709470939636231,
      "epoch": 3.694422389837237,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036878236005722954,
      "loss": 4.9114,
      "mean_token_accuracy": 0.19537943750619888,
      "num_tokens": 81574604.0,
      "step": 37225
    },
    {
      "entropy": 5.662421321868896,
      "epoch": 3.6949186184994045,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003687501518981713,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18780999332666398,
      "num_tokens": 81585575.0,
      "step": 37230
    },
    {
      "entropy": 5.6371338844299315,
      "epoch": 3.695414847161572,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003687179414143114,
      "loss": 4.9518,
      "mean_token_accuracy": 0.18485481441020965,
      "num_tokens": 81596114.0,
      "step": 37235
    },
    {
      "entropy": 5.652181720733642,
      "epoch": 3.6959110758237395,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036868572860644845,
      "loss": 4.9571,
      "mean_token_accuracy": 0.1858421415090561,
      "num_tokens": 81606380.0,
      "step": 37240
    },
    {
      "entropy": 5.6950983047485355,
      "epoch": 3.6964073044859074,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036865351347538123,
      "loss": 5.0177,
      "mean_token_accuracy": 0.17659500986337662,
      "num_tokens": 81615977.0,
      "step": 37245
    },
    {
      "entropy": 5.6465376853942875,
      "epoch": 3.696903533148075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003686212960219088,
      "loss": 4.9351,
      "mean_token_accuracy": 0.19294434934854507,
      "num_tokens": 81627505.0,
      "step": 37250
    },
    {
      "entropy": 5.6027219772338865,
      "epoch": 3.6973997618102423,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036858907624682987,
      "loss": 4.9302,
      "mean_token_accuracy": 0.19555899798870086,
      "num_tokens": 81637486.0,
      "step": 37255
    },
    {
      "entropy": 5.657325315475464,
      "epoch": 3.69789599047241,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003685568541509435,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19335216879844666,
      "num_tokens": 81648121.0,
      "step": 37260
    },
    {
      "entropy": 5.675260257720947,
      "epoch": 3.6983922191345773,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036852462973504884,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18788218796253203,
      "num_tokens": 81660468.0,
      "step": 37265
    },
    {
      "entropy": 5.679706621170044,
      "epoch": 3.6988884477967447,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003684924029999449,
      "loss": 4.9796,
      "mean_token_accuracy": 0.19242137521505356,
      "num_tokens": 81672516.0,
      "step": 37270
    },
    {
      "entropy": 5.677169132232666,
      "epoch": 3.6993846764589122,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036846017394643064,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2038411632180214,
      "num_tokens": 81683233.0,
      "step": 37275
    },
    {
      "entropy": 5.724870824813843,
      "epoch": 3.6998809051210797,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003684279425753057,
      "loss": 5.0071,
      "mean_token_accuracy": 0.19331008642911912,
      "num_tokens": 81694098.0,
      "step": 37280
    },
    {
      "entropy": 5.655303192138672,
      "epoch": 3.700377133783247,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036839570888736893,
      "loss": 4.9808,
      "mean_token_accuracy": 0.19104381203651427,
      "num_tokens": 81704453.0,
      "step": 37285
    },
    {
      "entropy": 5.680058431625366,
      "epoch": 3.7008733624454146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003683634728834199,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18391195088624954,
      "num_tokens": 81715041.0,
      "step": 37290
    },
    {
      "entropy": 5.659423065185547,
      "epoch": 3.701369591107582,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036833123456425794,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18479555547237397,
      "num_tokens": 81726453.0,
      "step": 37295
    },
    {
      "entropy": 5.676972389221191,
      "epoch": 3.70186581976975,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003682989939306824,
      "loss": 4.946,
      "mean_token_accuracy": 0.1953882172703743,
      "num_tokens": 81737268.0,
      "step": 37300
    },
    {
      "entropy": 5.698762369155884,
      "epoch": 3.7023620484319175,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003682667509834929,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18669929206371308,
      "num_tokens": 81749126.0,
      "step": 37305
    },
    {
      "entropy": 5.627850341796875,
      "epoch": 3.702858277094085,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036823450572348896,
      "loss": 4.9411,
      "mean_token_accuracy": 0.19545471668243408,
      "num_tokens": 81760305.0,
      "step": 37310
    },
    {
      "entropy": 5.758215665817261,
      "epoch": 3.7033545057562525,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003682022581514702,
      "loss": 5.0793,
      "mean_token_accuracy": 0.17958801686763765,
      "num_tokens": 81771700.0,
      "step": 37315
    },
    {
      "entropy": 5.658477306365967,
      "epoch": 3.70385073441842,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036817000826823616,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1917449101805687,
      "num_tokens": 81781431.0,
      "step": 37320
    },
    {
      "entropy": 5.737563276290894,
      "epoch": 3.7043469630805874,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036813775607458677,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18655088394880295,
      "num_tokens": 81792068.0,
      "step": 37325
    },
    {
      "entropy": 5.749075651168823,
      "epoch": 3.7048431917427553,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003681055015713217,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18510781228542328,
      "num_tokens": 81804311.0,
      "step": 37330
    },
    {
      "entropy": 5.708848714828491,
      "epoch": 3.705339420404923,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003680732447592408,
      "loss": 4.9387,
      "mean_token_accuracy": 0.19108337461948394,
      "num_tokens": 81814505.0,
      "step": 37335
    },
    {
      "entropy": 5.673727703094483,
      "epoch": 3.7058356490670903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036804098563914407,
      "loss": 5.0031,
      "mean_token_accuracy": 0.18456272929906845,
      "num_tokens": 81823960.0,
      "step": 37340
    },
    {
      "entropy": 5.663174057006836,
      "epoch": 3.7063318777292578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003680087242118313,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18462290465831757,
      "num_tokens": 81835086.0,
      "step": 37345
    },
    {
      "entropy": 5.7106201171875,
      "epoch": 3.7068281063914252,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003679764604781026,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18521535247564316,
      "num_tokens": 81845316.0,
      "step": 37350
    },
    {
      "entropy": 5.660580205917358,
      "epoch": 3.7073243350535927,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036794419443875805,
      "loss": 4.9836,
      "mean_token_accuracy": 0.18780554831027985,
      "num_tokens": 81856670.0,
      "step": 37355
    },
    {
      "entropy": 5.671510934829712,
      "epoch": 3.70782056371576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003679119260945977,
      "loss": 5.0466,
      "mean_token_accuracy": 0.1856653332710266,
      "num_tokens": 81869718.0,
      "step": 37360
    },
    {
      "entropy": 5.696794748306274,
      "epoch": 3.7083167923779277,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003678796554464219,
      "loss": 5.0579,
      "mean_token_accuracy": 0.19193999618291854,
      "num_tokens": 81879408.0,
      "step": 37365
    },
    {
      "entropy": 5.711381435394287,
      "epoch": 3.708813021040095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036784738249503063,
      "loss": 5.0465,
      "mean_token_accuracy": 0.1836067721247673,
      "num_tokens": 81889234.0,
      "step": 37370
    },
    {
      "entropy": 5.650112962722778,
      "epoch": 3.7093092497022626,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003678151072412245,
      "loss": 4.9738,
      "mean_token_accuracy": 0.18870442658662795,
      "num_tokens": 81898732.0,
      "step": 37375
    },
    {
      "entropy": 5.698671436309814,
      "epoch": 3.70980547836443,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036778282968580355,
      "loss": 5.0572,
      "mean_token_accuracy": 0.1872065156698227,
      "num_tokens": 81908524.0,
      "step": 37380
    },
    {
      "entropy": 5.690592527389526,
      "epoch": 3.710301707026598,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003677505498295685,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18340187817811965,
      "num_tokens": 81919750.0,
      "step": 37385
    },
    {
      "entropy": 5.666464710235596,
      "epoch": 3.7107979356887655,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036771826767331955,
      "loss": 4.9166,
      "mean_token_accuracy": 0.1947401851415634,
      "num_tokens": 81929799.0,
      "step": 37390
    },
    {
      "entropy": 5.690576982498169,
      "epoch": 3.711294164350933,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003676859832178575,
      "loss": 5.0618,
      "mean_token_accuracy": 0.18457115441560745,
      "num_tokens": 81941431.0,
      "step": 37395
    },
    {
      "entropy": 5.685776329040527,
      "epoch": 3.7117903930131004,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003676536964639827,
      "loss": 5.063,
      "mean_token_accuracy": 0.17900439202785492,
      "num_tokens": 81953113.0,
      "step": 37400
    },
    {
      "entropy": 5.650642251968383,
      "epoch": 3.712286621675268,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003676214074124958,
      "loss": 4.9741,
      "mean_token_accuracy": 0.1895705431699753,
      "num_tokens": 81963838.0,
      "step": 37405
    },
    {
      "entropy": 5.667581081390381,
      "epoch": 3.7127828503374354,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036758911606419773,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18413926661014557,
      "num_tokens": 81975349.0,
      "step": 37410
    },
    {
      "entropy": 5.731784439086914,
      "epoch": 3.713279078999603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003675568224198889,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18498398512601852,
      "num_tokens": 81985771.0,
      "step": 37415
    },
    {
      "entropy": 5.704026985168457,
      "epoch": 3.7137753076617708,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003675245264803705,
      "loss": 4.9913,
      "mean_token_accuracy": 0.19221652299165726,
      "num_tokens": 81996989.0,
      "step": 37420
    },
    {
      "entropy": 5.650719070434571,
      "epoch": 3.7142715363239382,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003674922282464432,
      "loss": 4.95,
      "mean_token_accuracy": 0.186215078830719,
      "num_tokens": 82008213.0,
      "step": 37425
    },
    {
      "entropy": 5.642379713058472,
      "epoch": 3.7147677649861057,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003674599277189078,
      "loss": 4.9972,
      "mean_token_accuracy": 0.1864299237728119,
      "num_tokens": 82019250.0,
      "step": 37430
    },
    {
      "entropy": 5.643792533874512,
      "epoch": 3.715263993648273,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036742762489856554,
      "loss": 4.9592,
      "mean_token_accuracy": 0.18848781138658524,
      "num_tokens": 82030752.0,
      "step": 37435
    },
    {
      "entropy": 5.7009118556976315,
      "epoch": 3.7157602223104407,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003673953197862173,
      "loss": 5.0108,
      "mean_token_accuracy": 0.19019654393196106,
      "num_tokens": 82041119.0,
      "step": 37440
    },
    {
      "entropy": 5.684094953536987,
      "epoch": 3.716256450972608,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036736301238266434,
      "loss": 5.0192,
      "mean_token_accuracy": 0.1867092341184616,
      "num_tokens": 82054358.0,
      "step": 37445
    },
    {
      "entropy": 5.697340297698974,
      "epoch": 3.7167526796347756,
      "grad_norm": 1.125,
      "learning_rate": 0.00036733070268870754,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18862800747156144,
      "num_tokens": 82066009.0,
      "step": 37450
    },
    {
      "entropy": 5.6154145240783695,
      "epoch": 3.717248908296943,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003672983907051483,
      "loss": 4.9177,
      "mean_token_accuracy": 0.19129274189472198,
      "num_tokens": 82076704.0,
      "step": 37455
    },
    {
      "entropy": 5.622593784332276,
      "epoch": 3.7177451369591106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036726607643278793,
      "loss": 4.9364,
      "mean_token_accuracy": 0.1848321184515953,
      "num_tokens": 82086445.0,
      "step": 37460
    },
    {
      "entropy": 5.671693468093872,
      "epoch": 3.718241365621278,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036723375987242763,
      "loss": 5.0232,
      "mean_token_accuracy": 0.19339277297258378,
      "num_tokens": 82097996.0,
      "step": 37465
    },
    {
      "entropy": 5.63032112121582,
      "epoch": 3.718737594283446,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036720144102486886,
      "loss": 4.9548,
      "mean_token_accuracy": 0.191514553129673,
      "num_tokens": 82108463.0,
      "step": 37470
    },
    {
      "entropy": 5.670805740356445,
      "epoch": 3.7192338229456134,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036716911989091304,
      "loss": 5.0168,
      "mean_token_accuracy": 0.1846352458000183,
      "num_tokens": 82118840.0,
      "step": 37475
    },
    {
      "entropy": 5.694333934783936,
      "epoch": 3.719730051607781,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036713679647136163,
      "loss": 4.9468,
      "mean_token_accuracy": 0.1983418270945549,
      "num_tokens": 82129240.0,
      "step": 37480
    },
    {
      "entropy": 5.7152848720550535,
      "epoch": 3.7202262802699484,
      "grad_norm": 1.25,
      "learning_rate": 0.0003671044707670161,
      "loss": 4.9893,
      "mean_token_accuracy": 0.19498854577541352,
      "num_tokens": 82139289.0,
      "step": 37485
    },
    {
      "entropy": 5.608434867858887,
      "epoch": 3.720722508932116,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036707214277867834,
      "loss": 4.8266,
      "mean_token_accuracy": 0.1999206468462944,
      "num_tokens": 82149103.0,
      "step": 37490
    },
    {
      "entropy": 5.638242101669311,
      "epoch": 3.7212187375942833,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003670398125071497,
      "loss": 4.9909,
      "mean_token_accuracy": 0.1896478682756424,
      "num_tokens": 82160527.0,
      "step": 37495
    },
    {
      "entropy": 5.6951006889343265,
      "epoch": 3.721714966256451,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036700747995323214,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1831248566508293,
      "num_tokens": 82171929.0,
      "step": 37500
    },
    {
      "entropy": 5.726068162918091,
      "epoch": 3.7222111949186187,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003669751451177273,
      "loss": 5.0722,
      "mean_token_accuracy": 0.19077035188674926,
      "num_tokens": 82183301.0,
      "step": 37505
    },
    {
      "entropy": 5.644747018814087,
      "epoch": 3.722707423580786,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000366942808001437,
      "loss": 4.8802,
      "mean_token_accuracy": 0.19855359345674514,
      "num_tokens": 82192719.0,
      "step": 37510
    },
    {
      "entropy": 5.694571781158447,
      "epoch": 3.7232036522429537,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036691046860516327,
      "loss": 5.0617,
      "mean_token_accuracy": 0.18786577582359315,
      "num_tokens": 82203842.0,
      "step": 37515
    },
    {
      "entropy": 5.714505290985107,
      "epoch": 3.723699880905121,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036687812692970796,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18520641326904297,
      "num_tokens": 82215280.0,
      "step": 37520
    },
    {
      "entropy": 5.697420358657837,
      "epoch": 3.7241961095672886,
      "grad_norm": 1.234375,
      "learning_rate": 0.000366845782975873,
      "loss": 5.035,
      "mean_token_accuracy": 0.17907359600067138,
      "num_tokens": 82226608.0,
      "step": 37525
    },
    {
      "entropy": 5.6792961120605465,
      "epoch": 3.724692338229456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036681343674446054,
      "loss": 4.9595,
      "mean_token_accuracy": 0.19285061210393906,
      "num_tokens": 82236186.0,
      "step": 37530
    },
    {
      "entropy": 5.697555065155029,
      "epoch": 3.7251885668916236,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003667810882362728,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1842886209487915,
      "num_tokens": 82247999.0,
      "step": 37535
    },
    {
      "entropy": 5.6488203525543215,
      "epoch": 3.725684795553791,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003667487374521117,
      "loss": 4.8895,
      "mean_token_accuracy": 0.1953355923295021,
      "num_tokens": 82258005.0,
      "step": 37540
    },
    {
      "entropy": 5.699429368972778,
      "epoch": 3.7261810242159585,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003667163843927797,
      "loss": 4.9861,
      "mean_token_accuracy": 0.19690550565719606,
      "num_tokens": 82269935.0,
      "step": 37545
    },
    {
      "entropy": 5.627181196212769,
      "epoch": 3.726677252878126,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036668402905907897,
      "loss": 4.9262,
      "mean_token_accuracy": 0.1925737276673317,
      "num_tokens": 82279480.0,
      "step": 37550
    },
    {
      "entropy": 5.664084339141846,
      "epoch": 3.727173481540294,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036665167145181186,
      "loss": 5.0176,
      "mean_token_accuracy": 0.18660940378904342,
      "num_tokens": 82291059.0,
      "step": 37555
    },
    {
      "entropy": 5.689393997192383,
      "epoch": 3.7276697102024614,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036661931157178085,
      "loss": 4.949,
      "mean_token_accuracy": 0.189563612639904,
      "num_tokens": 82301613.0,
      "step": 37560
    },
    {
      "entropy": 5.685317611694336,
      "epoch": 3.728165938864629,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00036658694941978825,
      "loss": 4.9125,
      "mean_token_accuracy": 0.19034821838140487,
      "num_tokens": 82311106.0,
      "step": 37565
    },
    {
      "entropy": 5.593799209594726,
      "epoch": 3.7286621675267964,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036655458499663675,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2025584489107132,
      "num_tokens": 82319997.0,
      "step": 37570
    },
    {
      "entropy": 5.668481922149658,
      "epoch": 3.729158396188964,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003665222183031288,
      "loss": 4.9566,
      "mean_token_accuracy": 0.19344426989555358,
      "num_tokens": 82331322.0,
      "step": 37575
    },
    {
      "entropy": 5.640417957305909,
      "epoch": 3.7296546248511313,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000366489849340067,
      "loss": 4.9554,
      "mean_token_accuracy": 0.1959759324789047,
      "num_tokens": 82342231.0,
      "step": 37580
    },
    {
      "entropy": 5.672274780273438,
      "epoch": 3.730150853513299,
      "grad_norm": 1.25,
      "learning_rate": 0.00036645747810825415,
      "loss": 4.9544,
      "mean_token_accuracy": 0.1883167415857315,
      "num_tokens": 82352656.0,
      "step": 37585
    },
    {
      "entropy": 5.6585277080535885,
      "epoch": 3.7306470821754667,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036642510460849285,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18534813523292543,
      "num_tokens": 82363855.0,
      "step": 37590
    },
    {
      "entropy": 5.694939804077149,
      "epoch": 3.731143310837634,
      "grad_norm": 1.1875,
      "learning_rate": 0.000366392728841586,
      "loss": 5.0189,
      "mean_token_accuracy": 0.18782471865415573,
      "num_tokens": 82374220.0,
      "step": 37595
    },
    {
      "entropy": 5.65072455406189,
      "epoch": 3.7316395394998016,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003663603508083364,
      "loss": 4.948,
      "mean_token_accuracy": 0.18332864642143248,
      "num_tokens": 82383838.0,
      "step": 37600
    },
    {
      "entropy": 5.64559907913208,
      "epoch": 3.732135768161969,
      "grad_norm": 1.15625,
      "learning_rate": 0.000366327970509547,
      "loss": 4.9815,
      "mean_token_accuracy": 0.19033066034317017,
      "num_tokens": 82396016.0,
      "step": 37605
    },
    {
      "entropy": 5.666850566864014,
      "epoch": 3.7326319968241366,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036629558794602063,
      "loss": 4.9515,
      "mean_token_accuracy": 0.1956949606537819,
      "num_tokens": 82405780.0,
      "step": 37610
    },
    {
      "entropy": 5.667845726013184,
      "epoch": 3.733128225486304,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036626320311856046,
      "loss": 4.9914,
      "mean_token_accuracy": 0.19154042154550552,
      "num_tokens": 82417171.0,
      "step": 37615
    },
    {
      "entropy": 5.650736951828003,
      "epoch": 3.7336244541484715,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036623081602796953,
      "loss": 4.955,
      "mean_token_accuracy": 0.18852079212665557,
      "num_tokens": 82428303.0,
      "step": 37620
    },
    {
      "entropy": 5.618891286849975,
      "epoch": 3.734120682810639,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036619842667505096,
      "loss": 4.8507,
      "mean_token_accuracy": 0.1913638159632683,
      "num_tokens": 82438867.0,
      "step": 37625
    },
    {
      "entropy": 5.606649684906006,
      "epoch": 3.7346169114728065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036616603506060786,
      "loss": 4.9412,
      "mean_token_accuracy": 0.1956599995493889,
      "num_tokens": 82449192.0,
      "step": 37630
    },
    {
      "entropy": 5.66806001663208,
      "epoch": 3.735113140134974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003661336411854436,
      "loss": 5.0485,
      "mean_token_accuracy": 0.182854326069355,
      "num_tokens": 82461895.0,
      "step": 37635
    },
    {
      "entropy": 5.621023464202881,
      "epoch": 3.7356093687971414,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003661012450503613,
      "loss": 4.9504,
      "mean_token_accuracy": 0.18703857511281968,
      "num_tokens": 82472476.0,
      "step": 37640
    },
    {
      "entropy": 5.6733112812042235,
      "epoch": 3.7361055974593094,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036606884665616455,
      "loss": 5.0549,
      "mean_token_accuracy": 0.18326596766710282,
      "num_tokens": 82483513.0,
      "step": 37645
    },
    {
      "entropy": 5.730916452407837,
      "epoch": 3.736601826121477,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003660364460036566,
      "loss": 5.0552,
      "mean_token_accuracy": 0.17945589870214462,
      "num_tokens": 82495303.0,
      "step": 37650
    },
    {
      "entropy": 5.723868989944458,
      "epoch": 3.7370980547836443,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003660040430936409,
      "loss": 5.0085,
      "mean_token_accuracy": 0.19022577106952668,
      "num_tokens": 82506457.0,
      "step": 37655
    },
    {
      "entropy": 5.7104973793029785,
      "epoch": 3.737594283445812,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036597163792692107,
      "loss": 5.0077,
      "mean_token_accuracy": 0.18911317884922027,
      "num_tokens": 82517708.0,
      "step": 37660
    },
    {
      "entropy": 5.694022369384766,
      "epoch": 3.7380905121079793,
      "grad_norm": 1.125,
      "learning_rate": 0.00036593923050430067,
      "loss": 4.9712,
      "mean_token_accuracy": 0.19403426796197892,
      "num_tokens": 82529147.0,
      "step": 37665
    },
    {
      "entropy": 5.692704916000366,
      "epoch": 3.7385867407701467,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036590682082658334,
      "loss": 5.0031,
      "mean_token_accuracy": 0.18625324070453644,
      "num_tokens": 82541148.0,
      "step": 37670
    },
    {
      "entropy": 5.652039003372193,
      "epoch": 3.7390829694323147,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036587440889457274,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18509554862976074,
      "num_tokens": 82552006.0,
      "step": 37675
    },
    {
      "entropy": 5.675874710083008,
      "epoch": 3.739579198094482,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036584199470907254,
      "loss": 4.9192,
      "mean_token_accuracy": 0.1930118352174759,
      "num_tokens": 82562637.0,
      "step": 37680
    },
    {
      "entropy": 5.671412324905395,
      "epoch": 3.7400754267566496,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003658095782708866,
      "loss": 4.9768,
      "mean_token_accuracy": 0.1834199920296669,
      "num_tokens": 82573389.0,
      "step": 37685
    },
    {
      "entropy": 5.631641054153443,
      "epoch": 3.740571655418817,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036577715958081885,
      "loss": 4.9182,
      "mean_token_accuracy": 0.1954685479402542,
      "num_tokens": 82584432.0,
      "step": 37690
    },
    {
      "entropy": 5.6129967212677006,
      "epoch": 3.7410678840809846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003657447386396732,
      "loss": 4.9466,
      "mean_token_accuracy": 0.18616099655628204,
      "num_tokens": 82595111.0,
      "step": 37695
    },
    {
      "entropy": 5.641240739822388,
      "epoch": 3.741564112743152,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003657123154482534,
      "loss": 5.0214,
      "mean_token_accuracy": 0.18898987174034118,
      "num_tokens": 82605348.0,
      "step": 37700
    },
    {
      "entropy": 5.602966785430908,
      "epoch": 3.7420603414053195,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036567989000736386,
      "loss": 4.9012,
      "mean_token_accuracy": 0.1895786091685295,
      "num_tokens": 82616295.0,
      "step": 37705
    },
    {
      "entropy": 5.62997088432312,
      "epoch": 3.742556570067487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036564746231780836,
      "loss": 4.9507,
      "mean_token_accuracy": 0.190984083712101,
      "num_tokens": 82627939.0,
      "step": 37710
    },
    {
      "entropy": 5.655132579803467,
      "epoch": 3.7430527987296545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003656150323803911,
      "loss": 4.9767,
      "mean_token_accuracy": 0.1939985379576683,
      "num_tokens": 82640431.0,
      "step": 37715
    },
    {
      "entropy": 5.706171989440918,
      "epoch": 3.743549027391822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036558260019591633,
      "loss": 5.0429,
      "mean_token_accuracy": 0.1931784600019455,
      "num_tokens": 82651095.0,
      "step": 37720
    },
    {
      "entropy": 5.65667085647583,
      "epoch": 3.7440452560539894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036555016576518816,
      "loss": 4.9395,
      "mean_token_accuracy": 0.19129811376333236,
      "num_tokens": 82660588.0,
      "step": 37725
    },
    {
      "entropy": 5.711937618255615,
      "epoch": 3.7445414847161573,
      "grad_norm": 1.125,
      "learning_rate": 0.0003655177290890111,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1901412546634674,
      "num_tokens": 82673217.0,
      "step": 37730
    },
    {
      "entropy": 5.709276723861694,
      "epoch": 3.745037713378325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036548529016818936,
      "loss": 4.9848,
      "mean_token_accuracy": 0.19731387943029405,
      "num_tokens": 82685729.0,
      "step": 37735
    },
    {
      "entropy": 5.6886265754699705,
      "epoch": 3.7455339420404923,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003654528490035274,
      "loss": 5.0524,
      "mean_token_accuracy": 0.18788220435380937,
      "num_tokens": 82697375.0,
      "step": 37740
    },
    {
      "entropy": 5.673688077926636,
      "epoch": 3.7460301707026598,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003654204055958298,
      "loss": 4.9153,
      "mean_token_accuracy": 0.1914492055773735,
      "num_tokens": 82707247.0,
      "step": 37745
    },
    {
      "entropy": 5.667896842956543,
      "epoch": 3.7465263993648272,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036538795994590093,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18439458012580873,
      "num_tokens": 82717827.0,
      "step": 37750
    },
    {
      "entropy": 5.656870698928833,
      "epoch": 3.7470226280269947,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036535551205454535,
      "loss": 5.03,
      "mean_token_accuracy": 0.19111379683017732,
      "num_tokens": 82728518.0,
      "step": 37755
    },
    {
      "entropy": 5.685105991363526,
      "epoch": 3.7475188566891626,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036532306192256774,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18624524772167206,
      "num_tokens": 82739687.0,
      "step": 37760
    },
    {
      "entropy": 5.630584001541138,
      "epoch": 3.74801508535133,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003652906095507729,
      "loss": 4.9323,
      "mean_token_accuracy": 0.1907113417983055,
      "num_tokens": 82749753.0,
      "step": 37765
    },
    {
      "entropy": 5.690666151046753,
      "epoch": 3.7485113140134976,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036525815493996557,
      "loss": 5.0559,
      "mean_token_accuracy": 0.18133843094110488,
      "num_tokens": 82760245.0,
      "step": 37770
    },
    {
      "entropy": 5.707627010345459,
      "epoch": 3.749007542675665,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036522569809095034,
      "loss": 4.9687,
      "mean_token_accuracy": 0.19364688396453858,
      "num_tokens": 82769508.0,
      "step": 37775
    },
    {
      "entropy": 5.628523683547973,
      "epoch": 3.7495037713378325,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036519323900453235,
      "loss": 4.9409,
      "mean_token_accuracy": 0.19288617819547654,
      "num_tokens": 82780258.0,
      "step": 37780
    },
    {
      "entropy": 5.633483839035034,
      "epoch": 3.75,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036516077768151636,
      "loss": 4.9494,
      "mean_token_accuracy": 0.19329769611358644,
      "num_tokens": 82791723.0,
      "step": 37785
    },
    {
      "entropy": 5.64320650100708,
      "epoch": 3.7504962286621675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003651283141227073,
      "loss": 4.9478,
      "mean_token_accuracy": 0.19683863073587418,
      "num_tokens": 82803169.0,
      "step": 37790
    },
    {
      "entropy": 5.63462643623352,
      "epoch": 3.750992457324335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003650958483289102,
      "loss": 4.9843,
      "mean_token_accuracy": 0.19145914912223816,
      "num_tokens": 82814115.0,
      "step": 37795
    },
    {
      "entropy": 5.681160497665405,
      "epoch": 3.7514886859865024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003650633803009303,
      "loss": 4.9592,
      "mean_token_accuracy": 0.19731679260730745,
      "num_tokens": 82826157.0,
      "step": 37800
    },
    {
      "entropy": 5.7198539733886715,
      "epoch": 3.75198491464867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036503091003957264,
      "loss": 4.973,
      "mean_token_accuracy": 0.19666629582643508,
      "num_tokens": 82837660.0,
      "step": 37805
    },
    {
      "entropy": 5.6338463306427,
      "epoch": 3.7524811433108374,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036499843754564235,
      "loss": 4.9466,
      "mean_token_accuracy": 0.1924939051270485,
      "num_tokens": 82848563.0,
      "step": 37810
    },
    {
      "entropy": 5.669944715499878,
      "epoch": 3.7529773719730053,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003649659628199448,
      "loss": 4.9599,
      "mean_token_accuracy": 0.19537925124168395,
      "num_tokens": 82859242.0,
      "step": 37815
    },
    {
      "entropy": 5.707666444778442,
      "epoch": 3.7534736006351728,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003649334858632852,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1867919608950615,
      "num_tokens": 82873726.0,
      "step": 37820
    },
    {
      "entropy": 5.650525188446045,
      "epoch": 3.7539698292973402,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000364901006676469,
      "loss": 4.8788,
      "mean_token_accuracy": 0.19761032462120057,
      "num_tokens": 82884960.0,
      "step": 37825
    },
    {
      "entropy": 5.622111797332764,
      "epoch": 3.7544660579595077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036486852526030147,
      "loss": 4.9361,
      "mean_token_accuracy": 0.1903286412358284,
      "num_tokens": 82897133.0,
      "step": 37830
    },
    {
      "entropy": 5.655135154724121,
      "epoch": 3.754962286621675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003648360416155882,
      "loss": 4.9359,
      "mean_token_accuracy": 0.19301411509513855,
      "num_tokens": 82906451.0,
      "step": 37835
    },
    {
      "entropy": 5.635616397857666,
      "epoch": 3.7554585152838427,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003648035557431347,
      "loss": 4.99,
      "mean_token_accuracy": 0.19199062436819075,
      "num_tokens": 82919255.0,
      "step": 37840
    },
    {
      "entropy": 5.7838561058044435,
      "epoch": 3.7559547439460106,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00036477106764374646,
      "loss": 5.0409,
      "mean_token_accuracy": 0.19324130564928055,
      "num_tokens": 82929747.0,
      "step": 37845
    },
    {
      "entropy": 5.750202178955078,
      "epoch": 3.756450972608178,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003647385773182292,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1866442859172821,
      "num_tokens": 82941505.0,
      "step": 37850
    },
    {
      "entropy": 5.674809885025025,
      "epoch": 3.7569472012703455,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036470608476738865,
      "loss": 5.0403,
      "mean_token_accuracy": 0.18911677598953247,
      "num_tokens": 82951825.0,
      "step": 37855
    },
    {
      "entropy": 5.6045207500457765,
      "epoch": 3.757443429932513,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003646735899920304,
      "loss": 4.9291,
      "mean_token_accuracy": 0.20297119319438933,
      "num_tokens": 82963614.0,
      "step": 37860
    },
    {
      "entropy": 5.632596254348755,
      "epoch": 3.7579396585946805,
      "grad_norm": 1.34375,
      "learning_rate": 0.00036464109299296047,
      "loss": 4.9363,
      "mean_token_accuracy": 0.19216389656066896,
      "num_tokens": 82973380.0,
      "step": 37865
    },
    {
      "entropy": 5.701120901107788,
      "epoch": 3.758435887256848,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003646085937709845,
      "loss": 4.9387,
      "mean_token_accuracy": 0.19522634595632554,
      "num_tokens": 82984456.0,
      "step": 37870
    },
    {
      "entropy": 5.665710592269898,
      "epoch": 3.7589321159190154,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003645760923269085,
      "loss": 5.0211,
      "mean_token_accuracy": 0.1887674331665039,
      "num_tokens": 82995603.0,
      "step": 37875
    },
    {
      "entropy": 5.66741304397583,
      "epoch": 3.759428344581183,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003645435886615385,
      "loss": 5.0049,
      "mean_token_accuracy": 0.1943971812725067,
      "num_tokens": 83007205.0,
      "step": 37880
    },
    {
      "entropy": 5.6454479694366455,
      "epoch": 3.7599245732433504,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036451108277568044,
      "loss": 4.9625,
      "mean_token_accuracy": 0.19205400943756104,
      "num_tokens": 83017572.0,
      "step": 37885
    },
    {
      "entropy": 5.6679259777069095,
      "epoch": 3.760420801905518,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036447857467014036,
      "loss": 4.9556,
      "mean_token_accuracy": 0.18872372657060624,
      "num_tokens": 83029169.0,
      "step": 37890
    },
    {
      "entropy": 5.652847003936768,
      "epoch": 3.7609170305676853,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003644460643457245,
      "loss": 4.9107,
      "mean_token_accuracy": 0.19518080800771714,
      "num_tokens": 83038970.0,
      "step": 37895
    },
    {
      "entropy": 5.633633947372436,
      "epoch": 3.7614132592298533,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036441355180323885,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19008784294128417,
      "num_tokens": 83050458.0,
      "step": 37900
    },
    {
      "entropy": 5.6890785694122314,
      "epoch": 3.7619094878920207,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003643810370434899,
      "loss": 5.1128,
      "mean_token_accuracy": 0.18582409918308257,
      "num_tokens": 83063057.0,
      "step": 37905
    },
    {
      "entropy": 5.63180832862854,
      "epoch": 3.762405716554188,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003643485200672838,
      "loss": 4.9196,
      "mean_token_accuracy": 0.19810879230499268,
      "num_tokens": 83074839.0,
      "step": 37910
    },
    {
      "entropy": 5.624255037307739,
      "epoch": 3.7629019452163557,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036431600087542695,
      "loss": 4.9415,
      "mean_token_accuracy": 0.19611052423715591,
      "num_tokens": 83085015.0,
      "step": 37915
    },
    {
      "entropy": 5.623102855682373,
      "epoch": 3.763398173878523,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003642834794687257,
      "loss": 5.0341,
      "mean_token_accuracy": 0.177721805870533,
      "num_tokens": 83095986.0,
      "step": 37920
    },
    {
      "entropy": 5.610351657867431,
      "epoch": 3.7638944025406906,
      "grad_norm": 1.125,
      "learning_rate": 0.0003642509558479865,
      "loss": 4.8595,
      "mean_token_accuracy": 0.2004381999373436,
      "num_tokens": 83105984.0,
      "step": 37925
    },
    {
      "entropy": 5.663722181320191,
      "epoch": 3.7643906312028586,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036421843001401607,
      "loss": 4.9085,
      "mean_token_accuracy": 0.19226088076829911,
      "num_tokens": 83116206.0,
      "step": 37930
    },
    {
      "entropy": 5.628343820571899,
      "epoch": 3.764886859865026,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036418590196762076,
      "loss": 4.8764,
      "mean_token_accuracy": 0.19572391211986542,
      "num_tokens": 83128064.0,
      "step": 37935
    },
    {
      "entropy": 5.604950284957885,
      "epoch": 3.7653830885271935,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036415337170960714,
      "loss": 4.9831,
      "mean_token_accuracy": 0.18794176131486892,
      "num_tokens": 83138646.0,
      "step": 37940
    },
    {
      "entropy": 5.69577202796936,
      "epoch": 3.765879317189361,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036412083924078223,
      "loss": 4.9521,
      "mean_token_accuracy": 0.1882582426071167,
      "num_tokens": 83148257.0,
      "step": 37945
    },
    {
      "entropy": 5.660703516006469,
      "epoch": 3.7663755458515285,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003640883045619523,
      "loss": 4.9501,
      "mean_token_accuracy": 0.19282418191432954,
      "num_tokens": 83158599.0,
      "step": 37950
    },
    {
      "entropy": 5.717898559570313,
      "epoch": 3.766871774513696,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003640557676739245,
      "loss": 5.0548,
      "mean_token_accuracy": 0.18799517005681993,
      "num_tokens": 83169698.0,
      "step": 37955
    },
    {
      "entropy": 5.654041719436646,
      "epoch": 3.7673680031758634,
      "grad_norm": 1.375,
      "learning_rate": 0.0003640232285775056,
      "loss": 4.926,
      "mean_token_accuracy": 0.1937662348151207,
      "num_tokens": 83179595.0,
      "step": 37960
    },
    {
      "entropy": 5.704063367843628,
      "epoch": 3.767864231838031,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036399068727350245,
      "loss": 5.0724,
      "mean_token_accuracy": 0.1851897045969963,
      "num_tokens": 83190626.0,
      "step": 37965
    },
    {
      "entropy": 5.636579751968384,
      "epoch": 3.7683604605001984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036395814376272194,
      "loss": 4.9416,
      "mean_token_accuracy": 0.19548614770174028,
      "num_tokens": 83202727.0,
      "step": 37970
    },
    {
      "entropy": 5.662483406066895,
      "epoch": 3.768856689162366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003639255980459713,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18495540469884872,
      "num_tokens": 83213334.0,
      "step": 37975
    },
    {
      "entropy": 5.718568897247314,
      "epoch": 3.7693529178245333,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003638930501240572,
      "loss": 5.0911,
      "mean_token_accuracy": 0.18411993831396103,
      "num_tokens": 83224885.0,
      "step": 37980
    },
    {
      "entropy": 5.653214025497436,
      "epoch": 3.769849146486701,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036386049999778724,
      "loss": 4.9331,
      "mean_token_accuracy": 0.1982780858874321,
      "num_tokens": 83235362.0,
      "step": 37985
    },
    {
      "entropy": 5.646080732345581,
      "epoch": 3.7703453751488687,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036382794766796824,
      "loss": 4.9961,
      "mean_token_accuracy": 0.19086461663246154,
      "num_tokens": 83246681.0,
      "step": 37990
    },
    {
      "entropy": 5.705072498321533,
      "epoch": 3.770841603811036,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036379539313540757,
      "loss": 5.0284,
      "mean_token_accuracy": 0.19017633944749832,
      "num_tokens": 83257474.0,
      "step": 37995
    },
    {
      "entropy": 5.701637172698975,
      "epoch": 3.7713378324732036,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003637628364009125,
      "loss": 4.9935,
      "mean_token_accuracy": 0.18957825601100922,
      "num_tokens": 83267335.0,
      "step": 38000
    },
    {
      "entropy": 5.671524381637573,
      "epoch": 3.771834061135371,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036373027746529037,
      "loss": 4.9811,
      "mean_token_accuracy": 0.19237171113491058,
      "num_tokens": 83278904.0,
      "step": 38005
    },
    {
      "entropy": 5.676113843917847,
      "epoch": 3.7723302897975386,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003636977163293485,
      "loss": 4.9713,
      "mean_token_accuracy": 0.19024337977170944,
      "num_tokens": 83288732.0,
      "step": 38010
    },
    {
      "entropy": 5.70328426361084,
      "epoch": 3.772826518459706,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003636651529938944,
      "loss": 4.978,
      "mean_token_accuracy": 0.1867981106042862,
      "num_tokens": 83299419.0,
      "step": 38015
    },
    {
      "entropy": 5.726906967163086,
      "epoch": 3.773322747121874,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003636325874597356,
      "loss": 5.0934,
      "mean_token_accuracy": 0.18190231621265412,
      "num_tokens": 83310352.0,
      "step": 38020
    },
    {
      "entropy": 5.7063535213470455,
      "epoch": 3.7738189757840415,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036360001972767956,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1831543803215027,
      "num_tokens": 83321520.0,
      "step": 38025
    },
    {
      "entropy": 5.619263744354248,
      "epoch": 3.774315204446209,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036356744979853406,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18926881104707718,
      "num_tokens": 83332328.0,
      "step": 38030
    },
    {
      "entropy": 5.701247310638427,
      "epoch": 3.7748114331083764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036353487767310655,
      "loss": 5.0526,
      "mean_token_accuracy": 0.17993392199277877,
      "num_tokens": 83342465.0,
      "step": 38035
    },
    {
      "entropy": 5.69619345664978,
      "epoch": 3.775307661770544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003635023033522049,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18801768869161606,
      "num_tokens": 83353423.0,
      "step": 38040
    },
    {
      "entropy": 5.697909116744995,
      "epoch": 3.7758038904327114,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003634697268366368,
      "loss": 4.9295,
      "mean_token_accuracy": 0.19253412932157515,
      "num_tokens": 83364235.0,
      "step": 38045
    },
    {
      "entropy": 5.716439199447632,
      "epoch": 3.776300119094879,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003634371481272101,
      "loss": 5.0442,
      "mean_token_accuracy": 0.18777037113904954,
      "num_tokens": 83375135.0,
      "step": 38050
    },
    {
      "entropy": 5.705961990356445,
      "epoch": 3.7767963477570463,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003634045672247327,
      "loss": 5.0217,
      "mean_token_accuracy": 0.19169188737869264,
      "num_tokens": 83385340.0,
      "step": 38055
    },
    {
      "entropy": 5.74348669052124,
      "epoch": 3.777292576419214,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003633719841300126,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18843438476324081,
      "num_tokens": 83396215.0,
      "step": 38060
    },
    {
      "entropy": 5.620566034317017,
      "epoch": 3.7777888050813813,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003633393988438576,
      "loss": 4.9467,
      "mean_token_accuracy": 0.19385550022125245,
      "num_tokens": 83407586.0,
      "step": 38065
    },
    {
      "entropy": 5.637872648239136,
      "epoch": 3.778285033743549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003633068113670759,
      "loss": 4.9272,
      "mean_token_accuracy": 0.19406328946352006,
      "num_tokens": 83417881.0,
      "step": 38070
    },
    {
      "entropy": 5.631580495834351,
      "epoch": 3.7787812624057167,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036327422170047565,
      "loss": 4.958,
      "mean_token_accuracy": 0.19255954325199126,
      "num_tokens": 83428794.0,
      "step": 38075
    },
    {
      "entropy": 5.618811178207397,
      "epoch": 3.779277491067884,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003632416298448648,
      "loss": 4.9307,
      "mean_token_accuracy": 0.19637101739645005,
      "num_tokens": 83439289.0,
      "step": 38080
    },
    {
      "entropy": 5.645553970336914,
      "epoch": 3.7797737197300516,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036320903580105176,
      "loss": 4.9728,
      "mean_token_accuracy": 0.18693810552358628,
      "num_tokens": 83449183.0,
      "step": 38085
    },
    {
      "entropy": 5.703620767593383,
      "epoch": 3.780269948392219,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036317643956984466,
      "loss": 5.0825,
      "mean_token_accuracy": 0.18490539640188217,
      "num_tokens": 83460225.0,
      "step": 38090
    },
    {
      "entropy": 5.697056007385254,
      "epoch": 3.7807661770543866,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036314384115205184,
      "loss": 5.0571,
      "mean_token_accuracy": 0.18669558614492415,
      "num_tokens": 83470831.0,
      "step": 38095
    },
    {
      "entropy": 5.597533893585205,
      "epoch": 3.781262405716554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003631112405484817,
      "loss": 4.9537,
      "mean_token_accuracy": 0.19380559176206588,
      "num_tokens": 83482472.0,
      "step": 38100
    },
    {
      "entropy": 5.671335124969483,
      "epoch": 3.781758634378722,
      "grad_norm": 1.125,
      "learning_rate": 0.00036307863775994275,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1918862745165825,
      "num_tokens": 83492838.0,
      "step": 38105
    },
    {
      "entropy": 5.6050199508667,
      "epoch": 3.7822548630408894,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036304603278724325,
      "loss": 4.9401,
      "mean_token_accuracy": 0.1914932906627655,
      "num_tokens": 83503055.0,
      "step": 38110
    },
    {
      "entropy": 5.66247787475586,
      "epoch": 3.782751091703057,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036301342563119194,
      "loss": 4.9601,
      "mean_token_accuracy": 0.19215754717588424,
      "num_tokens": 83512710.0,
      "step": 38115
    },
    {
      "entropy": 5.629550409317017,
      "epoch": 3.7832473203652244,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003629808162925973,
      "loss": 4.8918,
      "mean_token_accuracy": 0.19548716843128205,
      "num_tokens": 83522790.0,
      "step": 38120
    },
    {
      "entropy": 5.625388145446777,
      "epoch": 3.783743549027392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036294820477226797,
      "loss": 4.9737,
      "mean_token_accuracy": 0.1871588334441185,
      "num_tokens": 83534800.0,
      "step": 38125
    },
    {
      "entropy": 5.63240008354187,
      "epoch": 3.7842397776895593,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003629155910710127,
      "loss": 4.9564,
      "mean_token_accuracy": 0.1879636377096176,
      "num_tokens": 83544507.0,
      "step": 38130
    },
    {
      "entropy": 5.632778024673462,
      "epoch": 3.784736006351727,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003628829751896403,
      "loss": 4.9312,
      "mean_token_accuracy": 0.19351528882980346,
      "num_tokens": 83555053.0,
      "step": 38135
    },
    {
      "entropy": 5.6586620807647705,
      "epoch": 3.7852322350138943,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036285035712895934,
      "loss": 4.9449,
      "mean_token_accuracy": 0.1980431005358696,
      "num_tokens": 83564585.0,
      "step": 38140
    },
    {
      "entropy": 5.633138275146484,
      "epoch": 3.7857284636760618,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036281773688977894,
      "loss": 4.9508,
      "mean_token_accuracy": 0.18850650489330292,
      "num_tokens": 83574632.0,
      "step": 38145
    },
    {
      "entropy": 5.630981826782227,
      "epoch": 3.7862246923382292,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003627851144729078,
      "loss": 5.0543,
      "mean_token_accuracy": 0.18570600301027299,
      "num_tokens": 83586240.0,
      "step": 38150
    },
    {
      "entropy": 5.69580979347229,
      "epoch": 3.786720921000397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036275248987915497,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18802760243415834,
      "num_tokens": 83597669.0,
      "step": 38155
    },
    {
      "entropy": 5.742656183242798,
      "epoch": 3.7872171496625646,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036271986310932957,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1778021991252899,
      "num_tokens": 83608356.0,
      "step": 38160
    },
    {
      "entropy": 5.738620567321777,
      "epoch": 3.787713378324732,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003626872341642406,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18618321567773818,
      "num_tokens": 83620216.0,
      "step": 38165
    },
    {
      "entropy": 5.643303489685058,
      "epoch": 3.7882096069868996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003626546030446971,
      "loss": 4.9949,
      "mean_token_accuracy": 0.19121353328227997,
      "num_tokens": 83630807.0,
      "step": 38170
    },
    {
      "entropy": 5.644549798965454,
      "epoch": 3.788705835649067,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036262196975150835,
      "loss": 4.9652,
      "mean_token_accuracy": 0.1960174784064293,
      "num_tokens": 83643371.0,
      "step": 38175
    },
    {
      "entropy": 5.668812847137451,
      "epoch": 3.7892020643112345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003625893342854836,
      "loss": 4.9769,
      "mean_token_accuracy": 0.1950622022151947,
      "num_tokens": 83653489.0,
      "step": 38180
    },
    {
      "entropy": 5.618146657943726,
      "epoch": 3.789698292973402,
      "grad_norm": 1.125,
      "learning_rate": 0.00036255669664743195,
      "loss": 4.9262,
      "mean_token_accuracy": 0.19478649497032166,
      "num_tokens": 83664179.0,
      "step": 38185
    },
    {
      "entropy": 5.6648736000061035,
      "epoch": 3.79019452163557,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036252405683816306,
      "loss": 5.035,
      "mean_token_accuracy": 0.18486194908618928,
      "num_tokens": 83674912.0,
      "step": 38190
    },
    {
      "entropy": 5.681851005554199,
      "epoch": 3.7906907502977374,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003624914148584861,
      "loss": 4.9784,
      "mean_token_accuracy": 0.18907091170549392,
      "num_tokens": 83686187.0,
      "step": 38195
    },
    {
      "entropy": 5.663764715194702,
      "epoch": 3.791186978959905,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003624587707092105,
      "loss": 4.9368,
      "mean_token_accuracy": 0.1971246436238289,
      "num_tokens": 83696169.0,
      "step": 38200
    },
    {
      "entropy": 5.6323521614074705,
      "epoch": 3.7916832076220723,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036242612439114593,
      "loss": 5.0375,
      "mean_token_accuracy": 0.1884814366698265,
      "num_tokens": 83707127.0,
      "step": 38205
    },
    {
      "entropy": 5.619477367401123,
      "epoch": 3.79217943628424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036239347590510186,
      "loss": 4.8786,
      "mean_token_accuracy": 0.19050928801298142,
      "num_tokens": 83718081.0,
      "step": 38210
    },
    {
      "entropy": 5.712203884124756,
      "epoch": 3.7926756649464073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003623608252518879,
      "loss": 4.9973,
      "mean_token_accuracy": 0.19033613204956054,
      "num_tokens": 83729272.0,
      "step": 38215
    },
    {
      "entropy": 5.654443597793579,
      "epoch": 3.7931718936085748,
      "grad_norm": 1.25,
      "learning_rate": 0.0003623281724323137,
      "loss": 4.9541,
      "mean_token_accuracy": 0.19209375679492952,
      "num_tokens": 83740419.0,
      "step": 38220
    },
    {
      "entropy": 5.69622573852539,
      "epoch": 3.7936681222707422,
      "grad_norm": 1.21875,
      "learning_rate": 0.000362295517447189,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1882263347506523,
      "num_tokens": 83750348.0,
      "step": 38225
    },
    {
      "entropy": 5.6803192615509035,
      "epoch": 3.7941643509329097,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003622628602973236,
      "loss": 4.9846,
      "mean_token_accuracy": 0.19236861914396286,
      "num_tokens": 83760319.0,
      "step": 38230
    },
    {
      "entropy": 5.6196544647216795,
      "epoch": 3.794660579595077,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036223020098352724,
      "loss": 4.9446,
      "mean_token_accuracy": 0.19626190215349198,
      "num_tokens": 83772120.0,
      "step": 38235
    },
    {
      "entropy": 5.667178440093994,
      "epoch": 3.7951568082572447,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003621975395066099,
      "loss": 5.0265,
      "mean_token_accuracy": 0.187529456615448,
      "num_tokens": 83784306.0,
      "step": 38240
    },
    {
      "entropy": 5.627833986282349,
      "epoch": 3.7956530369194126,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036216487586738144,
      "loss": 4.8553,
      "mean_token_accuracy": 0.20174746811389924,
      "num_tokens": 83794397.0,
      "step": 38245
    },
    {
      "entropy": 5.704026508331299,
      "epoch": 3.79614926558158,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000362132210066652,
      "loss": 4.9406,
      "mean_token_accuracy": 0.19590207636356355,
      "num_tokens": 83805633.0,
      "step": 38250
    },
    {
      "entropy": 5.694241857528686,
      "epoch": 3.7966454942437475,
      "grad_norm": 1.125,
      "learning_rate": 0.00036209954210523137,
      "loss": 4.9979,
      "mean_token_accuracy": 0.1882918357849121,
      "num_tokens": 83816633.0,
      "step": 38255
    },
    {
      "entropy": 5.696560573577881,
      "epoch": 3.797141722905915,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003620668719839298,
      "loss": 5.014,
      "mean_token_accuracy": 0.18685114830732347,
      "num_tokens": 83826889.0,
      "step": 38260
    },
    {
      "entropy": 5.650158786773682,
      "epoch": 3.7976379515680825,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036203419970355747,
      "loss": 4.9075,
      "mean_token_accuracy": 0.19131799042224884,
      "num_tokens": 83836629.0,
      "step": 38265
    },
    {
      "entropy": 5.674249362945557,
      "epoch": 3.79813418023025,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003620015252649245,
      "loss": 5.0761,
      "mean_token_accuracy": 0.18491154462099074,
      "num_tokens": 83848790.0,
      "step": 38270
    },
    {
      "entropy": 5.600415372848511,
      "epoch": 3.798630408892418,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036196884866884115,
      "loss": 4.8943,
      "mean_token_accuracy": 0.1935485988855362,
      "num_tokens": 83859427.0,
      "step": 38275
    },
    {
      "entropy": 5.675079584121704,
      "epoch": 3.7991266375545854,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036193616991611784,
      "loss": 4.9678,
      "mean_token_accuracy": 0.19443043172359467,
      "num_tokens": 83869987.0,
      "step": 38280
    },
    {
      "entropy": 5.6666018009185795,
      "epoch": 3.799622866216753,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036190348900756475,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18442807644605635,
      "num_tokens": 83880512.0,
      "step": 38285
    },
    {
      "entropy": 5.695817565917968,
      "epoch": 3.8001190948789203,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003618708059439924,
      "loss": 4.9724,
      "mean_token_accuracy": 0.18949009031057357,
      "num_tokens": 83891884.0,
      "step": 38290
    },
    {
      "entropy": 5.719083404541015,
      "epoch": 3.8006153235410878,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003618381207262113,
      "loss": 5.0292,
      "mean_token_accuracy": 0.1838395655155182,
      "num_tokens": 83905079.0,
      "step": 38295
    },
    {
      "entropy": 5.62665548324585,
      "epoch": 3.8011115522032553,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003618054333550319,
      "loss": 4.9794,
      "mean_token_accuracy": 0.18727389574050904,
      "num_tokens": 83915539.0,
      "step": 38300
    },
    {
      "entropy": 5.687454652786255,
      "epoch": 3.8016077808654227,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036177274383126487,
      "loss": 5.0685,
      "mean_token_accuracy": 0.19022947996854783,
      "num_tokens": 83925779.0,
      "step": 38305
    },
    {
      "entropy": 5.617529344558716,
      "epoch": 3.80210400952759,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003617400521557207,
      "loss": 4.9115,
      "mean_token_accuracy": 0.1937125876545906,
      "num_tokens": 83935125.0,
      "step": 38310
    },
    {
      "entropy": 5.682462549209594,
      "epoch": 3.8026002381897577,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003617073583292101,
      "loss": 4.9355,
      "mean_token_accuracy": 0.19954591989517212,
      "num_tokens": 83945685.0,
      "step": 38315
    },
    {
      "entropy": 5.712564325332641,
      "epoch": 3.803096466851925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036167466235254394,
      "loss": 5.0795,
      "mean_token_accuracy": 0.1853422775864601,
      "num_tokens": 83956706.0,
      "step": 38320
    },
    {
      "entropy": 5.740531969070434,
      "epoch": 3.8035926955140926,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036164196422653293,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1880684643983841,
      "num_tokens": 83967081.0,
      "step": 38325
    },
    {
      "entropy": 5.630482482910156,
      "epoch": 3.8040889241762605,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036160926395198786,
      "loss": 4.9257,
      "mean_token_accuracy": 0.1942004531621933,
      "num_tokens": 83978788.0,
      "step": 38330
    },
    {
      "entropy": 5.673584461212158,
      "epoch": 3.804585152838428,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003615765615297197,
      "loss": 4.9573,
      "mean_token_accuracy": 0.191453018784523,
      "num_tokens": 83988696.0,
      "step": 38335
    },
    {
      "entropy": 5.6069543838500975,
      "epoch": 3.8050813815005955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003615438569605394,
      "loss": 4.9208,
      "mean_token_accuracy": 0.18867304772138596,
      "num_tokens": 84000616.0,
      "step": 38340
    },
    {
      "entropy": 5.673246669769287,
      "epoch": 3.805577610162763,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036151115024525777,
      "loss": 4.9823,
      "mean_token_accuracy": 0.1894165426492691,
      "num_tokens": 84012358.0,
      "step": 38345
    },
    {
      "entropy": 5.617692565917968,
      "epoch": 3.8060738388249304,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036147844138468616,
      "loss": 4.9526,
      "mean_token_accuracy": 0.1921518325805664,
      "num_tokens": 84023520.0,
      "step": 38350
    },
    {
      "entropy": 5.722918939590454,
      "epoch": 3.806570067487098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003614457303796356,
      "loss": 4.9899,
      "mean_token_accuracy": 0.1883169084787369,
      "num_tokens": 84034413.0,
      "step": 38355
    },
    {
      "entropy": 5.6170323371887205,
      "epoch": 3.807066296149266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003614130172309171,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18958460986614228,
      "num_tokens": 84046157.0,
      "step": 38360
    },
    {
      "entropy": 5.679438495635987,
      "epoch": 3.8075625248114333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003613803019393419,
      "loss": 5.0719,
      "mean_token_accuracy": 0.1852090448141098,
      "num_tokens": 84059506.0,
      "step": 38365
    },
    {
      "entropy": 5.718417596817017,
      "epoch": 3.808058753473601,
      "grad_norm": 1.25,
      "learning_rate": 0.00036134758450572153,
      "loss": 4.9911,
      "mean_token_accuracy": 0.1887483224272728,
      "num_tokens": 84070317.0,
      "step": 38370
    },
    {
      "entropy": 5.6822114944458,
      "epoch": 3.8085549821357683,
      "grad_norm": 1.171875,
      "learning_rate": 0.000361314864930867,
      "loss": 4.9729,
      "mean_token_accuracy": 0.1918485179543495,
      "num_tokens": 84081997.0,
      "step": 38375
    },
    {
      "entropy": 5.632814073562622,
      "epoch": 3.8090512107979357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003612821432155898,
      "loss": 4.939,
      "mean_token_accuracy": 0.187521855533123,
      "num_tokens": 84092468.0,
      "step": 38380
    },
    {
      "entropy": 5.679340171813965,
      "epoch": 3.809547439460103,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003612494193607015,
      "loss": 4.9492,
      "mean_token_accuracy": 0.1919543832540512,
      "num_tokens": 84102727.0,
      "step": 38385
    },
    {
      "entropy": 5.685619115829468,
      "epoch": 3.8100436681222707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003612166933670134,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18177189379930497,
      "num_tokens": 84114316.0,
      "step": 38390
    },
    {
      "entropy": 5.67602858543396,
      "epoch": 3.810539896784438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000361183965235337,
      "loss": 5.049,
      "mean_token_accuracy": 0.18851385414600372,
      "num_tokens": 84126375.0,
      "step": 38395
    },
    {
      "entropy": 5.681365299224853,
      "epoch": 3.8110361254466056,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003611512349664841,
      "loss": 4.9966,
      "mean_token_accuracy": 0.19064349085092544,
      "num_tokens": 84136898.0,
      "step": 38400
    },
    {
      "entropy": 5.579112195968628,
      "epoch": 3.811532354108773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003611185025612661,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2047694891691208,
      "num_tokens": 84148016.0,
      "step": 38405
    },
    {
      "entropy": 5.648340511322021,
      "epoch": 3.8120285827709406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003610857680204949,
      "loss": 5.013,
      "mean_token_accuracy": 0.1848191574215889,
      "num_tokens": 84159476.0,
      "step": 38410
    },
    {
      "entropy": 5.666786432266235,
      "epoch": 3.8125248114331085,
      "grad_norm": 1.125,
      "learning_rate": 0.00036105303134498207,
      "loss": 4.9738,
      "mean_token_accuracy": 0.1920041024684906,
      "num_tokens": 84170390.0,
      "step": 38415
    },
    {
      "entropy": 5.673123216629028,
      "epoch": 3.813021040095276,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036102029253553945,
      "loss": 5.0051,
      "mean_token_accuracy": 0.1894015073776245,
      "num_tokens": 84183265.0,
      "step": 38420
    },
    {
      "entropy": 5.645013856887817,
      "epoch": 3.8135172687574435,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003609875515929791,
      "loss": 4.9838,
      "mean_token_accuracy": 0.19104186296463013,
      "num_tokens": 84193251.0,
      "step": 38425
    },
    {
      "entropy": 5.681488180160523,
      "epoch": 3.814013497419611,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003609548085181126,
      "loss": 4.974,
      "mean_token_accuracy": 0.18753852993249892,
      "num_tokens": 84203131.0,
      "step": 38430
    },
    {
      "entropy": 5.7142048358917235,
      "epoch": 3.8145097260817784,
      "grad_norm": 1.25,
      "learning_rate": 0.0003609220633117521,
      "loss": 4.9768,
      "mean_token_accuracy": 0.19620784670114516,
      "num_tokens": 84213356.0,
      "step": 38435
    },
    {
      "entropy": 5.742343854904175,
      "epoch": 3.815005954743946,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003608893159747097,
      "loss": 5.1235,
      "mean_token_accuracy": 0.1807406648993492,
      "num_tokens": 84224304.0,
      "step": 38440
    },
    {
      "entropy": 5.683638381958008,
      "epoch": 3.815502183406114,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036085656650779716,
      "loss": 4.9736,
      "mean_token_accuracy": 0.19268364310264588,
      "num_tokens": 84235010.0,
      "step": 38445
    },
    {
      "entropy": 5.687301969528198,
      "epoch": 3.8159984120682813,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036082381491182685,
      "loss": 5.0105,
      "mean_token_accuracy": 0.1930270954966545,
      "num_tokens": 84246486.0,
      "step": 38450
    },
    {
      "entropy": 5.666319274902344,
      "epoch": 3.8164946407304488,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003607910611876109,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18558910489082336,
      "num_tokens": 84258081.0,
      "step": 38455
    },
    {
      "entropy": 5.652064085006714,
      "epoch": 3.8169908693926162,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003607583053359615,
      "loss": 4.9446,
      "mean_token_accuracy": 0.19800257980823516,
      "num_tokens": 84268609.0,
      "step": 38460
    },
    {
      "entropy": 5.76006293296814,
      "epoch": 3.8174870980547837,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003607255473576908,
      "loss": 5.0393,
      "mean_token_accuracy": 0.18762154281139373,
      "num_tokens": 84278655.0,
      "step": 38465
    },
    {
      "entropy": 5.687230491638184,
      "epoch": 3.817983326716951,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036069278725361144,
      "loss": 5.0368,
      "mean_token_accuracy": 0.18293385058641434,
      "num_tokens": 84289604.0,
      "step": 38470
    },
    {
      "entropy": 5.65771369934082,
      "epoch": 3.8184795553791187,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036066002502453556,
      "loss": 5.0013,
      "mean_token_accuracy": 0.19188869148492813,
      "num_tokens": 84300075.0,
      "step": 38475
    },
    {
      "entropy": 5.705559825897216,
      "epoch": 3.818975784041286,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003606272606712755,
      "loss": 5.0127,
      "mean_token_accuracy": 0.1872366800904274,
      "num_tokens": 84311317.0,
      "step": 38480
    },
    {
      "entropy": 5.661054277420044,
      "epoch": 3.8194720127034536,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036059449419464397,
      "loss": 4.9819,
      "mean_token_accuracy": 0.1850247636437416,
      "num_tokens": 84321354.0,
      "step": 38485
    },
    {
      "entropy": 5.66980938911438,
      "epoch": 3.819968241365621,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003605617255954535,
      "loss": 4.9842,
      "mean_token_accuracy": 0.19451007544994353,
      "num_tokens": 84332423.0,
      "step": 38490
    },
    {
      "entropy": 5.584043169021607,
      "epoch": 3.8204644700277886,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003605289548745164,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2020471513271332,
      "num_tokens": 84342697.0,
      "step": 38495
    },
    {
      "entropy": 5.644062089920044,
      "epoch": 3.8209606986899565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003604961820326457,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18709037452936172,
      "num_tokens": 84353509.0,
      "step": 38500
    },
    {
      "entropy": 5.684465312957764,
      "epoch": 3.821456927352124,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036046340707065387,
      "loss": 5.1078,
      "mean_token_accuracy": 0.18084113895893097,
      "num_tokens": 84365234.0,
      "step": 38505
    },
    {
      "entropy": 5.672574043273926,
      "epoch": 3.8219531560142914,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003604306299893537,
      "loss": 4.9175,
      "mean_token_accuracy": 0.19628805816173553,
      "num_tokens": 84375719.0,
      "step": 38510
    },
    {
      "entropy": 5.689174842834473,
      "epoch": 3.822449384676459,
      "grad_norm": 1.125,
      "learning_rate": 0.00036039785078955793,
      "loss": 5.0631,
      "mean_token_accuracy": 0.18198988288640977,
      "num_tokens": 84386326.0,
      "step": 38515
    },
    {
      "entropy": 5.619151926040649,
      "epoch": 3.8229456133386264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003603650694720796,
      "loss": 4.9331,
      "mean_token_accuracy": 0.1880045309662819,
      "num_tokens": 84398081.0,
      "step": 38520
    },
    {
      "entropy": 5.676185750961304,
      "epoch": 3.823441842000794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036033228603773136,
      "loss": 4.9876,
      "mean_token_accuracy": 0.18177227824926376,
      "num_tokens": 84408984.0,
      "step": 38525
    },
    {
      "entropy": 5.717316150665283,
      "epoch": 3.8239380706629618,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036029950048732635,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18533605188131333,
      "num_tokens": 84420848.0,
      "step": 38530
    },
    {
      "entropy": 5.666243362426758,
      "epoch": 3.8244342993251292,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036026671282167753,
      "loss": 4.9328,
      "mean_token_accuracy": 0.19045475274324417,
      "num_tokens": 84430221.0,
      "step": 38535
    },
    {
      "entropy": 5.644154834747314,
      "epoch": 3.8249305279872967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000360233923041598,
      "loss": 5.0041,
      "mean_token_accuracy": 0.19126246571540834,
      "num_tokens": 84440484.0,
      "step": 38540
    },
    {
      "entropy": 5.656496667861939,
      "epoch": 3.825426756649464,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036020113114790085,
      "loss": 4.9756,
      "mean_token_accuracy": 0.19144887030124663,
      "num_tokens": 84451264.0,
      "step": 38545
    },
    {
      "entropy": 5.678102779388428,
      "epoch": 3.8259229853116317,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003601683371413992,
      "loss": 5.0158,
      "mean_token_accuracy": 0.20075618326663972,
      "num_tokens": 84462957.0,
      "step": 38550
    },
    {
      "entropy": 5.729752492904663,
      "epoch": 3.826419213973799,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036013554102290636,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1849573016166687,
      "num_tokens": 84473662.0,
      "step": 38555
    },
    {
      "entropy": 5.662830257415772,
      "epoch": 3.8269154426359666,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003601027427932355,
      "loss": 5.0059,
      "mean_token_accuracy": 0.19318082481622695,
      "num_tokens": 84484904.0,
      "step": 38560
    },
    {
      "entropy": 5.597792768478394,
      "epoch": 3.827411671298134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003600699424532,
      "loss": 4.896,
      "mean_token_accuracy": 0.19902991503477097,
      "num_tokens": 84495189.0,
      "step": 38565
    },
    {
      "entropy": 5.6677999019622805,
      "epoch": 3.8279078999603016,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036003714000361326,
      "loss": 4.9549,
      "mean_token_accuracy": 0.19273137003183366,
      "num_tokens": 84506103.0,
      "step": 38570
    },
    {
      "entropy": 5.651500654220581,
      "epoch": 3.828404128622469,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036000433544528866,
      "loss": 4.9792,
      "mean_token_accuracy": 0.1879129260778427,
      "num_tokens": 84517614.0,
      "step": 38575
    },
    {
      "entropy": 5.618756484985352,
      "epoch": 3.8289003572846365,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003599715287790397,
      "loss": 4.8957,
      "mean_token_accuracy": 0.19133664965629577,
      "num_tokens": 84527737.0,
      "step": 38580
    },
    {
      "entropy": 5.675601291656494,
      "epoch": 3.8293965859468044,
      "grad_norm": 1.125,
      "learning_rate": 0.00035993872000567994,
      "loss": 4.9878,
      "mean_token_accuracy": 0.18672073930501937,
      "num_tokens": 84537652.0,
      "step": 38585
    },
    {
      "entropy": 5.655759477615357,
      "epoch": 3.829892814608972,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035990590912602304,
      "loss": 4.968,
      "mean_token_accuracy": 0.1882711634039879,
      "num_tokens": 84547500.0,
      "step": 38590
    },
    {
      "entropy": 5.668849515914917,
      "epoch": 3.8303890432711394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035987309614088245,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1850472405552864,
      "num_tokens": 84558415.0,
      "step": 38595
    },
    {
      "entropy": 5.744046401977539,
      "epoch": 3.830885271933307,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035984028105107195,
      "loss": 5.0632,
      "mean_token_accuracy": 0.1829160436987877,
      "num_tokens": 84569241.0,
      "step": 38600
    },
    {
      "entropy": 5.677428340911865,
      "epoch": 3.8313815005954743,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035980746385740536,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19516607969999314,
      "num_tokens": 84580821.0,
      "step": 38605
    },
    {
      "entropy": 5.7395445823669435,
      "epoch": 3.831877729257642,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035977464456069636,
      "loss": 5.0544,
      "mean_token_accuracy": 0.1838679239153862,
      "num_tokens": 84592428.0,
      "step": 38610
    },
    {
      "entropy": 5.698325681686401,
      "epoch": 3.8323739579198093,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035974182316175887,
      "loss": 4.986,
      "mean_token_accuracy": 0.19545787572860718,
      "num_tokens": 84602931.0,
      "step": 38615
    },
    {
      "entropy": 5.620042896270752,
      "epoch": 3.832870186581977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003597089996614068,
      "loss": 4.9668,
      "mean_token_accuracy": 0.1831803575158119,
      "num_tokens": 84614193.0,
      "step": 38620
    },
    {
      "entropy": 5.642746019363403,
      "epoch": 3.8333664152441447,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035967617406045406,
      "loss": 4.944,
      "mean_token_accuracy": 0.1897557035088539,
      "num_tokens": 84624440.0,
      "step": 38625
    },
    {
      "entropy": 5.657651615142822,
      "epoch": 3.833862643906312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035964334635971455,
      "loss": 4.9393,
      "mean_token_accuracy": 0.19578789919614792,
      "num_tokens": 84635661.0,
      "step": 38630
    },
    {
      "entropy": 5.657924699783325,
      "epoch": 3.8343588725684796,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003596105165600025,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18709564357995986,
      "num_tokens": 84647416.0,
      "step": 38635
    },
    {
      "entropy": 5.664909744262696,
      "epoch": 3.834855101230647,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035957768466213203,
      "loss": 4.9661,
      "mean_token_accuracy": 0.19199933856725693,
      "num_tokens": 84657507.0,
      "step": 38640
    },
    {
      "entropy": 5.661807823181152,
      "epoch": 3.8353513298928146,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035954485066691717,
      "loss": 4.9207,
      "mean_token_accuracy": 0.18804228603839873,
      "num_tokens": 84668471.0,
      "step": 38645
    },
    {
      "entropy": 5.606040906906128,
      "epoch": 3.835847558554982,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003595120145751723,
      "loss": 4.8465,
      "mean_token_accuracy": 0.1969015657901764,
      "num_tokens": 84679755.0,
      "step": 38650
    },
    {
      "entropy": 5.704650735855102,
      "epoch": 3.8363437872171495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035947917638771146,
      "loss": 5.0723,
      "mean_token_accuracy": 0.185339492559433,
      "num_tokens": 84691096.0,
      "step": 38655
    },
    {
      "entropy": 5.64655876159668,
      "epoch": 3.836840015879317,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035944633610534904,
      "loss": 4.9825,
      "mean_token_accuracy": 0.19045981168746948,
      "num_tokens": 84703977.0,
      "step": 38660
    },
    {
      "entropy": 5.694063758850097,
      "epoch": 3.8373362445414845,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035941349372889956,
      "loss": 5.0033,
      "mean_token_accuracy": 0.1885179504752159,
      "num_tokens": 84716132.0,
      "step": 38665
    },
    {
      "entropy": 5.661597681045532,
      "epoch": 3.8378324732036524,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035938064925917726,
      "loss": 4.9716,
      "mean_token_accuracy": 0.19040943533182145,
      "num_tokens": 84728658.0,
      "step": 38670
    },
    {
      "entropy": 5.703979969024658,
      "epoch": 3.83832870186582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003593478026969967,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18766611218452453,
      "num_tokens": 84739266.0,
      "step": 38675
    },
    {
      "entropy": 5.6432843685150145,
      "epoch": 3.8388249305279873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035931495404317243,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2035440430045128,
      "num_tokens": 84750923.0,
      "step": 38680
    },
    {
      "entropy": 5.568177080154419,
      "epoch": 3.839321159190155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000359282103298519,
      "loss": 4.9724,
      "mean_token_accuracy": 0.1869322657585144,
      "num_tokens": 84762755.0,
      "step": 38685
    },
    {
      "entropy": 5.678036689758301,
      "epoch": 3.8398173878523223,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035924925046385085,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1929445043206215,
      "num_tokens": 84773252.0,
      "step": 38690
    },
    {
      "entropy": 5.710522985458374,
      "epoch": 3.8403136165144898,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035921639553998297,
      "loss": 4.988,
      "mean_token_accuracy": 0.19185485541820527,
      "num_tokens": 84784526.0,
      "step": 38695
    },
    {
      "entropy": 5.679338788986206,
      "epoch": 3.8408098451766572,
      "grad_norm": 1.125,
      "learning_rate": 0.00035918353852772996,
      "loss": 5.022,
      "mean_token_accuracy": 0.19281063973903656,
      "num_tokens": 84796251.0,
      "step": 38700
    },
    {
      "entropy": 5.652444362640381,
      "epoch": 3.841306073838825,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003591506794279066,
      "loss": 4.9567,
      "mean_token_accuracy": 0.19319202303886412,
      "num_tokens": 84806947.0,
      "step": 38705
    },
    {
      "entropy": 5.690500402450562,
      "epoch": 3.8418023025009926,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003591178182413276,
      "loss": 5.029,
      "mean_token_accuracy": 0.18860018104314805,
      "num_tokens": 84819584.0,
      "step": 38710
    },
    {
      "entropy": 5.694978857040406,
      "epoch": 3.84229853116316,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00035908495496880804,
      "loss": 5.0401,
      "mean_token_accuracy": 0.18927372694015504,
      "num_tokens": 84829546.0,
      "step": 38715
    },
    {
      "entropy": 5.724910688400269,
      "epoch": 3.8427947598253276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003590520896111628,
      "loss": 5.0415,
      "mean_token_accuracy": 0.18929256051778792,
      "num_tokens": 84840843.0,
      "step": 38720
    },
    {
      "entropy": 5.647416734695435,
      "epoch": 3.843290988487495,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035901922216920683,
      "loss": 5.0089,
      "mean_token_accuracy": 0.19241406321525573,
      "num_tokens": 84852015.0,
      "step": 38725
    },
    {
      "entropy": 5.710695600509643,
      "epoch": 3.8437872171496625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003589863526437552,
      "loss": 4.9476,
      "mean_token_accuracy": 0.19587500095367433,
      "num_tokens": 84863720.0,
      "step": 38730
    },
    {
      "entropy": 5.6846668243408205,
      "epoch": 3.84428344581183,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003589534810356231,
      "loss": 4.9741,
      "mean_token_accuracy": 0.19038322418928147,
      "num_tokens": 84873181.0,
      "step": 38735
    },
    {
      "entropy": 5.64087233543396,
      "epoch": 3.8447796744739975,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035892060734562544,
      "loss": 4.9964,
      "mean_token_accuracy": 0.187871353328228,
      "num_tokens": 84884478.0,
      "step": 38740
    },
    {
      "entropy": 5.625946092605591,
      "epoch": 3.845275903136165,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00035888773157457756,
      "loss": 4.9595,
      "mean_token_accuracy": 0.1850600779056549,
      "num_tokens": 84894846.0,
      "step": 38745
    },
    {
      "entropy": 5.657539939880371,
      "epoch": 3.8457721317983324,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003588548537232947,
      "loss": 4.8774,
      "mean_token_accuracy": 0.20014103949069978,
      "num_tokens": 84905013.0,
      "step": 38750
    },
    {
      "entropy": 5.608162927627563,
      "epoch": 3.8462683604605004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035882197379259224,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2020984932780266,
      "num_tokens": 84916347.0,
      "step": 38755
    },
    {
      "entropy": 5.690939092636109,
      "epoch": 3.846764589122668,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003587890917832854,
      "loss": 5.0565,
      "mean_token_accuracy": 0.19075614213943481,
      "num_tokens": 84927770.0,
      "step": 38760
    },
    {
      "entropy": 5.679639482498169,
      "epoch": 3.8472608177848353,
      "grad_norm": 1.125,
      "learning_rate": 0.0003587562076961897,
      "loss": 4.9684,
      "mean_token_accuracy": 0.18834556341171266,
      "num_tokens": 84938323.0,
      "step": 38765
    },
    {
      "entropy": 5.695719814300537,
      "epoch": 3.847757046447003,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035872332153212046,
      "loss": 4.9713,
      "mean_token_accuracy": 0.19248559176921845,
      "num_tokens": 84948857.0,
      "step": 38770
    },
    {
      "entropy": 5.638695478439331,
      "epoch": 3.8482532751091703,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035869043329189314,
      "loss": 4.9562,
      "mean_token_accuracy": 0.18636986911296843,
      "num_tokens": 84959841.0,
      "step": 38775
    },
    {
      "entropy": 5.674744749069214,
      "epoch": 3.8487495037713377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003586575429763236,
      "loss": 5.0,
      "mean_token_accuracy": 0.191436830163002,
      "num_tokens": 84971013.0,
      "step": 38780
    },
    {
      "entropy": 5.712418270111084,
      "epoch": 3.849245732433505,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003586246505862272,
      "loss": 5.013,
      "mean_token_accuracy": 0.19008344709873198,
      "num_tokens": 84981277.0,
      "step": 38785
    },
    {
      "entropy": 5.655593347549439,
      "epoch": 3.849741961095673,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035859175612241957,
      "loss": 4.9689,
      "mean_token_accuracy": 0.1916593074798584,
      "num_tokens": 84993129.0,
      "step": 38790
    },
    {
      "entropy": 5.6384031772613525,
      "epoch": 3.8502381897578406,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035855885958571666,
      "loss": 4.9661,
      "mean_token_accuracy": 0.19520987719297409,
      "num_tokens": 85004152.0,
      "step": 38795
    },
    {
      "entropy": 5.670439624786377,
      "epoch": 3.850734418420008,
      "grad_norm": 1.15625,
      "learning_rate": 0.000358525960976934,
      "loss": 4.9767,
      "mean_token_accuracy": 0.1957280933856964,
      "num_tokens": 85016435.0,
      "step": 38800
    },
    {
      "entropy": 5.694642066955566,
      "epoch": 3.8512306470821756,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003584930602968875,
      "loss": 4.971,
      "mean_token_accuracy": 0.1841138243675232,
      "num_tokens": 85027107.0,
      "step": 38805
    },
    {
      "entropy": 5.702080345153808,
      "epoch": 3.851726875744343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035846015754639304,
      "loss": 5.1,
      "mean_token_accuracy": 0.18168012499809266,
      "num_tokens": 85039735.0,
      "step": 38810
    },
    {
      "entropy": 5.751820802688599,
      "epoch": 3.8522231044065105,
      "grad_norm": 1.25,
      "learning_rate": 0.0003584272527262665,
      "loss": 5.0387,
      "mean_token_accuracy": 0.19092777371406555,
      "num_tokens": 85049050.0,
      "step": 38815
    },
    {
      "entropy": 5.653997373580933,
      "epoch": 3.852719333068678,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035839434583732384,
      "loss": 4.8858,
      "mean_token_accuracy": 0.1970951110124588,
      "num_tokens": 85059229.0,
      "step": 38820
    },
    {
      "entropy": 5.660167646408081,
      "epoch": 3.8532155617308455,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00035836143688038107,
      "loss": 5.007,
      "mean_token_accuracy": 0.18575756549835204,
      "num_tokens": 85070022.0,
      "step": 38825
    },
    {
      "entropy": 5.728304862976074,
      "epoch": 3.853711790393013,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035832852585625436,
      "loss": 5.0749,
      "mean_token_accuracy": 0.1897353485226631,
      "num_tokens": 85081944.0,
      "step": 38830
    },
    {
      "entropy": 5.714842128753662,
      "epoch": 3.8542080190551804,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003582956127657598,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18985197991132735,
      "num_tokens": 85093182.0,
      "step": 38835
    },
    {
      "entropy": 5.659442996978759,
      "epoch": 3.8547042477173483,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003582626976097134,
      "loss": 4.97,
      "mean_token_accuracy": 0.19412271827459335,
      "num_tokens": 85103588.0,
      "step": 38840
    },
    {
      "entropy": 5.67170934677124,
      "epoch": 3.855200476379516,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003582297803889316,
      "loss": 5.0685,
      "mean_token_accuracy": 0.18758113384246827,
      "num_tokens": 85115403.0,
      "step": 38845
    },
    {
      "entropy": 5.64035930633545,
      "epoch": 3.8556967050416833,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003581968611042306,
      "loss": 4.93,
      "mean_token_accuracy": 0.19465560019016265,
      "num_tokens": 85125859.0,
      "step": 38850
    },
    {
      "entropy": 5.672080039978027,
      "epoch": 3.8561929337038507,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035816393975642675,
      "loss": 5.0239,
      "mean_token_accuracy": 0.18792459666728972,
      "num_tokens": 85136661.0,
      "step": 38855
    },
    {
      "entropy": 5.646504306793213,
      "epoch": 3.856689162366018,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035813101634633626,
      "loss": 5.0902,
      "mean_token_accuracy": 0.18192043155431747,
      "num_tokens": 85148570.0,
      "step": 38860
    },
    {
      "entropy": 5.693769073486328,
      "epoch": 3.8571853910281857,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035809809087477586,
      "loss": 4.9647,
      "mean_token_accuracy": 0.18747382164001464,
      "num_tokens": 85159505.0,
      "step": 38865
    },
    {
      "entropy": 5.697995233535766,
      "epoch": 3.857681619690353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003580651633425617,
      "loss": 4.9433,
      "mean_token_accuracy": 0.19362668246030806,
      "num_tokens": 85171323.0,
      "step": 38870
    },
    {
      "entropy": 5.682271194458008,
      "epoch": 3.858177848352521,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003580322337505106,
      "loss": 5.0328,
      "mean_token_accuracy": 0.19270373433828353,
      "num_tokens": 85182021.0,
      "step": 38875
    },
    {
      "entropy": 5.6446356773376465,
      "epoch": 3.8586740770146886,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00035799930209943906,
      "loss": 4.9362,
      "mean_token_accuracy": 0.19506387710571288,
      "num_tokens": 85192615.0,
      "step": 38880
    },
    {
      "entropy": 5.644015073776245,
      "epoch": 3.859170305676856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003579663683901636,
      "loss": 4.9659,
      "mean_token_accuracy": 0.17938806861639023,
      "num_tokens": 85204102.0,
      "step": 38885
    },
    {
      "entropy": 5.675383615493774,
      "epoch": 3.8596665343390235,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035793343262350094,
      "loss": 5.0045,
      "mean_token_accuracy": 0.19301500618457795,
      "num_tokens": 85213741.0,
      "step": 38890
    },
    {
      "entropy": 5.730546665191651,
      "epoch": 3.860162763001191,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003579004948002679,
      "loss": 4.998,
      "mean_token_accuracy": 0.19510674029588698,
      "num_tokens": 85225425.0,
      "step": 38895
    },
    {
      "entropy": 5.6581943988800045,
      "epoch": 3.8606589916633585,
      "grad_norm": 1.203125,
      "learning_rate": 0.00035786755492128134,
      "loss": 4.9373,
      "mean_token_accuracy": 0.1936718702316284,
      "num_tokens": 85236160.0,
      "step": 38900
    },
    {
      "entropy": 5.625641107559204,
      "epoch": 3.861155220325526,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003578346129873578,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1876315727829933,
      "num_tokens": 85246312.0,
      "step": 38905
    },
    {
      "entropy": 5.706946325302124,
      "epoch": 3.8616514489876934,
      "grad_norm": 1.15625,
      "learning_rate": 0.00035780166899931437,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17873653173446655,
      "num_tokens": 85258141.0,
      "step": 38910
    },
    {
      "entropy": 5.679397678375244,
      "epoch": 3.862147677649861,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003577687229579681,
      "loss": 4.9497,
      "mean_token_accuracy": 0.19314210414886473,
      "num_tokens": 85269000.0,
      "step": 38915
    },
    {
      "entropy": 5.733306169509888,
      "epoch": 3.8626439063120284,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003577357748641357,
      "loss": 4.9662,
      "mean_token_accuracy": 0.18998100012540817,
      "num_tokens": 85282290.0,
      "step": 38920
    },
    {
      "entropy": 5.615349817276001,
      "epoch": 3.863140134974196,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003577028247186344,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18499470502138138,
      "num_tokens": 85293517.0,
      "step": 38925
    },
    {
      "entropy": 5.681844329833984,
      "epoch": 3.8636363636363638,
      "grad_norm": 1.125,
      "learning_rate": 0.00035766987252228133,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1858261540532112,
      "num_tokens": 85304928.0,
      "step": 38930
    },
    {
      "entropy": 5.797132444381714,
      "epoch": 3.8641325922985312,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003576369182758935,
      "loss": 5.0618,
      "mean_token_accuracy": 0.1897641271352768,
      "num_tokens": 85315974.0,
      "step": 38935
    },
    {
      "entropy": 5.730095958709716,
      "epoch": 3.8646288209606987,
      "grad_norm": 1.25,
      "learning_rate": 0.0003576039619802882,
      "loss": 4.9544,
      "mean_token_accuracy": 0.18936989903450013,
      "num_tokens": 85325920.0,
      "step": 38940
    },
    {
      "entropy": 5.684758949279785,
      "epoch": 3.865125049622866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003575710036362826,
      "loss": 4.9643,
      "mean_token_accuracy": 0.1881942480802536,
      "num_tokens": 85337300.0,
      "step": 38945
    },
    {
      "entropy": 5.625458765029907,
      "epoch": 3.8656212782850337,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003575380432446941,
      "loss": 4.9395,
      "mean_token_accuracy": 0.19257063120603563,
      "num_tokens": 85348026.0,
      "step": 38950
    },
    {
      "entropy": 5.674749279022217,
      "epoch": 3.866117506947201,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003575050808063399,
      "loss": 5.0286,
      "mean_token_accuracy": 0.18954313695430755,
      "num_tokens": 85359835.0,
      "step": 38955
    },
    {
      "entropy": 5.6432898998260494,
      "epoch": 3.866613735609369,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035747211632203757,
      "loss": 4.945,
      "mean_token_accuracy": 0.19256002008914946,
      "num_tokens": 85370406.0,
      "step": 38960
    },
    {
      "entropy": 5.614468336105347,
      "epoch": 3.8671099642715365,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003574391497926044,
      "loss": 4.8706,
      "mean_token_accuracy": 0.19622685611248017,
      "num_tokens": 85381292.0,
      "step": 38965
    },
    {
      "entropy": 5.598035717010498,
      "epoch": 3.867606192933704,
      "grad_norm": 1.125,
      "learning_rate": 0.000357406181218858,
      "loss": 4.8904,
      "mean_token_accuracy": 0.19839369356632233,
      "num_tokens": 85392158.0,
      "step": 38970
    },
    {
      "entropy": 5.574085521697998,
      "epoch": 3.8681024215958715,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003573732106016158,
      "loss": 4.8848,
      "mean_token_accuracy": 0.19663214832544326,
      "num_tokens": 85403017.0,
      "step": 38975
    },
    {
      "entropy": 5.651254081726075,
      "epoch": 3.868598650258039,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003573402379416955,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1887944981455803,
      "num_tokens": 85412682.0,
      "step": 38980
    },
    {
      "entropy": 5.703349828720093,
      "epoch": 3.8690948789202064,
      "grad_norm": 1.25,
      "learning_rate": 0.00035730726323991474,
      "loss": 4.9576,
      "mean_token_accuracy": 0.18908325731754302,
      "num_tokens": 85422407.0,
      "step": 38985
    },
    {
      "entropy": 5.676379060745239,
      "epoch": 3.869591107582374,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003572742864970913,
      "loss": 4.9234,
      "mean_token_accuracy": 0.19635604172945023,
      "num_tokens": 85433277.0,
      "step": 38990
    },
    {
      "entropy": 5.636298274993896,
      "epoch": 3.8700873362445414,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003572413077140427,
      "loss": 5.0018,
      "mean_token_accuracy": 0.18911528289318086,
      "num_tokens": 85443995.0,
      "step": 38995
    },
    {
      "entropy": 5.677249765396118,
      "epoch": 3.870583564906709,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357208326891587,
      "loss": 4.902,
      "mean_token_accuracy": 0.19259553849697114,
      "num_tokens": 85454413.0,
      "step": 39000
    },
    {
      "epoch": 3.870583564906709,
      "eval_entropy": 5.464777504754877,
      "eval_loss": 5.110031604766846,
      "eval_mean_token_accuracy": 0.18963164878297967,
      "eval_num_tokens": 85454413.0,
      "eval_runtime": 20.1238,
      "eval_samples_per_second": 1673.095,
      "eval_steps_per_second": 209.156,
      "step": 39000
    }
  ],
  "logging_steps": 5,
  "max_steps": 100750,
  "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": 2.12988090986496e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}