ita-latn-10mb-after-ppt-shuff-dyck-100mb-ckpt500_seed455/checkpoint-6000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5954743946010321,
  "eval_steps": 3000,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.931299161911011,
      "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
    }
  ],
  "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": 3271861646622720.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}