{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 553,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 1.4297935962677002,
      "learning_rate": 0.0,
      "loss": 2.1269,
      "mean_token_accuracy": 0.5883483290672302,
      "num_tokens": 45860.0,
      "step": 1
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 1.434566855430603,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 2.0758,
      "mean_token_accuracy": 0.5955607295036316,
      "num_tokens": 91607.0,
      "step": 2
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 1.34773850440979,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 2.0995,
      "mean_token_accuracy": 0.5925061702728271,
      "num_tokens": 138803.0,
      "step": 3
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 1.2693862915039062,
      "learning_rate": 3.529411764705883e-05,
      "loss": 2.0646,
      "mean_token_accuracy": 0.592647910118103,
      "num_tokens": 185711.0,
      "step": 4
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.0771547555923462,
      "learning_rate": 4.705882352941177e-05,
      "loss": 1.9528,
      "mean_token_accuracy": 0.6024201512336731,
      "num_tokens": 229326.0,
      "step": 5
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.8461045026779175,
      "learning_rate": 5.882352941176471e-05,
      "loss": 1.8946,
      "mean_token_accuracy": 0.6050652861595154,
      "num_tokens": 273139.0,
      "step": 6
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.5019454956054688,
      "learning_rate": 7.058823529411765e-05,
      "loss": 1.833,
      "mean_token_accuracy": 0.6073743104934692,
      "num_tokens": 320421.0,
      "step": 7
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.5096880197525024,
      "learning_rate": 8.23529411764706e-05,
      "loss": 1.8264,
      "mean_token_accuracy": 0.604993462562561,
      "num_tokens": 368788.0,
      "step": 8
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.8427304029464722,
      "learning_rate": 9.411764705882353e-05,
      "loss": 1.8343,
      "mean_token_accuracy": 0.6039018630981445,
      "num_tokens": 415189.0,
      "step": 9
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.8095427751541138,
      "learning_rate": 0.00010588235294117647,
      "loss": 1.7456,
      "mean_token_accuracy": 0.6184370517730713,
      "num_tokens": 458774.0,
      "step": 10
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.5690692067146301,
      "learning_rate": 0.00011764705882352942,
      "loss": 1.761,
      "mean_token_accuracy": 0.6134573817253113,
      "num_tokens": 504300.0,
      "step": 11
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.42741021513938904,
      "learning_rate": 0.00012941176470588237,
      "loss": 1.7408,
      "mean_token_accuracy": 0.6172425150871277,
      "num_tokens": 550239.0,
      "step": 12
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.3621617257595062,
      "learning_rate": 0.0001411764705882353,
      "loss": 1.7292,
      "mean_token_accuracy": 0.6186677813529968,
      "num_tokens": 595311.0,
      "step": 13
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.356418639421463,
      "learning_rate": 0.00015294117647058822,
      "loss": 1.7433,
      "mean_token_accuracy": 0.618061900138855,
      "num_tokens": 643731.0,
      "step": 14
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.31577178835868835,
      "learning_rate": 0.0001647058823529412,
      "loss": 1.7376,
      "mean_token_accuracy": 0.6176530718803406,
      "num_tokens": 687946.0,
      "step": 15
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.28681808710098267,
      "learning_rate": 0.00017647058823529413,
      "loss": 1.7212,
      "mean_token_accuracy": 0.6185654997825623,
      "num_tokens": 733740.0,
      "step": 16
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.3112754821777344,
      "learning_rate": 0.00018823529411764707,
      "loss": 1.7352,
      "mean_token_accuracy": 0.6181626319885254,
      "num_tokens": 780185.0,
      "step": 17
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.40416133403778076,
      "learning_rate": 0.0002,
      "loss": 1.7291,
      "mean_token_accuracy": 0.618687093257904,
      "num_tokens": 827533.0,
      "step": 18
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.45459505915641785,
      "learning_rate": 0.0001999982823331779,
      "loss": 1.7223,
      "mean_token_accuracy": 0.6227962970733643,
      "num_tokens": 873543.0,
      "step": 19
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.37240275740623474,
      "learning_rate": 0.00019999312939171914,
      "loss": 1.7479,
      "mean_token_accuracy": 0.6146200895309448,
      "num_tokens": 923210.0,
      "step": 20
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.5440695881843567,
      "learning_rate": 0.00019998454135264444,
      "loss": 1.6538,
      "mean_token_accuracy": 0.6290217638015747,
      "num_tokens": 969460.0,
      "step": 21
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.2815147638320923,
      "learning_rate": 0.0001999725185109816,
      "loss": 1.6833,
      "mean_token_accuracy": 0.6285157799720764,
      "num_tokens": 1014215.0,
      "step": 22
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.27108052372932434,
      "learning_rate": 0.00019995706127975537,
      "loss": 1.6823,
      "mean_token_accuracy": 0.6276950836181641,
      "num_tokens": 1061587.0,
      "step": 23
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.28978854417800903,
      "learning_rate": 0.00019993817018997323,
      "loss": 1.7345,
      "mean_token_accuracy": 0.6147149205207825,
      "num_tokens": 1108746.0,
      "step": 24
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.26003706455230713,
      "learning_rate": 0.0001999158458906071,
      "loss": 1.6382,
      "mean_token_accuracy": 0.633102297782898,
      "num_tokens": 1150650.0,
      "step": 25
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.2689721882343292,
      "learning_rate": 0.00019989008914857116,
      "loss": 1.7528,
      "mean_token_accuracy": 0.6133280396461487,
      "num_tokens": 1198478.0,
      "step": 26
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.29233652353286743,
      "learning_rate": 0.00019986090084869545,
      "loss": 1.7128,
      "mean_token_accuracy": 0.6206752061843872,
      "num_tokens": 1248602.0,
      "step": 27
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.3018750846385956,
      "learning_rate": 0.00019982828199369541,
      "loss": 1.7215,
      "mean_token_accuracy": 0.6194838285446167,
      "num_tokens": 1298225.0,
      "step": 28
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.3065217435359955,
      "learning_rate": 0.00019979223370413763,
      "loss": 1.642,
      "mean_token_accuracy": 0.632573127746582,
      "num_tokens": 1346783.0,
      "step": 29
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.26041898131370544,
      "learning_rate": 0.00019975275721840103,
      "loss": 1.7038,
      "mean_token_accuracy": 0.6213216781616211,
      "num_tokens": 1394151.0,
      "step": 30
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.25361573696136475,
      "learning_rate": 0.00019970985389263467,
      "loss": 1.6846,
      "mean_token_accuracy": 0.6269124150276184,
      "num_tokens": 1440818.0,
      "step": 31
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.23808851838111877,
      "learning_rate": 0.0001996635252007109,
      "loss": 1.7003,
      "mean_token_accuracy": 0.6228302121162415,
      "num_tokens": 1486970.0,
      "step": 32
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.2258971631526947,
      "learning_rate": 0.00019961377273417487,
      "loss": 1.6915,
      "mean_token_accuracy": 0.6242671012878418,
      "num_tokens": 1535132.0,
      "step": 33
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.23645015060901642,
      "learning_rate": 0.00019956059820218982,
      "loss": 1.6701,
      "mean_token_accuracy": 0.6313917636871338,
      "num_tokens": 1581310.0,
      "step": 34
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.22955374419689178,
      "learning_rate": 0.00019950400343147833,
      "loss": 1.6651,
      "mean_token_accuracy": 0.6299709677696228,
      "num_tokens": 1626838.0,
      "step": 35
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.22750572860240936,
      "learning_rate": 0.0001994439903662596,
      "loss": 1.6982,
      "mean_token_accuracy": 0.6246239542961121,
      "num_tokens": 1674100.0,
      "step": 36
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.22190001606941223,
      "learning_rate": 0.00019938056106818261,
      "loss": 1.6975,
      "mean_token_accuracy": 0.6197924017906189,
      "num_tokens": 1722819.0,
      "step": 37
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.24010570347309113,
      "learning_rate": 0.00019931371771625544,
      "loss": 1.6365,
      "mean_token_accuracy": 0.6325092911720276,
      "num_tokens": 1770008.0,
      "step": 38
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.23719368875026703,
      "learning_rate": 0.0001992434626067702,
      "loss": 1.6906,
      "mean_token_accuracy": 0.6234063506126404,
      "num_tokens": 1815094.0,
      "step": 39
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.22259032726287842,
      "learning_rate": 0.00019916979815322433,
      "loss": 1.6798,
      "mean_token_accuracy": 0.6275040507316589,
      "num_tokens": 1865027.0,
      "step": 40
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.2300175577402115,
      "learning_rate": 0.00019909272688623756,
      "loss": 1.6783,
      "mean_token_accuracy": 0.6286372542381287,
      "num_tokens": 1911967.0,
      "step": 41
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.21260212361812592,
      "learning_rate": 0.0001990122514534651,
      "loss": 1.7155,
      "mean_token_accuracy": 0.6216686964035034,
      "num_tokens": 1963100.0,
      "step": 42
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.2217252254486084,
      "learning_rate": 0.00019892837461950652,
      "loss": 1.6599,
      "mean_token_accuracy": 0.6299470663070679,
      "num_tokens": 2010768.0,
      "step": 43
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.2244565337896347,
      "learning_rate": 0.00019884109926581096,
      "loss": 1.7049,
      "mean_token_accuracy": 0.622722864151001,
      "num_tokens": 2059303.0,
      "step": 44
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.23472081124782562,
      "learning_rate": 0.00019875042839057798,
      "loss": 1.7017,
      "mean_token_accuracy": 0.6219018697738647,
      "num_tokens": 2103829.0,
      "step": 45
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.22766339778900146,
      "learning_rate": 0.00019865636510865467,
      "loss": 1.7028,
      "mean_token_accuracy": 0.6218793392181396,
      "num_tokens": 2150558.0,
      "step": 46
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.228139728307724,
      "learning_rate": 0.0001985589126514286,
      "loss": 1.7016,
      "mean_token_accuracy": 0.6203511953353882,
      "num_tokens": 2201253.0,
      "step": 47
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.22668923437595367,
      "learning_rate": 0.0001984580743667168,
      "loss": 1.6959,
      "mean_token_accuracy": 0.6244654059410095,
      "num_tokens": 2247382.0,
      "step": 48
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.2259022295475006,
      "learning_rate": 0.0001983538537186508,
      "loss": 1.6974,
      "mean_token_accuracy": 0.6239613890647888,
      "num_tokens": 2295707.0,
      "step": 49
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.22462861239910126,
      "learning_rate": 0.0001982462542875576,
      "loss": 1.6612,
      "mean_token_accuracy": 0.6296576857566833,
      "num_tokens": 2340697.0,
      "step": 50
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.24034149944782257,
      "learning_rate": 0.0001981352797698367,
      "loss": 1.6784,
      "mean_token_accuracy": 0.6287301182746887,
      "num_tokens": 2382717.0,
      "step": 51
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.21968059241771698,
      "learning_rate": 0.00019802093397783296,
      "loss": 1.6612,
      "mean_token_accuracy": 0.6290526390075684,
      "num_tokens": 2430219.0,
      "step": 52
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.23564419150352478,
      "learning_rate": 0.0001979032208397059,
      "loss": 1.6276,
      "mean_token_accuracy": 0.6347538232803345,
      "num_tokens": 2473766.0,
      "step": 53
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.22665007412433624,
      "learning_rate": 0.00019778214439929452,
      "loss": 1.6862,
      "mean_token_accuracy": 0.6270169615745544,
      "num_tokens": 2521674.0,
      "step": 54
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.2169383466243744,
      "learning_rate": 0.00019765770881597855,
      "loss": 1.707,
      "mean_token_accuracy": 0.624442458152771,
      "num_tokens": 2570612.0,
      "step": 55
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.22436678409576416,
      "learning_rate": 0.00019752991836453543,
      "loss": 1.6428,
      "mean_token_accuracy": 0.6327121257781982,
      "num_tokens": 2620095.0,
      "step": 56
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.23180508613586426,
      "learning_rate": 0.00019739877743499352,
      "loss": 1.7243,
      "mean_token_accuracy": 0.6180028915405273,
      "num_tokens": 2665785.0,
      "step": 57
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.2244236022233963,
      "learning_rate": 0.0001972642905324813,
      "loss": 1.6619,
      "mean_token_accuracy": 0.6294060349464417,
      "num_tokens": 2712064.0,
      "step": 58
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.2267381250858307,
      "learning_rate": 0.00019712646227707263,
      "loss": 1.6287,
      "mean_token_accuracy": 0.6353709697723389,
      "num_tokens": 2760026.0,
      "step": 59
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.23568469285964966,
      "learning_rate": 0.00019698529740362785,
      "loss": 1.7067,
      "mean_token_accuracy": 0.6213864088058472,
      "num_tokens": 2806166.0,
      "step": 60
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.21780051290988922,
      "learning_rate": 0.00019684080076163142,
      "loss": 1.6478,
      "mean_token_accuracy": 0.630071759223938,
      "num_tokens": 2855572.0,
      "step": 61
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.23158034682273865,
      "learning_rate": 0.00019669297731502507,
      "loss": 1.6464,
      "mean_token_accuracy": 0.6330370903015137,
      "num_tokens": 2900646.0,
      "step": 62
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.22072935104370117,
      "learning_rate": 0.0001965418321420374,
      "loss": 1.6329,
      "mean_token_accuracy": 0.6327118277549744,
      "num_tokens": 2948656.0,
      "step": 63
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.23062308132648468,
      "learning_rate": 0.0001963873704350094,
      "loss": 1.6534,
      "mean_token_accuracy": 0.6293391585350037,
      "num_tokens": 2992709.0,
      "step": 64
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.21779538691043854,
      "learning_rate": 0.00019622959750021605,
      "loss": 1.618,
      "mean_token_accuracy": 0.6372920274734497,
      "num_tokens": 3038992.0,
      "step": 65
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.22693635523319244,
      "learning_rate": 0.000196068518757684,
      "loss": 1.7128,
      "mean_token_accuracy": 0.6208335161209106,
      "num_tokens": 3085558.0,
      "step": 66
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.23582525551319122,
      "learning_rate": 0.0001959041397410056,
      "loss": 1.6625,
      "mean_token_accuracy": 0.6306231617927551,
      "num_tokens": 3130170.0,
      "step": 67
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.2246374785900116,
      "learning_rate": 0.0001957364660971485,
      "loss": 1.6385,
      "mean_token_accuracy": 0.6291642189025879,
      "num_tokens": 3176311.0,
      "step": 68
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.22443550825119019,
      "learning_rate": 0.0001955655035862617,
      "loss": 1.6751,
      "mean_token_accuracy": 0.625806450843811,
      "num_tokens": 3222875.0,
      "step": 69
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.22177176177501678,
      "learning_rate": 0.0001953912580814779,
      "loss": 1.6653,
      "mean_token_accuracy": 0.6286900639533997,
      "num_tokens": 3271109.0,
      "step": 70
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.2268744707107544,
      "learning_rate": 0.0001952137355687116,
      "loss": 1.6222,
      "mean_token_accuracy": 0.6369240283966064,
      "num_tokens": 3317571.0,
      "step": 71
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.22358258068561554,
      "learning_rate": 0.00019503294214645337,
      "loss": 1.6428,
      "mean_token_accuracy": 0.633510172367096,
      "num_tokens": 3362493.0,
      "step": 72
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.2274637073278427,
      "learning_rate": 0.00019484888402556045,
      "loss": 1.6303,
      "mean_token_accuracy": 0.6295872926712036,
      "num_tokens": 3412450.0,
      "step": 73
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.2235092669725418,
      "learning_rate": 0.00019466156752904343,
      "loss": 1.6348,
      "mean_token_accuracy": 0.6336225867271423,
      "num_tokens": 3458150.0,
      "step": 74
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.22723203897476196,
      "learning_rate": 0.0001944709990918489,
      "loss": 1.6353,
      "mean_token_accuracy": 0.6303831934928894,
      "num_tokens": 3503285.0,
      "step": 75
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.23029087483882904,
      "learning_rate": 0.00019427718526063856,
      "loss": 1.6788,
      "mean_token_accuracy": 0.6246349215507507,
      "num_tokens": 3550258.0,
      "step": 76
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.21722035109996796,
      "learning_rate": 0.00019408013269356408,
      "loss": 1.6423,
      "mean_token_accuracy": 0.6343802809715271,
      "num_tokens": 3600054.0,
      "step": 77
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.22306939959526062,
      "learning_rate": 0.00019387984816003867,
      "loss": 1.6825,
      "mean_token_accuracy": 0.6262245774269104,
      "num_tokens": 3646563.0,
      "step": 78
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.24767781794071198,
      "learning_rate": 0.00019367633854050422,
      "loss": 1.6458,
      "mean_token_accuracy": 0.6306663751602173,
      "num_tokens": 3693284.0,
      "step": 79
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.2704238295555115,
      "learning_rate": 0.00019346961082619522,
      "loss": 1.5187,
      "mean_token_accuracy": 0.6579475402832031,
      "num_tokens": 3738689.0,
      "step": 80
    },
    {
      "epoch": 1.0253164556962024,
      "grad_norm": 0.22565050423145294,
      "learning_rate": 0.00019325967211889834,
      "loss": 1.5333,
      "mean_token_accuracy": 0.6507334113121033,
      "num_tokens": 3783544.0,
      "step": 81
    },
    {
      "epoch": 1.0379746835443038,
      "grad_norm": 0.21894116699695587,
      "learning_rate": 0.0001930465296307087,
      "loss": 1.4748,
      "mean_token_accuracy": 0.6617191433906555,
      "num_tokens": 3830433.0,
      "step": 82
    },
    {
      "epoch": 1.0506329113924051,
      "grad_norm": 0.2757681906223297,
      "learning_rate": 0.00019283019068378182,
      "loss": 1.5382,
      "mean_token_accuracy": 0.6483068466186523,
      "num_tokens": 3876830.0,
      "step": 83
    },
    {
      "epoch": 1.0632911392405062,
      "grad_norm": 0.26444658637046814,
      "learning_rate": 0.00019261066271008235,
      "loss": 1.4792,
      "mean_token_accuracy": 0.6634548902511597,
      "num_tokens": 3921833.0,
      "step": 84
    },
    {
      "epoch": 1.0759493670886076,
      "grad_norm": 0.23986731469631195,
      "learning_rate": 0.0001923879532511287,
      "loss": 1.5573,
      "mean_token_accuracy": 0.6481124758720398,
      "num_tokens": 3968122.0,
      "step": 85
    },
    {
      "epoch": 1.0886075949367089,
      "grad_norm": 0.23939338326454163,
      "learning_rate": 0.00019216206995773373,
      "loss": 1.5157,
      "mean_token_accuracy": 0.6560894846916199,
      "num_tokens": 4016466.0,
      "step": 86
    },
    {
      "epoch": 1.1012658227848102,
      "grad_norm": 0.24986766278743744,
      "learning_rate": 0.00019193302058974232,
      "loss": 1.5496,
      "mean_token_accuracy": 0.6462620496749878,
      "num_tokens": 4066437.0,
      "step": 87
    },
    {
      "epoch": 1.1139240506329113,
      "grad_norm": 0.23949536681175232,
      "learning_rate": 0.00019170081301576444,
      "loss": 1.5063,
      "mean_token_accuracy": 0.6563820242881775,
      "num_tokens": 4114793.0,
      "step": 88
    },
    {
      "epoch": 1.1265822784810127,
      "grad_norm": 0.23770886659622192,
      "learning_rate": 0.00019146545521290495,
      "loss": 1.4886,
      "mean_token_accuracy": 0.6591377854347229,
      "num_tokens": 4159441.0,
      "step": 89
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 0.2399303913116455,
      "learning_rate": 0.00019122695526648968,
      "loss": 1.5237,
      "mean_token_accuracy": 0.650536298751831,
      "num_tokens": 4208729.0,
      "step": 90
    },
    {
      "epoch": 1.1518987341772151,
      "grad_norm": 0.2637503147125244,
      "learning_rate": 0.00019098532136978754,
      "loss": 1.498,
      "mean_token_accuracy": 0.6567501425743103,
      "num_tokens": 4253466.0,
      "step": 91
    },
    {
      "epoch": 1.1645569620253164,
      "grad_norm": 0.2615419924259186,
      "learning_rate": 0.00019074056182372907,
      "loss": 1.5141,
      "mean_token_accuracy": 0.6535509824752808,
      "num_tokens": 4302784.0,
      "step": 92
    },
    {
      "epoch": 1.1772151898734178,
      "grad_norm": 0.2541172504425049,
      "learning_rate": 0.00019049268503662126,
      "loss": 1.4986,
      "mean_token_accuracy": 0.6578357815742493,
      "num_tokens": 4347917.0,
      "step": 93
    },
    {
      "epoch": 1.189873417721519,
      "grad_norm": 0.24259509146213531,
      "learning_rate": 0.00019024169952385885,
      "loss": 1.4655,
      "mean_token_accuracy": 0.6630852222442627,
      "num_tokens": 4394610.0,
      "step": 94
    },
    {
      "epoch": 1.2025316455696202,
      "grad_norm": 0.2532861530780792,
      "learning_rate": 0.00018998761390763154,
      "loss": 1.5144,
      "mean_token_accuracy": 0.6535828709602356,
      "num_tokens": 4443029.0,
      "step": 95
    },
    {
      "epoch": 1.2151898734177216,
      "grad_norm": 0.24317510426044464,
      "learning_rate": 0.00018973043691662803,
      "loss": 1.5196,
      "mean_token_accuracy": 0.6520761847496033,
      "num_tokens": 4490704.0,
      "step": 96
    },
    {
      "epoch": 1.2278481012658227,
      "grad_norm": 0.2610395848751068,
      "learning_rate": 0.000189470177385736,
      "loss": 1.5044,
      "mean_token_accuracy": 0.6601476669311523,
      "num_tokens": 4534384.0,
      "step": 97
    },
    {
      "epoch": 1.240506329113924,
      "grad_norm": 0.2677130103111267,
      "learning_rate": 0.00018920684425573865,
      "loss": 1.4518,
      "mean_token_accuracy": 0.6647972464561462,
      "num_tokens": 4578669.0,
      "step": 98
    },
    {
      "epoch": 1.2531645569620253,
      "grad_norm": 0.2610534429550171,
      "learning_rate": 0.00018894044657300765,
      "loss": 1.4535,
      "mean_token_accuracy": 0.6645317077636719,
      "num_tokens": 4624636.0,
      "step": 99
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 0.2596592903137207,
      "learning_rate": 0.00018867099348919217,
      "loss": 1.47,
      "mean_token_accuracy": 0.6623217463493347,
      "num_tokens": 4671958.0,
      "step": 100
    },
    {
      "epoch": 1.2784810126582278,
      "grad_norm": 0.2679537236690521,
      "learning_rate": 0.0001883984942609047,
      "loss": 1.473,
      "mean_token_accuracy": 0.6634280681610107,
      "num_tokens": 4718853.0,
      "step": 101
    },
    {
      "epoch": 1.2911392405063291,
      "grad_norm": 0.26461848616600037,
      "learning_rate": 0.00018812295824940285,
      "loss": 1.5248,
      "mean_token_accuracy": 0.6544753909111023,
      "num_tokens": 4767695.0,
      "step": 102
    },
    {
      "epoch": 1.3037974683544304,
      "grad_norm": 0.2688470482826233,
      "learning_rate": 0.00018784439492026798,
      "loss": 1.553,
      "mean_token_accuracy": 0.6471455693244934,
      "num_tokens": 4813880.0,
      "step": 103
    },
    {
      "epoch": 1.3164556962025316,
      "grad_norm": 0.2555462419986725,
      "learning_rate": 0.00018756281384307982,
      "loss": 1.4992,
      "mean_token_accuracy": 0.656396210193634,
      "num_tokens": 4860323.0,
      "step": 104
    },
    {
      "epoch": 1.3291139240506329,
      "grad_norm": 0.27267542481422424,
      "learning_rate": 0.0001872782246910879,
      "loss": 1.453,
      "mean_token_accuracy": 0.6634340882301331,
      "num_tokens": 4906274.0,
      "step": 105
    },
    {
      "epoch": 1.3417721518987342,
      "grad_norm": 0.2674024999141693,
      "learning_rate": 0.00018699063724087904,
      "loss": 1.5336,
      "mean_token_accuracy": 0.652473509311676,
      "num_tokens": 4954225.0,
      "step": 106
    },
    {
      "epoch": 1.3544303797468356,
      "grad_norm": 0.2559143602848053,
      "learning_rate": 0.0001867000613720417,
      "loss": 1.4769,
      "mean_token_accuracy": 0.6595107316970825,
      "num_tokens": 5001336.0,
      "step": 107
    },
    {
      "epoch": 1.3670886075949367,
      "grad_norm": 0.27574166655540466,
      "learning_rate": 0.0001864065070668265,
      "loss": 1.4803,
      "mean_token_accuracy": 0.661461591720581,
      "num_tokens": 5047445.0,
      "step": 108
    },
    {
      "epoch": 1.379746835443038,
      "grad_norm": 0.2721637189388275,
      "learning_rate": 0.00018610998440980324,
      "loss": 1.4955,
      "mean_token_accuracy": 0.6583252549171448,
      "num_tokens": 5092745.0,
      "step": 109
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 0.2680104076862335,
      "learning_rate": 0.00018581050358751445,
      "loss": 1.5764,
      "mean_token_accuracy": 0.6440534591674805,
      "num_tokens": 5140038.0,
      "step": 110
    },
    {
      "epoch": 1.4050632911392404,
      "grad_norm": 0.2686646580696106,
      "learning_rate": 0.00018550807488812562,
      "loss": 1.4896,
      "mean_token_accuracy": 0.6607694625854492,
      "num_tokens": 5185717.0,
      "step": 111
    },
    {
      "epoch": 1.4177215189873418,
      "grad_norm": 0.26218053698539734,
      "learning_rate": 0.00018520270870107166,
      "loss": 1.5112,
      "mean_token_accuracy": 0.6550983786582947,
      "num_tokens": 5230040.0,
      "step": 112
    },
    {
      "epoch": 1.4303797468354431,
      "grad_norm": 0.2599179446697235,
      "learning_rate": 0.00018489441551669986,
      "loss": 1.4914,
      "mean_token_accuracy": 0.6583897471427917,
      "num_tokens": 5279599.0,
      "step": 113
    },
    {
      "epoch": 1.4430379746835442,
      "grad_norm": 0.25936904549598694,
      "learning_rate": 0.00018458320592590975,
      "loss": 1.491,
      "mean_token_accuracy": 0.6570174098014832,
      "num_tokens": 5326817.0,
      "step": 114
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 0.2672724425792694,
      "learning_rate": 0.00018426909061978908,
      "loss": 1.5042,
      "mean_token_accuracy": 0.6552169322967529,
      "num_tokens": 5374131.0,
      "step": 115
    },
    {
      "epoch": 1.4683544303797469,
      "grad_norm": 0.26219993829727173,
      "learning_rate": 0.00018395208038924667,
      "loss": 1.457,
      "mean_token_accuracy": 0.664246141910553,
      "num_tokens": 5421155.0,
      "step": 116
    },
    {
      "epoch": 1.481012658227848,
      "grad_norm": 0.2645889222621918,
      "learning_rate": 0.00018363218612464158,
      "loss": 1.4865,
      "mean_token_accuracy": 0.6586675643920898,
      "num_tokens": 5468516.0,
      "step": 117
    },
    {
      "epoch": 1.4936708860759493,
      "grad_norm": 0.27145904302597046,
      "learning_rate": 0.00018330941881540915,
      "loss": 1.5393,
      "mean_token_accuracy": 0.6498931646347046,
      "num_tokens": 5519116.0,
      "step": 118
    },
    {
      "epoch": 1.5063291139240507,
      "grad_norm": 0.27031463384628296,
      "learning_rate": 0.00018298378954968337,
      "loss": 1.4729,
      "mean_token_accuracy": 0.6622894406318665,
      "num_tokens": 5566013.0,
      "step": 119
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.2636566460132599,
      "learning_rate": 0.0001826553095139159,
      "loss": 1.4852,
      "mean_token_accuracy": 0.6582340002059937,
      "num_tokens": 5613393.0,
      "step": 120
    },
    {
      "epoch": 1.5316455696202531,
      "grad_norm": 0.26901474595069885,
      "learning_rate": 0.00018232398999249192,
      "loss": 1.4343,
      "mean_token_accuracy": 0.6671172976493835,
      "num_tokens": 5658581.0,
      "step": 121
    },
    {
      "epoch": 1.5443037974683544,
      "grad_norm": 0.260540634393692,
      "learning_rate": 0.00018198984236734246,
      "loss": 1.5081,
      "mean_token_accuracy": 0.6527827978134155,
      "num_tokens": 5708631.0,
      "step": 122
    },
    {
      "epoch": 1.5569620253164556,
      "grad_norm": 0.267051100730896,
      "learning_rate": 0.0001816528781175533,
      "loss": 1.479,
      "mean_token_accuracy": 0.6584655046463013,
      "num_tokens": 5757428.0,
      "step": 123
    },
    {
      "epoch": 1.5696202531645569,
      "grad_norm": 0.2814350724220276,
      "learning_rate": 0.0001813131088189707,
      "loss": 1.4711,
      "mean_token_accuracy": 0.6626920104026794,
      "num_tokens": 5801689.0,
      "step": 124
    },
    {
      "epoch": 1.5822784810126582,
      "grad_norm": 0.2702229619026184,
      "learning_rate": 0.00018097054614380365,
      "loss": 1.4945,
      "mean_token_accuracy": 0.6577462553977966,
      "num_tokens": 5847827.0,
      "step": 125
    },
    {
      "epoch": 1.5949367088607596,
      "grad_norm": 0.2745739221572876,
      "learning_rate": 0.000180625201860223,
      "loss": 1.504,
      "mean_token_accuracy": 0.6541758179664612,
      "num_tokens": 5895606.0,
      "step": 126
    },
    {
      "epoch": 1.6075949367088609,
      "grad_norm": 0.2730543315410614,
      "learning_rate": 0.0001802770878319571,
      "loss": 1.4956,
      "mean_token_accuracy": 0.6596698760986328,
      "num_tokens": 5942501.0,
      "step": 127
    },
    {
      "epoch": 1.620253164556962,
      "grad_norm": 0.26320144534111023,
      "learning_rate": 0.00017992621601788428,
      "loss": 1.4677,
      "mean_token_accuracy": 0.6615061163902283,
      "num_tokens": 5988684.0,
      "step": 128
    },
    {
      "epoch": 1.6329113924050633,
      "grad_norm": 0.27375948429107666,
      "learning_rate": 0.00017957259847162205,
      "loss": 1.4883,
      "mean_token_accuracy": 0.6593698263168335,
      "num_tokens": 6036577.0,
      "step": 129
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 0.27551186084747314,
      "learning_rate": 0.00017921624734111292,
      "loss": 1.5127,
      "mean_token_accuracy": 0.6532179713249207,
      "num_tokens": 6082866.0,
      "step": 130
    },
    {
      "epoch": 1.6582278481012658,
      "grad_norm": 0.2854526937007904,
      "learning_rate": 0.00017885717486820722,
      "loss": 1.4834,
      "mean_token_accuracy": 0.6589472889900208,
      "num_tokens": 6128184.0,
      "step": 131
    },
    {
      "epoch": 1.6708860759493671,
      "grad_norm": 0.2762332558631897,
      "learning_rate": 0.00017849539338824231,
      "loss": 1.4762,
      "mean_token_accuracy": 0.6631950736045837,
      "num_tokens": 6175777.0,
      "step": 132
    },
    {
      "epoch": 1.6835443037974684,
      "grad_norm": 0.28698286414146423,
      "learning_rate": 0.0001781309153296192,
      "loss": 1.4743,
      "mean_token_accuracy": 0.6608595252037048,
      "num_tokens": 6221704.0,
      "step": 133
    },
    {
      "epoch": 1.6962025316455698,
      "grad_norm": 0.2894686758518219,
      "learning_rate": 0.00017776375321337521,
      "loss": 1.4985,
      "mean_token_accuracy": 0.65559983253479,
      "num_tokens": 6267430.0,
      "step": 134
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.2916942834854126,
      "learning_rate": 0.00017739391965275404,
      "loss": 1.4675,
      "mean_token_accuracy": 0.6640766859054565,
      "num_tokens": 6309965.0,
      "step": 135
    },
    {
      "epoch": 1.721518987341772,
      "grad_norm": 0.27421343326568604,
      "learning_rate": 0.00017702142735277247,
      "loss": 1.497,
      "mean_token_accuracy": 0.6565625071525574,
      "num_tokens": 6358326.0,
      "step": 136
    },
    {
      "epoch": 1.7341772151898733,
      "grad_norm": 0.2747173309326172,
      "learning_rate": 0.00017664628910978375,
      "loss": 1.4995,
      "mean_token_accuracy": 0.655762255191803,
      "num_tokens": 6405160.0,
      "step": 137
    },
    {
      "epoch": 1.7468354430379747,
      "grad_norm": 0.27782154083251953,
      "learning_rate": 0.0001762685178110382,
      "loss": 1.4358,
      "mean_token_accuracy": 0.667519211769104,
      "num_tokens": 6449795.0,
      "step": 138
    },
    {
      "epoch": 1.759493670886076,
      "grad_norm": 0.2791241705417633,
      "learning_rate": 0.00017588812643424032,
      "loss": 1.5493,
      "mean_token_accuracy": 0.6500073671340942,
      "num_tokens": 6497300.0,
      "step": 139
    },
    {
      "epoch": 1.7721518987341773,
      "grad_norm": 0.2714492380619049,
      "learning_rate": 0.0001755051280471031,
      "loss": 1.4895,
      "mean_token_accuracy": 0.6580542922019958,
      "num_tokens": 6546518.0,
      "step": 140
    },
    {
      "epoch": 1.7848101265822784,
      "grad_norm": 0.27369654178619385,
      "learning_rate": 0.00017511953580689888,
      "loss": 1.5208,
      "mean_token_accuracy": 0.6513364315032959,
      "num_tokens": 6593048.0,
      "step": 141
    },
    {
      "epoch": 1.7974683544303798,
      "grad_norm": 0.2905201017856598,
      "learning_rate": 0.00017473136296000772,
      "loss": 1.5268,
      "mean_token_accuracy": 0.6505144238471985,
      "num_tokens": 6638988.0,
      "step": 142
    },
    {
      "epoch": 1.810126582278481,
      "grad_norm": 0.2680700719356537,
      "learning_rate": 0.000174340622841462,
      "loss": 1.4976,
      "mean_token_accuracy": 0.6543448567390442,
      "num_tokens": 6686177.0,
      "step": 143
    },
    {
      "epoch": 1.8227848101265822,
      "grad_norm": 0.28409284353256226,
      "learning_rate": 0.00017394732887448847,
      "loss": 1.4865,
      "mean_token_accuracy": 0.6597988605499268,
      "num_tokens": 6731379.0,
      "step": 144
    },
    {
      "epoch": 1.8354430379746836,
      "grad_norm": 0.2727820575237274,
      "learning_rate": 0.00017355149457004709,
      "loss": 1.4922,
      "mean_token_accuracy": 0.6577648520469666,
      "num_tokens": 6779860.0,
      "step": 145
    },
    {
      "epoch": 1.8481012658227849,
      "grad_norm": 0.2895708382129669,
      "learning_rate": 0.0001731531335263669,
      "loss": 1.4861,
      "mean_token_accuracy": 0.6559346318244934,
      "num_tokens": 6822631.0,
      "step": 146
    },
    {
      "epoch": 1.8607594936708862,
      "grad_norm": 0.2730681300163269,
      "learning_rate": 0.0001727522594284789,
      "loss": 1.5365,
      "mean_token_accuracy": 0.6517869830131531,
      "num_tokens": 6871240.0,
      "step": 147
    },
    {
      "epoch": 1.8734177215189873,
      "grad_norm": 0.27399033308029175,
      "learning_rate": 0.00017234888604774574,
      "loss": 1.5494,
      "mean_token_accuracy": 0.6449940800666809,
      "num_tokens": 6920506.0,
      "step": 148
    },
    {
      "epoch": 1.8860759493670884,
      "grad_norm": 0.28052183985710144,
      "learning_rate": 0.00017194302724138903,
      "loss": 1.5205,
      "mean_token_accuracy": 0.6526179313659668,
      "num_tokens": 6968451.0,
      "step": 149
    },
    {
      "epoch": 1.8987341772151898,
      "grad_norm": 0.2830485999584198,
      "learning_rate": 0.00017153469695201277,
      "loss": 1.534,
      "mean_token_accuracy": 0.6482208371162415,
      "num_tokens": 7015391.0,
      "step": 150
    },
    {
      "epoch": 1.9113924050632911,
      "grad_norm": 0.2714807689189911,
      "learning_rate": 0.0001711239092071248,
      "loss": 1.5007,
      "mean_token_accuracy": 0.6547030210494995,
      "num_tokens": 7061957.0,
      "step": 151
    },
    {
      "epoch": 1.9240506329113924,
      "grad_norm": 0.28191933035850525,
      "learning_rate": 0.00017071067811865476,
      "loss": 1.506,
      "mean_token_accuracy": 0.6569375395774841,
      "num_tokens": 7107246.0,
      "step": 152
    },
    {
      "epoch": 1.9367088607594938,
      "grad_norm": 0.2704729437828064,
      "learning_rate": 0.00017029501788246924,
      "loss": 1.4832,
      "mean_token_accuracy": 0.6607176065444946,
      "num_tokens": 7155553.0,
      "step": 153
    },
    {
      "epoch": 1.9493670886075949,
      "grad_norm": 0.27189669013023376,
      "learning_rate": 0.00016987694277788417,
      "loss": 1.4983,
      "mean_token_accuracy": 0.6550452709197998,
      "num_tokens": 7202255.0,
      "step": 154
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 0.288601279258728,
      "learning_rate": 0.0001694564671671743,
      "loss": 1.5234,
      "mean_token_accuracy": 0.651522159576416,
      "num_tokens": 7250081.0,
      "step": 155
    },
    {
      "epoch": 1.9746835443037973,
      "grad_norm": 0.28722238540649414,
      "learning_rate": 0.0001690336054950797,
      "loss": 1.4838,
      "mean_token_accuracy": 0.6596331596374512,
      "num_tokens": 7295452.0,
      "step": 156
    },
    {
      "epoch": 1.9873417721518987,
      "grad_norm": 0.28106075525283813,
      "learning_rate": 0.00016860837228830974,
      "loss": 1.4449,
      "mean_token_accuracy": 0.6669970750808716,
      "num_tokens": 7340912.0,
      "step": 157
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.29809167981147766,
      "learning_rate": 0.0001681807821550438,
      "loss": 1.3563,
      "mean_token_accuracy": 0.6853920221328735,
      "num_tokens": 7386725.0,
      "step": 158
    },
    {
      "epoch": 2.0126582278481013,
      "grad_norm": 0.302463173866272,
      "learning_rate": 0.00016775084978442955,
      "loss": 1.2736,
      "mean_token_accuracy": 0.7005270719528198,
      "num_tokens": 7431374.0,
      "step": 159
    },
    {
      "epoch": 2.0253164556962027,
      "grad_norm": 0.3087366819381714,
      "learning_rate": 0.00016731858994607838,
      "loss": 1.2788,
      "mean_token_accuracy": 0.6979596018791199,
      "num_tokens": 7477018.0,
      "step": 160
    },
    {
      "epoch": 2.037974683544304,
      "grad_norm": 0.31179991364479065,
      "learning_rate": 0.00016688401748955802,
      "loss": 1.3065,
      "mean_token_accuracy": 0.6944866180419922,
      "num_tokens": 7525800.0,
      "step": 161
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.3638923466205597,
      "learning_rate": 0.00016644714734388217,
      "loss": 1.2457,
      "mean_token_accuracy": 0.7046037912368774,
      "num_tokens": 7573759.0,
      "step": 162
    },
    {
      "epoch": 2.0632911392405062,
      "grad_norm": 0.3726513385772705,
      "learning_rate": 0.00016600799451699802,
      "loss": 1.215,
      "mean_token_accuracy": 0.7081204652786255,
      "num_tokens": 7617478.0,
      "step": 163
    },
    {
      "epoch": 2.0759493670886076,
      "grad_norm": 0.316274493932724,
      "learning_rate": 0.0001655665740952703,
      "loss": 1.2713,
      "mean_token_accuracy": 0.6999913454055786,
      "num_tokens": 7663624.0,
      "step": 164
    },
    {
      "epoch": 2.088607594936709,
      "grad_norm": 0.31808707118034363,
      "learning_rate": 0.00016512290124296336,
      "loss": 1.2328,
      "mean_token_accuracy": 0.7067911028862,
      "num_tokens": 7708158.0,
      "step": 165
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 0.304767370223999,
      "learning_rate": 0.00016467699120171987,
      "loss": 1.2526,
      "mean_token_accuracy": 0.7050590515136719,
      "num_tokens": 7755306.0,
      "step": 166
    },
    {
      "epoch": 2.1139240506329116,
      "grad_norm": 0.28329142928123474,
      "learning_rate": 0.00016422885929003758,
      "loss": 1.2241,
      "mean_token_accuracy": 0.7089524269104004,
      "num_tokens": 7803692.0,
      "step": 167
    },
    {
      "epoch": 2.1265822784810124,
      "grad_norm": 0.2999579906463623,
      "learning_rate": 0.00016377852090274276,
      "loss": 1.2834,
      "mean_token_accuracy": 0.7007805109024048,
      "num_tokens": 7851547.0,
      "step": 168
    },
    {
      "epoch": 2.1392405063291138,
      "grad_norm": 0.32791537046432495,
      "learning_rate": 0.0001633259915104616,
      "loss": 1.225,
      "mean_token_accuracy": 0.7079063057899475,
      "num_tokens": 7896107.0,
      "step": 169
    },
    {
      "epoch": 2.151898734177215,
      "grad_norm": 0.35149338841438293,
      "learning_rate": 0.0001628712866590885,
      "loss": 1.1993,
      "mean_token_accuracy": 0.7159732580184937,
      "num_tokens": 7938836.0,
      "step": 170
    },
    {
      "epoch": 2.1645569620253164,
      "grad_norm": 0.3519878387451172,
      "learning_rate": 0.00016241442196925223,
      "loss": 1.267,
      "mean_token_accuracy": 0.6996027827262878,
      "num_tokens": 7985475.0,
      "step": 171
    },
    {
      "epoch": 2.1772151898734178,
      "grad_norm": 0.32906627655029297,
      "learning_rate": 0.00016195541313577923,
      "loss": 1.2498,
      "mean_token_accuracy": 0.7009555697441101,
      "num_tokens": 8034933.0,
      "step": 172
    },
    {
      "epoch": 2.189873417721519,
      "grad_norm": 0.322187215089798,
      "learning_rate": 0.00016149427592715432,
      "loss": 1.269,
      "mean_token_accuracy": 0.6975268721580505,
      "num_tokens": 8082912.0,
      "step": 173
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.33761700987815857,
      "learning_rate": 0.00016103102618497922,
      "loss": 1.245,
      "mean_token_accuracy": 0.7072176337242126,
      "num_tokens": 8129099.0,
      "step": 174
    },
    {
      "epoch": 2.2151898734177213,
      "grad_norm": 0.31815439462661743,
      "learning_rate": 0.00016056567982342817,
      "loss": 1.2356,
      "mean_token_accuracy": 0.7027240991592407,
      "num_tokens": 8175931.0,
      "step": 175
    },
    {
      "epoch": 2.2278481012658227,
      "grad_norm": 0.3221277594566345,
      "learning_rate": 0.00016009825282870126,
      "loss": 1.2474,
      "mean_token_accuracy": 0.7036200165748596,
      "num_tokens": 8224227.0,
      "step": 176
    },
    {
      "epoch": 2.240506329113924,
      "grad_norm": 0.32305020093917847,
      "learning_rate": 0.00015962876125847535,
      "loss": 1.2577,
      "mean_token_accuracy": 0.7010858654975891,
      "num_tokens": 8273285.0,
      "step": 177
    },
    {
      "epoch": 2.2531645569620253,
      "grad_norm": 0.35282593965530396,
      "learning_rate": 0.00015915722124135227,
      "loss": 1.2303,
      "mean_token_accuracy": 0.7082141041755676,
      "num_tokens": 8316689.0,
      "step": 178
    },
    {
      "epoch": 2.2658227848101267,
      "grad_norm": 0.3499389886856079,
      "learning_rate": 0.0001586836489763049,
      "loss": 1.275,
      "mean_token_accuracy": 0.6990365386009216,
      "num_tokens": 8365327.0,
      "step": 179
    },
    {
      "epoch": 2.278481012658228,
      "grad_norm": 0.34207987785339355,
      "learning_rate": 0.00015820806073212055,
      "loss": 1.2536,
      "mean_token_accuracy": 0.702049970626831,
      "num_tokens": 8411392.0,
      "step": 180
    },
    {
      "epoch": 2.291139240506329,
      "grad_norm": 0.32787948846817017,
      "learning_rate": 0.0001577304728468422,
      "loss": 1.2539,
      "mean_token_accuracy": 0.7007697820663452,
      "num_tokens": 8460562.0,
      "step": 181
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 0.3327817916870117,
      "learning_rate": 0.0001572509017272072,
      "loss": 1.2562,
      "mean_token_accuracy": 0.7036707401275635,
      "num_tokens": 8506693.0,
      "step": 182
    },
    {
      "epoch": 2.3164556962025316,
      "grad_norm": 0.3491711914539337,
      "learning_rate": 0.00015676936384808354,
      "loss": 1.2257,
      "mean_token_accuracy": 0.7093122005462646,
      "num_tokens": 8549239.0,
      "step": 183
    },
    {
      "epoch": 2.329113924050633,
      "grad_norm": 0.33453595638275146,
      "learning_rate": 0.00015628587575190395,
      "loss": 1.2653,
      "mean_token_accuracy": 0.6975212693214417,
      "num_tokens": 8594850.0,
      "step": 184
    },
    {
      "epoch": 2.3417721518987342,
      "grad_norm": 0.331360399723053,
      "learning_rate": 0.00015580045404809772,
      "loss": 1.283,
      "mean_token_accuracy": 0.6986226439476013,
      "num_tokens": 8642542.0,
      "step": 185
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.3299597501754761,
      "learning_rate": 0.00015531311541251995,
      "loss": 1.2375,
      "mean_token_accuracy": 0.704332709312439,
      "num_tokens": 8690890.0,
      "step": 186
    },
    {
      "epoch": 2.367088607594937,
      "grad_norm": 0.3463394343852997,
      "learning_rate": 0.00015482387658687875,
      "loss": 1.2434,
      "mean_token_accuracy": 0.7066888809204102,
      "num_tokens": 8737195.0,
      "step": 187
    },
    {
      "epoch": 2.379746835443038,
      "grad_norm": 0.3440077304840088,
      "learning_rate": 0.00015433275437816004,
      "loss": 1.2422,
      "mean_token_accuracy": 0.7034814357757568,
      "num_tokens": 8784251.0,
      "step": 188
    },
    {
      "epoch": 2.392405063291139,
      "grad_norm": 0.34045225381851196,
      "learning_rate": 0.00015383976565805035,
      "loss": 1.2669,
      "mean_token_accuracy": 0.6978285312652588,
      "num_tokens": 8831103.0,
      "step": 189
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.34536802768707275,
      "learning_rate": 0.00015334492736235705,
      "loss": 1.2757,
      "mean_token_accuracy": 0.699856698513031,
      "num_tokens": 8878611.0,
      "step": 190
    },
    {
      "epoch": 2.4177215189873418,
      "grad_norm": 0.3353114426136017,
      "learning_rate": 0.00015284825649042655,
      "loss": 1.2864,
      "mean_token_accuracy": 0.6976174712181091,
      "num_tokens": 8927279.0,
      "step": 191
    },
    {
      "epoch": 2.430379746835443,
      "grad_norm": 0.3356870710849762,
      "learning_rate": 0.00015234977010456047,
      "loss": 1.2366,
      "mean_token_accuracy": 0.7043005228042603,
      "num_tokens": 8974198.0,
      "step": 192
    },
    {
      "epoch": 2.4430379746835444,
      "grad_norm": 0.347060889005661,
      "learning_rate": 0.00015184948532942928,
      "loss": 1.2653,
      "mean_token_accuracy": 0.7007454633712769,
      "num_tokens": 9018395.0,
      "step": 193
    },
    {
      "epoch": 2.4556962025316453,
      "grad_norm": 0.33887261152267456,
      "learning_rate": 0.0001513474193514842,
      "loss": 1.2701,
      "mean_token_accuracy": 0.7008969187736511,
      "num_tokens": 9069077.0,
      "step": 194
    },
    {
      "epoch": 2.4683544303797467,
      "grad_norm": 0.348005473613739,
      "learning_rate": 0.0001508435894183667,
      "loss": 1.2134,
      "mean_token_accuracy": 0.711205244064331,
      "num_tokens": 9113754.0,
      "step": 195
    },
    {
      "epoch": 2.481012658227848,
      "grad_norm": 0.3352693021297455,
      "learning_rate": 0.00015033801283831596,
      "loss": 1.2823,
      "mean_token_accuracy": 0.6957764029502869,
      "num_tokens": 9164557.0,
      "step": 196
    },
    {
      "epoch": 2.4936708860759493,
      "grad_norm": 0.33518290519714355,
      "learning_rate": 0.00014983070697957438,
      "loss": 1.2649,
      "mean_token_accuracy": 0.69664067029953,
      "num_tokens": 9211773.0,
      "step": 197
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.33229929208755493,
      "learning_rate": 0.00014932168926979074,
      "loss": 1.2585,
      "mean_token_accuracy": 0.7020613551139832,
      "num_tokens": 9260300.0,
      "step": 198
    },
    {
      "epoch": 2.518987341772152,
      "grad_norm": 0.3351842164993286,
      "learning_rate": 0.00014881097719542173,
      "loss": 1.259,
      "mean_token_accuracy": 0.7015478014945984,
      "num_tokens": 9309789.0,
      "step": 199
    },
    {
      "epoch": 2.5316455696202533,
      "grad_norm": 0.34552860260009766,
      "learning_rate": 0.000148298588301131,
      "loss": 1.2738,
      "mean_token_accuracy": 0.7004240155220032,
      "num_tokens": 9357260.0,
      "step": 200
    },
    {
      "epoch": 2.5443037974683547,
      "grad_norm": 0.3504265546798706,
      "learning_rate": 0.0001477845401891867,
      "loss": 1.2694,
      "mean_token_accuracy": 0.6969807744026184,
      "num_tokens": 9406608.0,
      "step": 201
    },
    {
      "epoch": 2.5569620253164556,
      "grad_norm": 0.35586681962013245,
      "learning_rate": 0.00014726885051885653,
      "loss": 1.2553,
      "mean_token_accuracy": 0.7009937167167664,
      "num_tokens": 9451253.0,
      "step": 202
    },
    {
      "epoch": 2.569620253164557,
      "grad_norm": 0.34470343589782715,
      "learning_rate": 0.00014675153700580124,
      "loss": 1.2595,
      "mean_token_accuracy": 0.7024118304252625,
      "num_tokens": 9497216.0,
      "step": 203
    },
    {
      "epoch": 2.5822784810126582,
      "grad_norm": 0.335822194814682,
      "learning_rate": 0.00014623261742146602,
      "loss": 1.2525,
      "mean_token_accuracy": 0.7015378475189209,
      "num_tokens": 9544033.0,
      "step": 204
    },
    {
      "epoch": 2.5949367088607596,
      "grad_norm": 0.3584577441215515,
      "learning_rate": 0.00014571210959246988,
      "loss": 1.2333,
      "mean_token_accuracy": 0.706198513507843,
      "num_tokens": 9586333.0,
      "step": 205
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 0.34092119336128235,
      "learning_rate": 0.00014519003139999337,
      "loss": 1.2542,
      "mean_token_accuracy": 0.7034091353416443,
      "num_tokens": 9634298.0,
      "step": 206
    },
    {
      "epoch": 2.620253164556962,
      "grad_norm": 0.35801050066947937,
      "learning_rate": 0.0001446664007791644,
      "loss": 1.2988,
      "mean_token_accuracy": 0.6928357481956482,
      "num_tokens": 9682141.0,
      "step": 207
    },
    {
      "epoch": 2.632911392405063,
      "grad_norm": 0.3394937515258789,
      "learning_rate": 0.00014414123571844178,
      "loss": 1.2624,
      "mean_token_accuracy": 0.7046942710876465,
      "num_tokens": 9731222.0,
      "step": 208
    },
    {
      "epoch": 2.6455696202531644,
      "grad_norm": 0.35818547010421753,
      "learning_rate": 0.00014361455425899756,
      "loss": 1.2537,
      "mean_token_accuracy": 0.7017344832420349,
      "num_tokens": 9778851.0,
      "step": 209
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.369809627532959,
      "learning_rate": 0.00014308637449409706,
      "loss": 1.2001,
      "mean_token_accuracy": 0.7117389440536499,
      "num_tokens": 9822556.0,
      "step": 210
    },
    {
      "epoch": 2.670886075949367,
      "grad_norm": 0.3590123653411865,
      "learning_rate": 0.00014255671456847748,
      "loss": 1.2923,
      "mean_token_accuracy": 0.6946203708648682,
      "num_tokens": 9870151.0,
      "step": 211
    },
    {
      "epoch": 2.6835443037974684,
      "grad_norm": 0.3501306176185608,
      "learning_rate": 0.00014202559267772444,
      "loss": 1.2468,
      "mean_token_accuracy": 0.701853334903717,
      "num_tokens": 9916833.0,
      "step": 212
    },
    {
      "epoch": 2.6962025316455698,
      "grad_norm": 0.3504582643508911,
      "learning_rate": 0.00014149302706764697,
      "loss": 1.279,
      "mean_token_accuracy": 0.6964161396026611,
      "num_tokens": 9965727.0,
      "step": 213
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 0.3543940782546997,
      "learning_rate": 0.00014095903603365066,
      "loss": 1.2258,
      "mean_token_accuracy": 0.704728364944458,
      "num_tokens": 10009527.0,
      "step": 214
    },
    {
      "epoch": 2.721518987341772,
      "grad_norm": 0.3418155014514923,
      "learning_rate": 0.0001404236379201091,
      "loss": 1.2177,
      "mean_token_accuracy": 0.7080258727073669,
      "num_tokens": 10055667.0,
      "step": 215
    },
    {
      "epoch": 2.7341772151898733,
      "grad_norm": 0.34693628549575806,
      "learning_rate": 0.00013988685111973384,
      "loss": 1.254,
      "mean_token_accuracy": 0.7021883726119995,
      "num_tokens": 10104077.0,
      "step": 216
    },
    {
      "epoch": 2.7468354430379747,
      "grad_norm": 0.3545077443122864,
      "learning_rate": 0.00013934869407294245,
      "loss": 1.3069,
      "mean_token_accuracy": 0.6918914318084717,
      "num_tokens": 10152374.0,
      "step": 217
    },
    {
      "epoch": 2.759493670886076,
      "grad_norm": 0.3604845702648163,
      "learning_rate": 0.00013880918526722497,
      "loss": 1.2708,
      "mean_token_accuracy": 0.6986167430877686,
      "num_tokens": 10198416.0,
      "step": 218
    },
    {
      "epoch": 2.7721518987341773,
      "grad_norm": 0.34998974204063416,
      "learning_rate": 0.000138268343236509,
      "loss": 1.2194,
      "mean_token_accuracy": 0.7079833745956421,
      "num_tokens": 10243724.0,
      "step": 219
    },
    {
      "epoch": 2.7848101265822782,
      "grad_norm": 0.3521496057510376,
      "learning_rate": 0.0001377261865605227,
      "loss": 1.2717,
      "mean_token_accuracy": 0.6980299949645996,
      "num_tokens": 10290286.0,
      "step": 220
    },
    {
      "epoch": 2.7974683544303796,
      "grad_norm": 0.3567959666252136,
      "learning_rate": 0.0001371827338641568,
      "loss": 1.2435,
      "mean_token_accuracy": 0.7042776942253113,
      "num_tokens": 10335000.0,
      "step": 221
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.3671683371067047,
      "learning_rate": 0.00013663800381682464,
      "loss": 1.2715,
      "mean_token_accuracy": 0.6983711123466492,
      "num_tokens": 10381661.0,
      "step": 222
    },
    {
      "epoch": 2.8227848101265822,
      "grad_norm": 0.3411732614040375,
      "learning_rate": 0.00013609201513182075,
      "loss": 1.2401,
      "mean_token_accuracy": 0.7075390815734863,
      "num_tokens": 10427314.0,
      "step": 223
    },
    {
      "epoch": 2.8354430379746836,
      "grad_norm": 0.3450396656990051,
      "learning_rate": 0.00013554478656567818,
      "loss": 1.2947,
      "mean_token_accuracy": 0.6972159147262573,
      "num_tokens": 10476119.0,
      "step": 224
    },
    {
      "epoch": 2.848101265822785,
      "grad_norm": 0.3609869182109833,
      "learning_rate": 0.0001349963369175239,
      "loss": 1.2503,
      "mean_token_accuracy": 0.7017799615859985,
      "num_tokens": 10521408.0,
      "step": 225
    },
    {
      "epoch": 2.8607594936708862,
      "grad_norm": 0.3500053584575653,
      "learning_rate": 0.0001344466850284333,
      "loss": 1.2872,
      "mean_token_accuracy": 0.6957990527153015,
      "num_tokens": 10569128.0,
      "step": 226
    },
    {
      "epoch": 2.8734177215189876,
      "grad_norm": 0.3592340052127838,
      "learning_rate": 0.00013389584978078258,
      "loss": 1.2613,
      "mean_token_accuracy": 0.7002545595169067,
      "num_tokens": 10615151.0,
      "step": 227
    },
    {
      "epoch": 2.8860759493670884,
      "grad_norm": 0.35870033502578735,
      "learning_rate": 0.00013334385009760032,
      "loss": 1.2487,
      "mean_token_accuracy": 0.7038562893867493,
      "num_tokens": 10659636.0,
      "step": 228
    },
    {
      "epoch": 2.8987341772151898,
      "grad_norm": 0.3581511676311493,
      "learning_rate": 0.00013279070494191737,
      "loss": 1.2662,
      "mean_token_accuracy": 0.7002348303794861,
      "num_tokens": 10703988.0,
      "step": 229
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.3540744185447693,
      "learning_rate": 0.00013223643331611537,
      "loss": 1.2613,
      "mean_token_accuracy": 0.7001772522926331,
      "num_tokens": 10750306.0,
      "step": 230
    },
    {
      "epoch": 2.9240506329113924,
      "grad_norm": 0.37011268734931946,
      "learning_rate": 0.000131681054261274,
      "loss": 1.2902,
      "mean_token_accuracy": 0.697233259677887,
      "num_tokens": 10795550.0,
      "step": 231
    },
    {
      "epoch": 2.9367088607594938,
      "grad_norm": 0.3456755578517914,
      "learning_rate": 0.00013112458685651668,
      "loss": 1.3182,
      "mean_token_accuracy": 0.6913251876831055,
      "num_tokens": 10845252.0,
      "step": 232
    },
    {
      "epoch": 2.9493670886075947,
      "grad_norm": 0.34918084740638733,
      "learning_rate": 0.00013056705021835546,
      "loss": 1.2464,
      "mean_token_accuracy": 0.7025464773178101,
      "num_tokens": 10892204.0,
      "step": 233
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.3588140606880188,
      "learning_rate": 0.0001300084635000341,
      "loss": 1.3022,
      "mean_token_accuracy": 0.6909407377243042,
      "num_tokens": 10938188.0,
      "step": 234
    },
    {
      "epoch": 2.9746835443037973,
      "grad_norm": 0.35461536049842834,
      "learning_rate": 0.00012944884589086993,
      "loss": 1.3174,
      "mean_token_accuracy": 0.6889122128486633,
      "num_tokens": 10986801.0,
      "step": 235
    },
    {
      "epoch": 2.9873417721518987,
      "grad_norm": 0.3585621416568756,
      "learning_rate": 0.00012888821661559508,
      "loss": 1.2524,
      "mean_token_accuracy": 0.7034770846366882,
      "num_tokens": 11033772.0,
      "step": 236
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.3551687002182007,
      "learning_rate": 0.00012832659493369557,
      "loss": 1.0758,
      "mean_token_accuracy": 0.739671528339386,
      "num_tokens": 11079559.0,
      "step": 237
    },
    {
      "epoch": 3.0126582278481013,
      "grad_norm": 0.4075370728969574,
      "learning_rate": 0.00012776400013875006,
      "loss": 1.032,
      "mean_token_accuracy": 0.7526733875274658,
      "num_tokens": 11127596.0,
      "step": 238
    },
    {
      "epoch": 3.0253164556962027,
      "grad_norm": 0.3849814832210541,
      "learning_rate": 0.0001272004515577668,
      "loss": 1.004,
      "mean_token_accuracy": 0.7555179595947266,
      "num_tokens": 11172559.0,
      "step": 239
    },
    {
      "epoch": 3.037974683544304,
      "grad_norm": 0.4353029131889343,
      "learning_rate": 0.0001266359685505198,
      "loss": 0.99,
      "mean_token_accuracy": 0.7563720345497131,
      "num_tokens": 11218566.0,
      "step": 240
    },
    {
      "epoch": 3.050632911392405,
      "grad_norm": 0.600694477558136,
      "learning_rate": 0.0001260705705088837,
      "loss": 0.9818,
      "mean_token_accuracy": 0.7581696510314941,
      "num_tokens": 11268632.0,
      "step": 241
    },
    {
      "epoch": 3.0632911392405062,
      "grad_norm": 0.5432077050209045,
      "learning_rate": 0.00012550427685616765,
      "loss": 1.0269,
      "mean_token_accuracy": 0.7526742219924927,
      "num_tokens": 11315719.0,
      "step": 242
    },
    {
      "epoch": 3.0759493670886076,
      "grad_norm": 0.4427173435688019,
      "learning_rate": 0.00012493710704644805,
      "loss": 0.9989,
      "mean_token_accuracy": 0.75676429271698,
      "num_tokens": 11362573.0,
      "step": 243
    },
    {
      "epoch": 3.088607594936709,
      "grad_norm": 0.3981163501739502,
      "learning_rate": 0.0001243690805639002,
      "loss": 0.9723,
      "mean_token_accuracy": 0.7642591595649719,
      "num_tokens": 11408853.0,
      "step": 244
    },
    {
      "epoch": 3.1012658227848102,
      "grad_norm": 0.3880089223384857,
      "learning_rate": 0.00012380021692212894,
      "loss": 0.9801,
      "mean_token_accuracy": 0.7625755667686462,
      "num_tokens": 11455555.0,
      "step": 245
    },
    {
      "epoch": 3.1139240506329116,
      "grad_norm": 0.38175487518310547,
      "learning_rate": 0.00012323053566349834,
      "loss": 0.9998,
      "mean_token_accuracy": 0.7584809064865112,
      "num_tokens": 11503963.0,
      "step": 246
    },
    {
      "epoch": 3.1265822784810124,
      "grad_norm": 0.39173728227615356,
      "learning_rate": 0.00012266005635846037,
      "loss": 0.9341,
      "mean_token_accuracy": 0.7685783505439758,
      "num_tokens": 11549537.0,
      "step": 247
    },
    {
      "epoch": 3.1392405063291138,
      "grad_norm": 0.435289591550827,
      "learning_rate": 0.0001220887986048825,
      "loss": 0.948,
      "mean_token_accuracy": 0.7657555341720581,
      "num_tokens": 11595442.0,
      "step": 248
    },
    {
      "epoch": 3.151898734177215,
      "grad_norm": 0.480478972196579,
      "learning_rate": 0.00012151678202737456,
      "loss": 0.9671,
      "mean_token_accuracy": 0.7640275955200195,
      "num_tokens": 11638867.0,
      "step": 249
    },
    {
      "epoch": 3.1645569620253164,
      "grad_norm": 0.5007439851760864,
      "learning_rate": 0.00012094402627661447,
      "loss": 0.9631,
      "mean_token_accuracy": 0.7621972560882568,
      "num_tokens": 11684515.0,
      "step": 250
    },
    {
      "epoch": 3.1772151898734178,
      "grad_norm": 0.4938323497772217,
      "learning_rate": 0.00012037055102867321,
      "loss": 0.9863,
      "mean_token_accuracy": 0.7565716505050659,
      "num_tokens": 11731295.0,
      "step": 251
    },
    {
      "epoch": 3.189873417721519,
      "grad_norm": 0.46317818760871887,
      "learning_rate": 0.00011979637598433899,
      "loss": 0.9599,
      "mean_token_accuracy": 0.7661996483802795,
      "num_tokens": 11775897.0,
      "step": 252
    },
    {
      "epoch": 3.2025316455696204,
      "grad_norm": 0.4412788152694702,
      "learning_rate": 0.00011922152086844023,
      "loss": 0.9711,
      "mean_token_accuracy": 0.7613348364830017,
      "num_tokens": 11821749.0,
      "step": 253
    },
    {
      "epoch": 3.2151898734177213,
      "grad_norm": 0.4221232235431671,
      "learning_rate": 0.00011864600542916813,
      "loss": 0.9382,
      "mean_token_accuracy": 0.7676687240600586,
      "num_tokens": 11868449.0,
      "step": 254
    },
    {
      "epoch": 3.2278481012658227,
      "grad_norm": 0.4161496162414551,
      "learning_rate": 0.00011806984943739821,
      "loss": 0.9391,
      "mean_token_accuracy": 0.7664811015129089,
      "num_tokens": 11915036.0,
      "step": 255
    },
    {
      "epoch": 3.240506329113924,
      "grad_norm": 0.40381163358688354,
      "learning_rate": 0.00011749307268601111,
      "loss": 1.0414,
      "mean_token_accuracy": 0.7476540207862854,
      "num_tokens": 11965079.0,
      "step": 256
    },
    {
      "epoch": 3.2531645569620253,
      "grad_norm": 0.41820257902145386,
      "learning_rate": 0.00011691569498921264,
      "loss": 0.9629,
      "mean_token_accuracy": 0.7615616917610168,
      "num_tokens": 12011784.0,
      "step": 257
    },
    {
      "epoch": 3.2658227848101267,
      "grad_norm": 0.44741806387901306,
      "learning_rate": 0.00011633773618185302,
      "loss": 0.977,
      "mean_token_accuracy": 0.7589461207389832,
      "num_tokens": 12058153.0,
      "step": 258
    },
    {
      "epoch": 3.278481012658228,
      "grad_norm": 0.4427592158317566,
      "learning_rate": 0.00011575921611874565,
      "loss": 0.9571,
      "mean_token_accuracy": 0.7660679221153259,
      "num_tokens": 12104427.0,
      "step": 259
    },
    {
      "epoch": 3.291139240506329,
      "grad_norm": 0.4691466987133026,
      "learning_rate": 0.00011518015467398489,
      "loss": 0.9336,
      "mean_token_accuracy": 0.7664586901664734,
      "num_tokens": 12149528.0,
      "step": 260
    },
    {
      "epoch": 3.3037974683544302,
      "grad_norm": 0.47362270951271057,
      "learning_rate": 0.00011460057174026335,
      "loss": 0.9858,
      "mean_token_accuracy": 0.7595200538635254,
      "num_tokens": 12196598.0,
      "step": 261
    },
    {
      "epoch": 3.3164556962025316,
      "grad_norm": 0.46093323826789856,
      "learning_rate": 0.0001140204872281886,
      "loss": 0.9882,
      "mean_token_accuracy": 0.7542067766189575,
      "num_tokens": 12245036.0,
      "step": 262
    },
    {
      "epoch": 3.329113924050633,
      "grad_norm": 0.43514880537986755,
      "learning_rate": 0.00011343992106559898,
      "loss": 1.0037,
      "mean_token_accuracy": 0.7565442323684692,
      "num_tokens": 12294189.0,
      "step": 263
    },
    {
      "epoch": 3.3417721518987342,
      "grad_norm": 0.4213711619377136,
      "learning_rate": 0.00011285889319687923,
      "loss": 0.9712,
      "mean_token_accuracy": 0.7629369497299194,
      "num_tokens": 12343569.0,
      "step": 264
    },
    {
      "epoch": 3.3544303797468356,
      "grad_norm": 0.42522549629211426,
      "learning_rate": 0.00011227742358227522,
      "loss": 0.9892,
      "mean_token_accuracy": 0.7599421143531799,
      "num_tokens": 12392717.0,
      "step": 265
    },
    {
      "epoch": 3.367088607594937,
      "grad_norm": 0.43775442242622375,
      "learning_rate": 0.00011169553219720828,
      "loss": 0.9598,
      "mean_token_accuracy": 0.7645099759101868,
      "num_tokens": 12437216.0,
      "step": 266
    },
    {
      "epoch": 3.379746835443038,
      "grad_norm": 0.45032966136932373,
      "learning_rate": 0.00011111323903158885,
      "loss": 0.988,
      "mean_token_accuracy": 0.7559954524040222,
      "num_tokens": 12484566.0,
      "step": 267
    },
    {
      "epoch": 3.392405063291139,
      "grad_norm": 0.4129181206226349,
      "learning_rate": 0.00011053056408913001,
      "loss": 0.9857,
      "mean_token_accuracy": 0.758560299873352,
      "num_tokens": 12534891.0,
      "step": 268
    },
    {
      "epoch": 3.4050632911392404,
      "grad_norm": 0.4736068546772003,
      "learning_rate": 0.0001099475273866601,
      "loss": 0.9717,
      "mean_token_accuracy": 0.7591570615768433,
      "num_tokens": 12579893.0,
      "step": 269
    },
    {
      "epoch": 3.4177215189873418,
      "grad_norm": 0.4686090350151062,
      "learning_rate": 0.0001093641489534351,
      "loss": 0.9763,
      "mean_token_accuracy": 0.7623862028121948,
      "num_tokens": 12625552.0,
      "step": 270
    },
    {
      "epoch": 3.430379746835443,
      "grad_norm": 0.44861164689064026,
      "learning_rate": 0.0001087804488304506,
      "loss": 0.9643,
      "mean_token_accuracy": 0.7631211876869202,
      "num_tokens": 12671496.0,
      "step": 271
    },
    {
      "epoch": 3.4430379746835444,
      "grad_norm": 0.45744118094444275,
      "learning_rate": 0.00010819644706975332,
      "loss": 0.9521,
      "mean_token_accuracy": 0.7642855644226074,
      "num_tokens": 12716763.0,
      "step": 272
    },
    {
      "epoch": 3.4556962025316453,
      "grad_norm": 0.48478421568870544,
      "learning_rate": 0.00010761216373375221,
      "loss": 0.9767,
      "mean_token_accuracy": 0.7615178823471069,
      "num_tokens": 12760151.0,
      "step": 273
    },
    {
      "epoch": 3.4683544303797467,
      "grad_norm": 0.4677796959877014,
      "learning_rate": 0.0001070276188945293,
      "loss": 0.9978,
      "mean_token_accuracy": 0.7565178871154785,
      "num_tokens": 12805130.0,
      "step": 274
    },
    {
      "epoch": 3.481012658227848,
      "grad_norm": 0.4537658095359802,
      "learning_rate": 0.00010644283263315014,
      "loss": 0.9788,
      "mean_token_accuracy": 0.7601967453956604,
      "num_tokens": 12851949.0,
      "step": 275
    },
    {
      "epoch": 3.4936708860759493,
      "grad_norm": 0.4506645202636719,
      "learning_rate": 0.00010585782503897388,
      "loss": 0.9727,
      "mean_token_accuracy": 0.7614617347717285,
      "num_tokens": 12898123.0,
      "step": 276
    },
    {
      "epoch": 3.5063291139240507,
      "grad_norm": 0.42953184247016907,
      "learning_rate": 0.00010527261620896323,
      "loss": 0.9731,
      "mean_token_accuracy": 0.7647458910942078,
      "num_tokens": 12945268.0,
      "step": 277
    },
    {
      "epoch": 3.518987341772152,
      "grad_norm": 0.4598475396633148,
      "learning_rate": 0.00010468722624699401,
      "loss": 1.0024,
      "mean_token_accuracy": 0.7568365931510925,
      "num_tokens": 12991371.0,
      "step": 278
    },
    {
      "epoch": 3.5316455696202533,
      "grad_norm": 0.44531142711639404,
      "learning_rate": 0.00010410167526316457,
      "loss": 0.9627,
      "mean_token_accuracy": 0.7636065483093262,
      "num_tokens": 13039004.0,
      "step": 279
    },
    {
      "epoch": 3.5443037974683547,
      "grad_norm": 0.4507937729358673,
      "learning_rate": 0.00010351598337310482,
      "loss": 0.9912,
      "mean_token_accuracy": 0.7541834115982056,
      "num_tokens": 13085920.0,
      "step": 280
    },
    {
      "epoch": 3.5569620253164556,
      "grad_norm": 0.4690151512622833,
      "learning_rate": 0.00010293017069728535,
      "loss": 1.0278,
      "mean_token_accuracy": 0.7486876845359802,
      "num_tokens": 13133610.0,
      "step": 281
    },
    {
      "epoch": 3.569620253164557,
      "grad_norm": 0.4618782103061676,
      "learning_rate": 0.00010234425736032607,
      "loss": 0.9648,
      "mean_token_accuracy": 0.7642068862915039,
      "num_tokens": 13178353.0,
      "step": 282
    },
    {
      "epoch": 3.5822784810126582,
      "grad_norm": 0.45085033774375916,
      "learning_rate": 0.00010175826349030496,
      "loss": 1.0038,
      "mean_token_accuracy": 0.7567628026008606,
      "num_tokens": 13227509.0,
      "step": 283
    },
    {
      "epoch": 3.5949367088607596,
      "grad_norm": 0.44593507051467896,
      "learning_rate": 0.00010117220921806664,
      "loss": 0.9988,
      "mean_token_accuracy": 0.752589225769043,
      "num_tokens": 13276140.0,
      "step": 284
    },
    {
      "epoch": 3.607594936708861,
      "grad_norm": 0.448966383934021,
      "learning_rate": 0.00010058611467653066,
      "loss": 0.9782,
      "mean_token_accuracy": 0.7603657841682434,
      "num_tokens": 13322896.0,
      "step": 285
    },
    {
      "epoch": 3.620253164556962,
      "grad_norm": 0.43764910101890564,
      "learning_rate": 0.0001,
      "loss": 0.9499,
      "mean_token_accuracy": 0.7665926218032837,
      "num_tokens": 13369788.0,
      "step": 286
    },
    {
      "epoch": 3.632911392405063,
      "grad_norm": 0.46153128147125244,
      "learning_rate": 9.941388532346934e-05,
      "loss": 0.9586,
      "mean_token_accuracy": 0.7638452053070068,
      "num_tokens": 13414073.0,
      "step": 287
    },
    {
      "epoch": 3.6455696202531644,
      "grad_norm": 0.44592994451522827,
      "learning_rate": 9.882779078193339e-05,
      "loss": 0.9965,
      "mean_token_accuracy": 0.7588433027267456,
      "num_tokens": 13462931.0,
      "step": 288
    },
    {
      "epoch": 3.6582278481012658,
      "grad_norm": 0.4681696593761444,
      "learning_rate": 9.824173650969507e-05,
      "loss": 0.9741,
      "mean_token_accuracy": 0.7605555057525635,
      "num_tokens": 13509369.0,
      "step": 289
    },
    {
      "epoch": 3.670886075949367,
      "grad_norm": 0.46631893515586853,
      "learning_rate": 9.765574263967396e-05,
      "loss": 0.9998,
      "mean_token_accuracy": 0.7567450404167175,
      "num_tokens": 13553206.0,
      "step": 290
    },
    {
      "epoch": 3.6835443037974684,
      "grad_norm": 0.44251787662506104,
      "learning_rate": 9.706982930271465e-05,
      "loss": 0.9894,
      "mean_token_accuracy": 0.7563201785087585,
      "num_tokens": 13601370.0,
      "step": 291
    },
    {
      "epoch": 3.6962025316455698,
      "grad_norm": 0.4594095051288605,
      "learning_rate": 9.648401662689521e-05,
      "loss": 0.9922,
      "mean_token_accuracy": 0.7575803995132446,
      "num_tokens": 13647210.0,
      "step": 292
    },
    {
      "epoch": 3.708860759493671,
      "grad_norm": 0.4474162757396698,
      "learning_rate": 9.589832473683547e-05,
      "loss": 0.9915,
      "mean_token_accuracy": 0.7562800049781799,
      "num_tokens": 13694726.0,
      "step": 293
    },
    {
      "epoch": 3.721518987341772,
      "grad_norm": 0.4731695353984833,
      "learning_rate": 9.531277375300599e-05,
      "loss": 0.9804,
      "mean_token_accuracy": 0.758215606212616,
      "num_tokens": 13737605.0,
      "step": 294
    },
    {
      "epoch": 3.7341772151898733,
      "grad_norm": 0.4519980847835541,
      "learning_rate": 9.472738379103682e-05,
      "loss": 1.004,
      "mean_token_accuracy": 0.7548936605453491,
      "num_tokens": 13784873.0,
      "step": 295
    },
    {
      "epoch": 3.7468354430379747,
      "grad_norm": 0.45873039960861206,
      "learning_rate": 9.414217496102614e-05,
      "loss": 0.9526,
      "mean_token_accuracy": 0.7641096711158752,
      "num_tokens": 13831802.0,
      "step": 296
    },
    {
      "epoch": 3.759493670886076,
      "grad_norm": 0.46528834104537964,
      "learning_rate": 9.355716736684988e-05,
      "loss": 0.9592,
      "mean_token_accuracy": 0.7640624642372131,
      "num_tokens": 13878035.0,
      "step": 297
    },
    {
      "epoch": 3.7721518987341773,
      "grad_norm": 0.45268163084983826,
      "learning_rate": 9.297238110547074e-05,
      "loss": 1.0181,
      "mean_token_accuracy": 0.7509171366691589,
      "num_tokens": 13927436.0,
      "step": 298
    },
    {
      "epoch": 3.7848101265822782,
      "grad_norm": 0.4594506323337555,
      "learning_rate": 9.238783626624781e-05,
      "loss": 0.9415,
      "mean_token_accuracy": 0.7657912373542786,
      "num_tokens": 13972430.0,
      "step": 299
    },
    {
      "epoch": 3.7974683544303796,
      "grad_norm": 0.46437937021255493,
      "learning_rate": 9.180355293024669e-05,
      "loss": 0.9485,
      "mean_token_accuracy": 0.766374945640564,
      "num_tokens": 14019227.0,
      "step": 300
    },
    {
      "epoch": 3.810126582278481,
      "grad_norm": 0.4636267125606537,
      "learning_rate": 9.121955116954942e-05,
      "loss": 0.9802,
      "mean_token_accuracy": 0.7597395181655884,
      "num_tokens": 14064750.0,
      "step": 301
    },
    {
      "epoch": 3.8227848101265822,
      "grad_norm": 0.48009157180786133,
      "learning_rate": 9.063585104656493e-05,
      "loss": 0.9248,
      "mean_token_accuracy": 0.7727100253105164,
      "num_tokens": 14109629.0,
      "step": 302
    },
    {
      "epoch": 3.8354430379746836,
      "grad_norm": 0.46061643958091736,
      "learning_rate": 9.005247261333993e-05,
      "loss": 0.9856,
      "mean_token_accuracy": 0.7584598660469055,
      "num_tokens": 14157714.0,
      "step": 303
    },
    {
      "epoch": 3.848101265822785,
      "grad_norm": 0.46670177578926086,
      "learning_rate": 8.946943591087e-05,
      "loss": 1.0039,
      "mean_token_accuracy": 0.7559428811073303,
      "num_tokens": 14202874.0,
      "step": 304
    },
    {
      "epoch": 3.8607594936708862,
      "grad_norm": 0.4689050018787384,
      "learning_rate": 8.88867609684112e-05,
      "loss": 0.9811,
      "mean_token_accuracy": 0.7581244707107544,
      "num_tokens": 14247064.0,
      "step": 305
    },
    {
      "epoch": 3.8734177215189876,
      "grad_norm": 0.44281500577926636,
      "learning_rate": 8.830446780279176e-05,
      "loss": 0.9928,
      "mean_token_accuracy": 0.7555981874465942,
      "num_tokens": 14296653.0,
      "step": 306
    },
    {
      "epoch": 3.8860759493670884,
      "grad_norm": 0.45278170704841614,
      "learning_rate": 8.772257641772479e-05,
      "loss": 0.9772,
      "mean_token_accuracy": 0.7612497806549072,
      "num_tokens": 14342518.0,
      "step": 307
    },
    {
      "epoch": 3.8987341772151898,
      "grad_norm": 0.4528560936450958,
      "learning_rate": 8.71411068031208e-05,
      "loss": 0.9495,
      "mean_token_accuracy": 0.7653767466545105,
      "num_tokens": 14389406.0,
      "step": 308
    },
    {
      "epoch": 3.911392405063291,
      "grad_norm": 0.4473622143268585,
      "learning_rate": 8.656007893440106e-05,
      "loss": 0.9523,
      "mean_token_accuracy": 0.7669484615325928,
      "num_tokens": 14437734.0,
      "step": 309
    },
    {
      "epoch": 3.9240506329113924,
      "grad_norm": 0.4741891324520111,
      "learning_rate": 8.597951277181142e-05,
      "loss": 1.003,
      "mean_token_accuracy": 0.7522197961807251,
      "num_tokens": 14485437.0,
      "step": 310
    },
    {
      "epoch": 3.9367088607594938,
      "grad_norm": 0.4523335099220276,
      "learning_rate": 8.539942825973666e-05,
      "loss": 0.9808,
      "mean_token_accuracy": 0.7603780031204224,
      "num_tokens": 14534282.0,
      "step": 311
    },
    {
      "epoch": 3.9493670886075947,
      "grad_norm": 0.469784677028656,
      "learning_rate": 8.481984532601515e-05,
      "loss": 0.9904,
      "mean_token_accuracy": 0.7573358416557312,
      "num_tokens": 14580319.0,
      "step": 312
    },
    {
      "epoch": 3.962025316455696,
      "grad_norm": 0.47262197732925415,
      "learning_rate": 8.424078388125436e-05,
      "loss": 0.9906,
      "mean_token_accuracy": 0.7568280696868896,
      "num_tokens": 14628896.0,
      "step": 313
    },
    {
      "epoch": 3.9746835443037973,
      "grad_norm": 0.47215375304222107,
      "learning_rate": 8.366226381814697e-05,
      "loss": 1.0217,
      "mean_token_accuracy": 0.7515872716903687,
      "num_tokens": 14677943.0,
      "step": 314
    },
    {
      "epoch": 3.9873417721518987,
      "grad_norm": 0.4638935625553131,
      "learning_rate": 8.308430501078739e-05,
      "loss": 1.0244,
      "mean_token_accuracy": 0.7499192953109741,
      "num_tokens": 14727559.0,
      "step": 315
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.43590956926345825,
      "learning_rate": 8.25069273139889e-05,
      "loss": 0.8203,
      "mean_token_accuracy": 0.7978315949440002,
      "num_tokens": 14774109.0,
      "step": 316
    },
    {
      "epoch": 4.012658227848101,
      "grad_norm": 0.4966140389442444,
      "learning_rate": 8.19301505626018e-05,
      "loss": 0.7698,
      "mean_token_accuracy": 0.8127477765083313,
      "num_tokens": 14821852.0,
      "step": 317
    },
    {
      "epoch": 4.025316455696203,
      "grad_norm": 0.48176905512809753,
      "learning_rate": 8.13539945708319e-05,
      "loss": 0.7578,
      "mean_token_accuracy": 0.8137043118476868,
      "num_tokens": 14867274.0,
      "step": 318
    },
    {
      "epoch": 4.037974683544304,
      "grad_norm": 0.49787062406539917,
      "learning_rate": 8.07784791315598e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.8246570229530334,
      "num_tokens": 14913037.0,
      "step": 319
    },
    {
      "epoch": 4.050632911392405,
      "grad_norm": 0.6012403964996338,
      "learning_rate": 8.020362401566103e-05,
      "loss": 0.7109,
      "mean_token_accuracy": 0.8215330839157104,
      "num_tokens": 14959569.0,
      "step": 320
    },
    {
      "epoch": 4.063291139240507,
      "grad_norm": 0.7662049531936646,
      "learning_rate": 7.962944897132678e-05,
      "loss": 0.7184,
      "mean_token_accuracy": 0.8202931880950928,
      "num_tokens": 15006565.0,
      "step": 321
    },
    {
      "epoch": 4.075949367088608,
      "grad_norm": 0.707722008228302,
      "learning_rate": 7.905597372338558e-05,
      "loss": 0.7138,
      "mean_token_accuracy": 0.8212239146232605,
      "num_tokens": 15053218.0,
      "step": 322
    },
    {
      "epoch": 4.0886075949367084,
      "grad_norm": 0.5710378289222717,
      "learning_rate": 7.848321797262548e-05,
      "loss": 0.6948,
      "mean_token_accuracy": 0.8260384202003479,
      "num_tokens": 15101626.0,
      "step": 323
    },
    {
      "epoch": 4.10126582278481,
      "grad_norm": 0.536343514919281,
      "learning_rate": 7.791120139511752e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.8271479606628418,
      "num_tokens": 15147012.0,
      "step": 324
    },
    {
      "epoch": 4.113924050632911,
      "grad_norm": 0.5122034549713135,
      "learning_rate": 7.733994364153969e-05,
      "loss": 0.7082,
      "mean_token_accuracy": 0.8272866010665894,
      "num_tokens": 15192996.0,
      "step": 325
    },
    {
      "epoch": 4.1265822784810124,
      "grad_norm": 0.49738800525665283,
      "learning_rate": 7.67694643365017e-05,
      "loss": 0.7172,
      "mean_token_accuracy": 0.8232470750808716,
      "num_tokens": 15238270.0,
      "step": 326
    },
    {
      "epoch": 4.139240506329114,
      "grad_norm": 0.5137494206428528,
      "learning_rate": 7.619978307787108e-05,
      "loss": 0.6974,
      "mean_token_accuracy": 0.8292660713195801,
      "num_tokens": 15281770.0,
      "step": 327
    },
    {
      "epoch": 4.151898734177215,
      "grad_norm": 0.5142786502838135,
      "learning_rate": 7.563091943609984e-05,
      "loss": 0.7247,
      "mean_token_accuracy": 0.8209677934646606,
      "num_tokens": 15328105.0,
      "step": 328
    },
    {
      "epoch": 4.1645569620253164,
      "grad_norm": 0.5584180951118469,
      "learning_rate": 7.506289295355198e-05,
      "loss": 0.709,
      "mean_token_accuracy": 0.8250913619995117,
      "num_tokens": 15374679.0,
      "step": 329
    },
    {
      "epoch": 4.177215189873418,
      "grad_norm": 0.5827005505561829,
      "learning_rate": 7.449572314383237e-05,
      "loss": 0.7321,
      "mean_token_accuracy": 0.8182151317596436,
      "num_tokens": 15421166.0,
      "step": 330
    },
    {
      "epoch": 4.189873417721519,
      "grad_norm": 0.6163628101348877,
      "learning_rate": 7.392942949111633e-05,
      "loss": 0.7458,
      "mean_token_accuracy": 0.8168110847473145,
      "num_tokens": 15469246.0,
      "step": 331
    },
    {
      "epoch": 4.2025316455696204,
      "grad_norm": 0.5932377576828003,
      "learning_rate": 7.336403144948022e-05,
      "loss": 0.7449,
      "mean_token_accuracy": 0.8152413368225098,
      "num_tokens": 15516117.0,
      "step": 332
    },
    {
      "epoch": 4.215189873417722,
      "grad_norm": 0.5681145787239075,
      "learning_rate": 7.279954844223323e-05,
      "loss": 0.7052,
      "mean_token_accuracy": 0.8263698816299438,
      "num_tokens": 15562590.0,
      "step": 333
    },
    {
      "epoch": 4.227848101265823,
      "grad_norm": 0.5275646448135376,
      "learning_rate": 7.223599986124994e-05,
      "loss": 0.7144,
      "mean_token_accuracy": 0.8233181238174438,
      "num_tokens": 15611380.0,
      "step": 334
    },
    {
      "epoch": 4.2405063291139244,
      "grad_norm": 0.549479067325592,
      "learning_rate": 7.167340506630445e-05,
      "loss": 0.6902,
      "mean_token_accuracy": 0.8262238502502441,
      "num_tokens": 15656853.0,
      "step": 335
    },
    {
      "epoch": 4.253164556962025,
      "grad_norm": 0.5401619672775269,
      "learning_rate": 7.111178338440496e-05,
      "loss": 0.6795,
      "mean_token_accuracy": 0.8296888470649719,
      "num_tokens": 15702968.0,
      "step": 336
    },
    {
      "epoch": 4.265822784810126,
      "grad_norm": 0.5113623738288879,
      "learning_rate": 7.055115410913009e-05,
      "loss": 0.6953,
      "mean_token_accuracy": 0.8251818418502808,
      "num_tokens": 15748811.0,
      "step": 337
    },
    {
      "epoch": 4.2784810126582276,
      "grad_norm": 0.5283024311065674,
      "learning_rate": 6.999153649996595e-05,
      "loss": 0.7128,
      "mean_token_accuracy": 0.8234080672264099,
      "num_tokens": 15796380.0,
      "step": 338
    },
    {
      "epoch": 4.291139240506329,
      "grad_norm": 0.5316226482391357,
      "learning_rate": 6.943294978164454e-05,
      "loss": 0.6752,
      "mean_token_accuracy": 0.8321078419685364,
      "num_tokens": 15842956.0,
      "step": 339
    },
    {
      "epoch": 4.30379746835443,
      "grad_norm": 0.5778771638870239,
      "learning_rate": 6.887541314348333e-05,
      "loss": 0.7031,
      "mean_token_accuracy": 0.8234809637069702,
      "num_tokens": 15889134.0,
      "step": 340
    },
    {
      "epoch": 4.3164556962025316,
      "grad_norm": 0.5819162130355835,
      "learning_rate": 6.831894573872601e-05,
      "loss": 0.7429,
      "mean_token_accuracy": 0.8154833912849426,
      "num_tokens": 15936229.0,
      "step": 341
    },
    {
      "epoch": 4.329113924050633,
      "grad_norm": 0.5713794827461243,
      "learning_rate": 6.776356668388464e-05,
      "loss": 0.7124,
      "mean_token_accuracy": 0.8215630650520325,
      "num_tokens": 15981973.0,
      "step": 342
    },
    {
      "epoch": 4.341772151898734,
      "grad_norm": 0.5627953410148621,
      "learning_rate": 6.720929505808265e-05,
      "loss": 0.7298,
      "mean_token_accuracy": 0.8181149959564209,
      "num_tokens": 16031035.0,
      "step": 343
    },
    {
      "epoch": 4.3544303797468356,
      "grad_norm": 0.5609976053237915,
      "learning_rate": 6.665614990239969e-05,
      "loss": 0.6939,
      "mean_token_accuracy": 0.828144371509552,
      "num_tokens": 16077929.0,
      "step": 344
    },
    {
      "epoch": 4.367088607594937,
      "grad_norm": 0.5497515797615051,
      "learning_rate": 6.610415021921747e-05,
      "loss": 0.7098,
      "mean_token_accuracy": 0.8211853504180908,
      "num_tokens": 16124544.0,
      "step": 345
    },
    {
      "epoch": 4.379746835443038,
      "grad_norm": 0.5421944260597229,
      "learning_rate": 6.555331497156672e-05,
      "loss": 0.6964,
      "mean_token_accuracy": 0.8250083923339844,
      "num_tokens": 16172376.0,
      "step": 346
    },
    {
      "epoch": 4.3924050632911396,
      "grad_norm": 0.535250723361969,
      "learning_rate": 6.50036630824761e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.8260812163352966,
      "num_tokens": 16217829.0,
      "step": 347
    },
    {
      "epoch": 4.405063291139241,
      "grad_norm": 0.5412235260009766,
      "learning_rate": 6.445521343432188e-05,
      "loss": 0.7066,
      "mean_token_accuracy": 0.8224402666091919,
      "num_tokens": 16263962.0,
      "step": 348
    },
    {
      "epoch": 4.417721518987342,
      "grad_norm": 0.5575072169303894,
      "learning_rate": 6.390798486817929e-05,
      "loss": 0.705,
      "mean_token_accuracy": 0.8242610096931458,
      "num_tokens": 16308410.0,
      "step": 349
    },
    {
      "epoch": 4.430379746835443,
      "grad_norm": 0.5557534098625183,
      "learning_rate": 6.336199618317537e-05,
      "loss": 0.7195,
      "mean_token_accuracy": 0.8202704787254333,
      "num_tokens": 16355133.0,
      "step": 350
    },
    {
      "epoch": 4.443037974683544,
      "grad_norm": 0.5342419743537903,
      "learning_rate": 6.281726613584321e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.8249329924583435,
      "num_tokens": 16401842.0,
      "step": 351
    },
    {
      "epoch": 4.455696202531645,
      "grad_norm": 0.5405949950218201,
      "learning_rate": 6.227381343947733e-05,
      "loss": 0.6922,
      "mean_token_accuracy": 0.8285226821899414,
      "num_tokens": 16449370.0,
      "step": 352
    },
    {
      "epoch": 4.468354430379747,
      "grad_norm": 0.552777886390686,
      "learning_rate": 6.173165676349103e-05,
      "loss": 0.7142,
      "mean_token_accuracy": 0.8212454915046692,
      "num_tokens": 16497416.0,
      "step": 353
    },
    {
      "epoch": 4.481012658227848,
      "grad_norm": 0.5529714226722717,
      "learning_rate": 6.119081473277501e-05,
      "loss": 0.6991,
      "mean_token_accuracy": 0.8240037560462952,
      "num_tokens": 16544231.0,
      "step": 354
    },
    {
      "epoch": 4.493670886075949,
      "grad_norm": 0.5485042929649353,
      "learning_rate": 6.065130592705759e-05,
      "loss": 0.7307,
      "mean_token_accuracy": 0.8205800652503967,
      "num_tokens": 16592701.0,
      "step": 355
    },
    {
      "epoch": 4.506329113924051,
      "grad_norm": 0.5457609295845032,
      "learning_rate": 6.01131488802662e-05,
      "loss": 0.7112,
      "mean_token_accuracy": 0.8206736445426941,
      "num_tokens": 16639646.0,
      "step": 356
    },
    {
      "epoch": 4.518987341772152,
      "grad_norm": 0.5910729765892029,
      "learning_rate": 5.9576362079890925e-05,
      "loss": 0.7012,
      "mean_token_accuracy": 0.8259279727935791,
      "num_tokens": 16685375.0,
      "step": 357
    },
    {
      "epoch": 4.531645569620253,
      "grad_norm": 0.5504463911056519,
      "learning_rate": 5.904096396634935e-05,
      "loss": 0.7073,
      "mean_token_accuracy": 0.8246215581893921,
      "num_tokens": 16733267.0,
      "step": 358
    },
    {
      "epoch": 4.544303797468355,
      "grad_norm": 0.5555917024612427,
      "learning_rate": 5.8506972932353035e-05,
      "loss": 0.6635,
      "mean_token_accuracy": 0.8324669599533081,
      "num_tokens": 16777961.0,
      "step": 359
    },
    {
      "epoch": 4.556962025316456,
      "grad_norm": 0.5440343618392944,
      "learning_rate": 5.797440732227555e-05,
      "loss": 0.7234,
      "mean_token_accuracy": 0.8209396004676819,
      "num_tokens": 16827472.0,
      "step": 360
    },
    {
      "epoch": 4.569620253164557,
      "grad_norm": 0.5303831100463867,
      "learning_rate": 5.744328543152253e-05,
      "loss": 0.662,
      "mean_token_accuracy": 0.8331236243247986,
      "num_tokens": 16872839.0,
      "step": 361
    },
    {
      "epoch": 4.582278481012658,
      "grad_norm": 0.563178539276123,
      "learning_rate": 5.691362550590297e-05,
      "loss": 0.6625,
      "mean_token_accuracy": 0.8346900343894958,
      "num_tokens": 16915163.0,
      "step": 362
    },
    {
      "epoch": 4.594936708860759,
      "grad_norm": 0.5247779488563538,
      "learning_rate": 5.638544574100249e-05,
      "loss": 0.7099,
      "mean_token_accuracy": 0.823519766330719,
      "num_tokens": 16963986.0,
      "step": 363
    },
    {
      "epoch": 4.6075949367088604,
      "grad_norm": 0.5180805325508118,
      "learning_rate": 5.585876428155824e-05,
      "loss": 0.6988,
      "mean_token_accuracy": 0.8249117732048035,
      "num_tokens": 17012220.0,
      "step": 364
    },
    {
      "epoch": 4.620253164556962,
      "grad_norm": 0.5352941751480103,
      "learning_rate": 5.533359922083562e-05,
      "loss": 0.6849,
      "mean_token_accuracy": 0.8279665112495422,
      "num_tokens": 17058711.0,
      "step": 365
    },
    {
      "epoch": 4.632911392405063,
      "grad_norm": 0.5777114033699036,
      "learning_rate": 5.4809968600006635e-05,
      "loss": 0.678,
      "mean_token_accuracy": 0.8303003907203674,
      "num_tokens": 17102682.0,
      "step": 366
    },
    {
      "epoch": 4.6455696202531644,
      "grad_norm": 0.5614490509033203,
      "learning_rate": 5.4287890407530175e-05,
      "loss": 0.7108,
      "mean_token_accuracy": 0.8221400380134583,
      "num_tokens": 17149530.0,
      "step": 367
    },
    {
      "epoch": 4.658227848101266,
      "grad_norm": 0.5888735055923462,
      "learning_rate": 5.3767382578534e-05,
      "loss": 0.6856,
      "mean_token_accuracy": 0.828814685344696,
      "num_tokens": 17193634.0,
      "step": 368
    },
    {
      "epoch": 4.670886075949367,
      "grad_norm": 0.5642009377479553,
      "learning_rate": 5.324846299419879e-05,
      "loss": 0.7221,
      "mean_token_accuracy": 0.8182055950164795,
      "num_tokens": 17243375.0,
      "step": 369
    },
    {
      "epoch": 4.6835443037974684,
      "grad_norm": 0.5699864625930786,
      "learning_rate": 5.273114948114346e-05,
      "loss": 0.6912,
      "mean_token_accuracy": 0.8257492780685425,
      "num_tokens": 17288116.0,
      "step": 370
    },
    {
      "epoch": 4.69620253164557,
      "grad_norm": 0.5664955973625183,
      "learning_rate": 5.2215459810813306e-05,
      "loss": 0.7255,
      "mean_token_accuracy": 0.8178353309631348,
      "num_tokens": 17335928.0,
      "step": 371
    },
    {
      "epoch": 4.708860759493671,
      "grad_norm": 0.5631005167961121,
      "learning_rate": 5.170141169886904e-05,
      "loss": 0.6987,
      "mean_token_accuracy": 0.8256130218505859,
      "num_tokens": 17382320.0,
      "step": 372
    },
    {
      "epoch": 4.7215189873417724,
      "grad_norm": 0.5475056767463684,
      "learning_rate": 5.118902280457829e-05,
      "loss": 0.7383,
      "mean_token_accuracy": 0.8166825771331787,
      "num_tokens": 17432712.0,
      "step": 373
    },
    {
      "epoch": 4.734177215189874,
      "grad_norm": 0.5569179654121399,
      "learning_rate": 5.0678310730209275e-05,
      "loss": 0.7157,
      "mean_token_accuracy": 0.8199924230575562,
      "num_tokens": 17480174.0,
      "step": 374
    },
    {
      "epoch": 4.746835443037975,
      "grad_norm": 0.5541711449623108,
      "learning_rate": 5.016929302042563e-05,
      "loss": 0.6978,
      "mean_token_accuracy": 0.826197624206543,
      "num_tokens": 17528563.0,
      "step": 375
    },
    {
      "epoch": 4.759493670886076,
      "grad_norm": 0.5538228750228882,
      "learning_rate": 4.9661987161684045e-05,
      "loss": 0.693,
      "mean_token_accuracy": 0.8276927471160889,
      "num_tokens": 17573303.0,
      "step": 376
    },
    {
      "epoch": 4.772151898734177,
      "grad_norm": 0.5440813899040222,
      "learning_rate": 4.9156410581633317e-05,
      "loss": 0.699,
      "mean_token_accuracy": 0.8246811628341675,
      "num_tokens": 17620099.0,
      "step": 377
    },
    {
      "epoch": 4.784810126582278,
      "grad_norm": 0.5575851202011108,
      "learning_rate": 4.865258064851579e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.8187082409858704,
      "num_tokens": 17666100.0,
      "step": 378
    },
    {
      "epoch": 4.7974683544303796,
      "grad_norm": 0.5372059941291809,
      "learning_rate": 4.8150514670570754e-05,
      "loss": 0.7156,
      "mean_token_accuracy": 0.8224175572395325,
      "num_tokens": 17716107.0,
      "step": 379
    },
    {
      "epoch": 4.810126582278481,
      "grad_norm": 0.5502781271934509,
      "learning_rate": 4.765022989543958e-05,
      "loss": 0.6808,
      "mean_token_accuracy": 0.8300414085388184,
      "num_tokens": 17762547.0,
      "step": 380
    },
    {
      "epoch": 4.822784810126582,
      "grad_norm": 0.5721256732940674,
      "learning_rate": 4.7151743509573444e-05,
      "loss": 0.7167,
      "mean_token_accuracy": 0.8231954574584961,
      "num_tokens": 17808939.0,
      "step": 381
    },
    {
      "epoch": 4.8354430379746836,
      "grad_norm": 0.557105541229248,
      "learning_rate": 4.665507263764299e-05,
      "loss": 0.6788,
      "mean_token_accuracy": 0.831015944480896,
      "num_tokens": 17854457.0,
      "step": 382
    },
    {
      "epoch": 4.848101265822785,
      "grad_norm": 0.560921311378479,
      "learning_rate": 4.6160234341949646e-05,
      "loss": 0.7473,
      "mean_token_accuracy": 0.8154394030570984,
      "num_tokens": 17901958.0,
      "step": 383
    },
    {
      "epoch": 4.860759493670886,
      "grad_norm": 0.5427091717720032,
      "learning_rate": 4.5667245621839974e-05,
      "loss": 0.6843,
      "mean_token_accuracy": 0.8294286727905273,
      "num_tokens": 17949316.0,
      "step": 384
    },
    {
      "epoch": 4.8734177215189876,
      "grad_norm": 0.54844731092453,
      "learning_rate": 4.5176123413121284e-05,
      "loss": 0.7267,
      "mean_token_accuracy": 0.8211008310317993,
      "num_tokens": 17998525.0,
      "step": 385
    },
    {
      "epoch": 4.886075949367089,
      "grad_norm": 0.5478676557540894,
      "learning_rate": 4.468688458748006e-05,
      "loss": 0.7032,
      "mean_token_accuracy": 0.8235698342323303,
      "num_tokens": 18045174.0,
      "step": 386
    },
    {
      "epoch": 4.89873417721519,
      "grad_norm": 0.5715044140815735,
      "learning_rate": 4.4199545951902286e-05,
      "loss": 0.7351,
      "mean_token_accuracy": 0.8134607076644897,
      "num_tokens": 18094745.0,
      "step": 387
    },
    {
      "epoch": 4.911392405063291,
      "grad_norm": 0.5567689538002014,
      "learning_rate": 4.3714124248096067e-05,
      "loss": 0.72,
      "mean_token_accuracy": 0.8220183253288269,
      "num_tokens": 18142769.0,
      "step": 388
    },
    {
      "epoch": 4.924050632911392,
      "grad_norm": 0.5515459179878235,
      "learning_rate": 4.3230636151916495e-05,
      "loss": 0.7314,
      "mean_token_accuracy": 0.8181837201118469,
      "num_tokens": 18190392.0,
      "step": 389
    },
    {
      "epoch": 4.936708860759493,
      "grad_norm": 0.5599684715270996,
      "learning_rate": 4.274909827279283e-05,
      "loss": 0.6852,
      "mean_token_accuracy": 0.8275984525680542,
      "num_tokens": 18238292.0,
      "step": 390
    },
    {
      "epoch": 4.949367088607595,
      "grad_norm": 0.581655740737915,
      "learning_rate": 4.226952715315779e-05,
      "loss": 0.6965,
      "mean_token_accuracy": 0.8275977969169617,
      "num_tokens": 18282932.0,
      "step": 391
    },
    {
      "epoch": 4.962025316455696,
      "grad_norm": 0.5861680507659912,
      "learning_rate": 4.17919392678795e-05,
      "loss": 0.6881,
      "mean_token_accuracy": 0.8262351155281067,
      "num_tokens": 18325542.0,
      "step": 392
    },
    {
      "epoch": 4.974683544303797,
      "grad_norm": 0.5522813200950623,
      "learning_rate": 4.131635102369513e-05,
      "loss": 0.7188,
      "mean_token_accuracy": 0.8196642398834229,
      "num_tokens": 18373317.0,
      "step": 393
    },
    {
      "epoch": 4.987341772151899,
      "grad_norm": 0.5678694248199463,
      "learning_rate": 4.084277875864776e-05,
      "loss": 0.7326,
      "mean_token_accuracy": 0.8180271983146667,
      "num_tokens": 18422185.0,
      "step": 394
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.4936027526855469,
      "learning_rate": 4.037123874152472e-05,
      "loss": 0.5711,
      "mean_token_accuracy": 0.8599328398704529,
      "num_tokens": 18468998.0,
      "step": 395
    },
    {
      "epoch": 5.012658227848101,
      "grad_norm": 0.5264456272125244,
      "learning_rate": 3.9901747171298766e-05,
      "loss": 0.5328,
      "mean_token_accuracy": 0.8718881011009216,
      "num_tokens": 18514452.0,
      "step": 396
    },
    {
      "epoch": 5.025316455696203,
      "grad_norm": 0.5222989320755005,
      "learning_rate": 3.943432017657186e-05,
      "loss": 0.5193,
      "mean_token_accuracy": 0.8739416599273682,
      "num_tokens": 18561288.0,
      "step": 397
    },
    {
      "epoch": 5.037974683544304,
      "grad_norm": 0.5219823718070984,
      "learning_rate": 3.8968973815020806e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.884689211845398,
      "num_tokens": 18607731.0,
      "step": 398
    },
    {
      "epoch": 5.050632911392405,
      "grad_norm": 0.5559691786766052,
      "learning_rate": 3.850572407284569e-05,
      "loss": 0.4868,
      "mean_token_accuracy": 0.8799352645874023,
      "num_tokens": 18650422.0,
      "step": 399
    },
    {
      "epoch": 5.063291139240507,
      "grad_norm": 0.6586586236953735,
      "learning_rate": 3.80445868642208e-05,
      "loss": 0.5133,
      "mean_token_accuracy": 0.8707754015922546,
      "num_tokens": 18694425.0,
      "step": 400
    },
    {
      "epoch": 5.075949367088608,
      "grad_norm": 0.7043012976646423,
      "learning_rate": 3.7585578030747744e-05,
      "loss": 0.4937,
      "mean_token_accuracy": 0.8765274882316589,
      "num_tokens": 18743180.0,
      "step": 401
    },
    {
      "epoch": 5.0886075949367084,
      "grad_norm": 0.7497896552085876,
      "learning_rate": 3.7128713340911535e-05,
      "loss": 0.4978,
      "mean_token_accuracy": 0.8755159974098206,
      "num_tokens": 18792905.0,
      "step": 402
    },
    {
      "epoch": 5.10126582278481,
      "grad_norm": 0.7151162028312683,
      "learning_rate": 3.667400848953845e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.875111997127533,
      "num_tokens": 18839835.0,
      "step": 403
    },
    {
      "epoch": 5.113924050632911,
      "grad_norm": 0.6478233337402344,
      "learning_rate": 3.622147909725724e-05,
      "loss": 0.4802,
      "mean_token_accuracy": 0.8800956010818481,
      "num_tokens": 18888029.0,
      "step": 404
    },
    {
      "epoch": 5.1265822784810124,
      "grad_norm": 0.6064746975898743,
      "learning_rate": 3.577114070996247e-05,
      "loss": 0.5117,
      "mean_token_accuracy": 0.8728470802307129,
      "num_tokens": 18933322.0,
      "step": 405
    },
    {
      "epoch": 5.139240506329114,
      "grad_norm": 0.5626401305198669,
      "learning_rate": 3.532300879828013e-05,
      "loss": 0.4832,
      "mean_token_accuracy": 0.8781822919845581,
      "num_tokens": 18980522.0,
      "step": 406
    },
    {
      "epoch": 5.151898734177215,
      "grad_norm": 0.546474039554596,
      "learning_rate": 3.4877098757036665e-05,
      "loss": 0.5058,
      "mean_token_accuracy": 0.8778883218765259,
      "num_tokens": 19028624.0,
      "step": 407
    },
    {
      "epoch": 5.1645569620253164,
      "grad_norm": 0.5445472002029419,
      "learning_rate": 3.44334259047297e-05,
      "loss": 0.4888,
      "mean_token_accuracy": 0.8800671696662903,
      "num_tokens": 19074522.0,
      "step": 408
    },
    {
      "epoch": 5.177215189873418,
      "grad_norm": 0.531647801399231,
      "learning_rate": 3.3992005483002e-05,
      "loss": 0.4776,
      "mean_token_accuracy": 0.8834218978881836,
      "num_tokens": 19122571.0,
      "step": 409
    },
    {
      "epoch": 5.189873417721519,
      "grad_norm": 0.5426664352416992,
      "learning_rate": 3.355285265611784e-05,
      "loss": 0.5024,
      "mean_token_accuracy": 0.8779997825622559,
      "num_tokens": 19171389.0,
      "step": 410
    },
    {
      "epoch": 5.2025316455696204,
      "grad_norm": 0.5555723905563354,
      "learning_rate": 3.3115982510442014e-05,
      "loss": 0.5103,
      "mean_token_accuracy": 0.8771188855171204,
      "num_tokens": 19219296.0,
      "step": 411
    },
    {
      "epoch": 5.215189873417722,
      "grad_norm": 0.5892623066902161,
      "learning_rate": 3.268141005392163e-05,
      "loss": 0.4948,
      "mean_token_accuracy": 0.8789801597595215,
      "num_tokens": 19262386.0,
      "step": 412
    },
    {
      "epoch": 5.227848101265823,
      "grad_norm": 0.5882421731948853,
      "learning_rate": 3.224915021557049e-05,
      "loss": 0.4821,
      "mean_token_accuracy": 0.8825682401657104,
      "num_tokens": 19307259.0,
      "step": 413
    },
    {
      "epoch": 5.2405063291139244,
      "grad_norm": 0.6074917316436768,
      "learning_rate": 3.1819217844956214e-05,
      "loss": 0.4925,
      "mean_token_accuracy": 0.8789646625518799,
      "num_tokens": 19355152.0,
      "step": 414
    },
    {
      "epoch": 5.253164556962025,
      "grad_norm": 0.6312512159347534,
      "learning_rate": 3.1391627711690296e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8776240944862366,
      "num_tokens": 19400470.0,
      "step": 415
    },
    {
      "epoch": 5.265822784810126,
      "grad_norm": 0.6164311766624451,
      "learning_rate": 3.0966394504920317e-05,
      "loss": 0.4695,
      "mean_token_accuracy": 0.8826056718826294,
      "num_tokens": 19446848.0,
      "step": 416
    },
    {
      "epoch": 5.2784810126582276,
      "grad_norm": 0.6131919026374817,
      "learning_rate": 3.0543532832825715e-05,
      "loss": 0.4808,
      "mean_token_accuracy": 0.8802450299263,
      "num_tokens": 19493432.0,
      "step": 417
    },
    {
      "epoch": 5.291139240506329,
      "grad_norm": 0.6002056002616882,
      "learning_rate": 3.0123057222115836e-05,
      "loss": 0.4986,
      "mean_token_accuracy": 0.8771020174026489,
      "num_tokens": 19542496.0,
      "step": 418
    },
    {
      "epoch": 5.30379746835443,
      "grad_norm": 0.6012871265411377,
      "learning_rate": 2.9704982117530777e-05,
      "loss": 0.5124,
      "mean_token_accuracy": 0.8735976815223694,
      "num_tokens": 19590605.0,
      "step": 419
    },
    {
      "epoch": 5.3164556962025316,
      "grad_norm": 0.5826122164726257,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8811623454093933,
      "num_tokens": 19636816.0,
      "step": 420
    },
    {
      "epoch": 5.329113924050633,
      "grad_norm": 0.5988377332687378,
      "learning_rate": 2.887609079287521e-05,
      "loss": 0.4867,
      "mean_token_accuracy": 0.8805438280105591,
      "num_tokens": 19683441.0,
      "step": 421
    },
    {
      "epoch": 5.341772151898734,
      "grad_norm": 0.5812557339668274,
      "learning_rate": 2.8465303047987267e-05,
      "loss": 0.4878,
      "mean_token_accuracy": 0.8807899951934814,
      "num_tokens": 19731605.0,
      "step": 422
    },
    {
      "epoch": 5.3544303797468356,
      "grad_norm": 0.5604175329208374,
      "learning_rate": 2.805697275861101e-05,
      "loss": 0.4961,
      "mean_token_accuracy": 0.8778889179229736,
      "num_tokens": 19780346.0,
      "step": 423
    },
    {
      "epoch": 5.367088607594937,
      "grad_norm": 0.5724864602088928,
      "learning_rate": 2.765111395225424e-05,
      "loss": 0.4884,
      "mean_token_accuracy": 0.8789756894111633,
      "num_tokens": 19828326.0,
      "step": 424
    },
    {
      "epoch": 5.379746835443038,
      "grad_norm": 0.5703694224357605,
      "learning_rate": 2.7247740571521118e-05,
      "loss": 0.4896,
      "mean_token_accuracy": 0.8801313042640686,
      "num_tokens": 19875600.0,
      "step": 425
    },
    {
      "epoch": 5.3924050632911396,
      "grad_norm": 0.5990126132965088,
      "learning_rate": 2.6846866473633125e-05,
      "loss": 0.4726,
      "mean_token_accuracy": 0.8829923868179321,
      "num_tokens": 19919174.0,
      "step": 426
    },
    {
      "epoch": 5.405063291139241,
      "grad_norm": 0.6109516024589539,
      "learning_rate": 2.6448505429952917e-05,
      "loss": 0.4907,
      "mean_token_accuracy": 0.8786324262619019,
      "num_tokens": 19963492.0,
      "step": 427
    },
    {
      "epoch": 5.417721518987342,
      "grad_norm": 0.5848973393440247,
      "learning_rate": 2.605267112551154e-05,
      "loss": 0.5069,
      "mean_token_accuracy": 0.8753526210784912,
      "num_tokens": 20011058.0,
      "step": 428
    },
    {
      "epoch": 5.430379746835443,
      "grad_norm": 0.590811014175415,
      "learning_rate": 2.5659377158538012e-05,
      "loss": 0.499,
      "mean_token_accuracy": 0.8782621622085571,
      "num_tokens": 20058905.0,
      "step": 429
    },
    {
      "epoch": 5.443037974683544,
      "grad_norm": 0.6152106523513794,
      "learning_rate": 2.5268637039992293e-05,
      "loss": 0.4988,
      "mean_token_accuracy": 0.875298023223877,
      "num_tokens": 20105945.0,
      "step": 430
    },
    {
      "epoch": 5.455696202531645,
      "grad_norm": 0.592876136302948,
      "learning_rate": 2.488046419310114e-05,
      "loss": 0.4965,
      "mean_token_accuracy": 0.877647340297699,
      "num_tokens": 20153462.0,
      "step": 431
    },
    {
      "epoch": 5.468354430379747,
      "grad_norm": 0.6070770025253296,
      "learning_rate": 2.4494871952896947e-05,
      "loss": 0.4933,
      "mean_token_accuracy": 0.8780992031097412,
      "num_tokens": 20197529.0,
      "step": 432
    },
    {
      "epoch": 5.481012658227848,
      "grad_norm": 0.5933142900466919,
      "learning_rate": 2.411187356575969e-05,
      "loss": 0.4813,
      "mean_token_accuracy": 0.880434513092041,
      "num_tokens": 20241594.0,
      "step": 433
    },
    {
      "epoch": 5.493670886075949,
      "grad_norm": 0.5929135084152222,
      "learning_rate": 2.3731482188961818e-05,
      "loss": 0.4783,
      "mean_token_accuracy": 0.8830586671829224,
      "num_tokens": 20288083.0,
      "step": 434
    },
    {
      "epoch": 5.506329113924051,
      "grad_norm": 0.5733610391616821,
      "learning_rate": 2.335371089021623e-05,
      "loss": 0.49,
      "mean_token_accuracy": 0.8810379505157471,
      "num_tokens": 20337667.0,
      "step": 435
    },
    {
      "epoch": 5.518987341772152,
      "grad_norm": 0.5939366221427917,
      "learning_rate": 2.297857264722756e-05,
      "loss": 0.4756,
      "mean_token_accuracy": 0.8813409209251404,
      "num_tokens": 20385161.0,
      "step": 436
    },
    {
      "epoch": 5.531645569620253,
      "grad_norm": 0.5886082649230957,
      "learning_rate": 2.260608034724595e-05,
      "loss": 0.5091,
      "mean_token_accuracy": 0.8750845789909363,
      "num_tokens": 20432529.0,
      "step": 437
    },
    {
      "epoch": 5.544303797468355,
      "grad_norm": 0.6086618900299072,
      "learning_rate": 2.2236246786624792e-05,
      "loss": 0.4902,
      "mean_token_accuracy": 0.8768316507339478,
      "num_tokens": 20476679.0,
      "step": 438
    },
    {
      "epoch": 5.556962025316456,
      "grad_norm": 0.5850755572319031,
      "learning_rate": 2.1869084670380835e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8765684962272644,
      "num_tokens": 20524559.0,
      "step": 439
    },
    {
      "epoch": 5.569620253164557,
      "grad_norm": 0.5923234224319458,
      "learning_rate": 2.150460661175768e-05,
      "loss": 0.4736,
      "mean_token_accuracy": 0.8819867372512817,
      "num_tokens": 20568152.0,
      "step": 440
    },
    {
      "epoch": 5.582278481012658,
      "grad_norm": 0.5886726975440979,
      "learning_rate": 2.114282513179281e-05,
      "loss": 0.4752,
      "mean_token_accuracy": 0.8854075074195862,
      "num_tokens": 20613760.0,
      "step": 441
    },
    {
      "epoch": 5.594936708860759,
      "grad_norm": 0.6090703010559082,
      "learning_rate": 2.0783752658887066e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8821261525154114,
      "num_tokens": 20658957.0,
      "step": 442
    },
    {
      "epoch": 5.6075949367088604,
      "grad_norm": 0.6041176915168762,
      "learning_rate": 2.0427401528377953e-05,
      "loss": 0.4982,
      "mean_token_accuracy": 0.8764042258262634,
      "num_tokens": 20704330.0,
      "step": 443
    },
    {
      "epoch": 5.620253164556962,
      "grad_norm": 0.60368812084198,
      "learning_rate": 2.0073783982115723e-05,
      "loss": 0.5062,
      "mean_token_accuracy": 0.8757708668708801,
      "num_tokens": 20752555.0,
      "step": 444
    },
    {
      "epoch": 5.632911392405063,
      "grad_norm": 0.5774247050285339,
      "learning_rate": 1.9722912168042897e-05,
      "loss": 0.511,
      "mean_token_accuracy": 0.8744285106658936,
      "num_tokens": 20805115.0,
      "step": 445
    },
    {
      "epoch": 5.6455696202531644,
      "grad_norm": 0.6094748973846436,
      "learning_rate": 1.937479813977703e-05,
      "loss": 0.4879,
      "mean_token_accuracy": 0.8806090950965881,
      "num_tokens": 20849571.0,
      "step": 446
    },
    {
      "epoch": 5.658227848101266,
      "grad_norm": 0.6145944595336914,
      "learning_rate": 1.9029453856196376e-05,
      "loss": 0.5112,
      "mean_token_accuracy": 0.8738813400268555,
      "num_tokens": 20896678.0,
      "step": 447
    },
    {
      "epoch": 5.670886075949367,
      "grad_norm": 0.5843153595924377,
      "learning_rate": 1.868689118102931e-05,
      "loss": 0.4848,
      "mean_token_accuracy": 0.8810961246490479,
      "num_tokens": 20942796.0,
      "step": 448
    },
    {
      "epoch": 5.6835443037974684,
      "grad_norm": 0.5841859579086304,
      "learning_rate": 1.8347121882446717e-05,
      "loss": 0.4932,
      "mean_token_accuracy": 0.8793487548828125,
      "num_tokens": 20989971.0,
      "step": 449
    },
    {
      "epoch": 5.69620253164557,
      "grad_norm": 0.606610894203186,
      "learning_rate": 1.8010157632657543e-05,
      "loss": 0.4786,
      "mean_token_accuracy": 0.8826842904090881,
      "num_tokens": 21037258.0,
      "step": 450
    },
    {
      "epoch": 5.708860759493671,
      "grad_norm": 0.6054370999336243,
      "learning_rate": 1.7676010007508092e-05,
      "loss": 0.4914,
      "mean_token_accuracy": 0.8785582184791565,
      "num_tokens": 21082183.0,
      "step": 451
    },
    {
      "epoch": 5.7215189873417724,
      "grad_norm": 0.6061049699783325,
      "learning_rate": 1.7344690486084137e-05,
      "loss": 0.5013,
      "mean_token_accuracy": 0.8739482760429382,
      "num_tokens": 21126697.0,
      "step": 452
    },
    {
      "epoch": 5.734177215189874,
      "grad_norm": 0.5808444023132324,
      "learning_rate": 1.701621045031666e-05,
      "loss": 0.5065,
      "mean_token_accuracy": 0.8762862086296082,
      "num_tokens": 21177006.0,
      "step": 453
    },
    {
      "epoch": 5.746835443037975,
      "grad_norm": 0.5869651436805725,
      "learning_rate": 1.6690581184590858e-05,
      "loss": 0.5022,
      "mean_token_accuracy": 0.8760119080543518,
      "num_tokens": 21224010.0,
      "step": 454
    },
    {
      "epoch": 5.759493670886076,
      "grad_norm": 0.5848543047904968,
      "learning_rate": 1.636781387535843e-05,
      "loss": 0.5056,
      "mean_token_accuracy": 0.8729737401008606,
      "num_tokens": 21273056.0,
      "step": 455
    },
    {
      "epoch": 5.772151898734177,
      "grad_norm": 0.6035214066505432,
      "learning_rate": 1.604791961075336e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8771805167198181,
      "num_tokens": 21318121.0,
      "step": 456
    },
    {
      "epoch": 5.784810126582278,
      "grad_norm": 0.587157666683197,
      "learning_rate": 1.5730909380210945e-05,
      "loss": 0.485,
      "mean_token_accuracy": 0.8795527815818787,
      "num_tokens": 21366754.0,
      "step": 457
    },
    {
      "epoch": 5.7974683544303796,
      "grad_norm": 0.5755473971366882,
      "learning_rate": 1.5416794074090258e-05,
      "loss": 0.495,
      "mean_token_accuracy": 0.8767243027687073,
      "num_tokens": 21415895.0,
      "step": 458
    },
    {
      "epoch": 5.810126582278481,
      "grad_norm": 0.5992995500564575,
      "learning_rate": 1.5105584483300162e-05,
      "loss": 0.462,
      "mean_token_accuracy": 0.88564532995224,
      "num_tokens": 21461554.0,
      "step": 459
    },
    {
      "epoch": 5.822784810126582,
      "grad_norm": 0.6188696622848511,
      "learning_rate": 1.479729129892835e-05,
      "loss": 0.4892,
      "mean_token_accuracy": 0.8776053190231323,
      "num_tokens": 21506718.0,
      "step": 460
    },
    {
      "epoch": 5.8354430379746836,
      "grad_norm": 0.6524665951728821,
      "learning_rate": 1.4491925111874383e-05,
      "loss": 0.4889,
      "mean_token_accuracy": 0.8804559111595154,
      "num_tokens": 21551527.0,
      "step": 461
    },
    {
      "epoch": 5.848101265822785,
      "grad_norm": 0.6141889095306396,
      "learning_rate": 1.4189496412485592e-05,
      "loss": 0.4781,
      "mean_token_accuracy": 0.8806585073471069,
      "num_tokens": 21597694.0,
      "step": 462
    },
    {
      "epoch": 5.860759493670886,
      "grad_norm": 0.6022236943244934,
      "learning_rate": 1.3890015590196803e-05,
      "loss": 0.4957,
      "mean_token_accuracy": 0.8784832954406738,
      "num_tokens": 21644912.0,
      "step": 463
    },
    {
      "epoch": 5.8734177215189876,
      "grad_norm": 0.5849525332450867,
      "learning_rate": 1.3593492933173512e-05,
      "loss": 0.4981,
      "mean_token_accuracy": 0.8747273683547974,
      "num_tokens": 21694492.0,
      "step": 464
    },
    {
      "epoch": 5.886075949367089,
      "grad_norm": 0.5798141360282898,
      "learning_rate": 1.3299938627958297e-05,
      "loss": 0.5031,
      "mean_token_accuracy": 0.8752538561820984,
      "num_tokens": 21745772.0,
      "step": 465
    },
    {
      "epoch": 5.89873417721519,
      "grad_norm": 0.6097593903541565,
      "learning_rate": 1.300936275912098e-05,
      "loss": 0.4789,
      "mean_token_accuracy": 0.8802154660224915,
      "num_tokens": 21790391.0,
      "step": 466
    },
    {
      "epoch": 5.911392405063291,
      "grad_norm": 0.6151306629180908,
      "learning_rate": 1.2721775308912132e-05,
      "loss": 0.5194,
      "mean_token_accuracy": 0.8713752031326294,
      "num_tokens": 21837009.0,
      "step": 467
    },
    {
      "epoch": 5.924050632911392,
      "grad_norm": 0.5922465324401855,
      "learning_rate": 1.2437186156920167e-05,
      "loss": 0.5141,
      "mean_token_accuracy": 0.8723000288009644,
      "num_tokens": 21884990.0,
      "step": 468
    },
    {
      "epoch": 5.936708860759493,
      "grad_norm": 0.5962823033332825,
      "learning_rate": 1.2155605079732046e-05,
      "loss": 0.5089,
      "mean_token_accuracy": 0.8751126527786255,
      "num_tokens": 21931648.0,
      "step": 469
    },
    {
      "epoch": 5.949367088607595,
      "grad_norm": 0.6032353639602661,
      "learning_rate": 1.1877041750597173e-05,
      "loss": 0.4964,
      "mean_token_accuracy": 0.8769810795783997,
      "num_tokens": 21978656.0,
      "step": 470
    },
    {
      "epoch": 5.962025316455696,
      "grad_norm": 0.6158297657966614,
      "learning_rate": 1.160150573909532e-05,
      "loss": 0.5099,
      "mean_token_accuracy": 0.8746172189712524,
      "num_tokens": 22026071.0,
      "step": 471
    },
    {
      "epoch": 5.974683544303797,
      "grad_norm": 0.5986502766609192,
      "learning_rate": 1.132900651080785e-05,
      "loss": 0.4691,
      "mean_token_accuracy": 0.8836339712142944,
      "num_tokens": 22070263.0,
      "step": 472
    },
    {
      "epoch": 5.987341772151899,
      "grad_norm": 0.6113969087600708,
      "learning_rate": 1.1059553426992365e-05,
      "loss": 0.5074,
      "mean_token_accuracy": 0.8743844032287598,
      "num_tokens": 22117638.0,
      "step": 473
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5078223347663879,
      "learning_rate": 1.0793155744261351e-05,
      "loss": 0.4089,
      "mean_token_accuracy": 0.9015847444534302,
      "num_tokens": 22162756.0,
      "step": 474
    },
    {
      "epoch": 6.012658227848101,
      "grad_norm": 0.5180269479751587,
      "learning_rate": 1.0529822614264018e-05,
      "loss": 0.4232,
      "mean_token_accuracy": 0.8996778726577759,
      "num_tokens": 22210317.0,
      "step": 475
    },
    {
      "epoch": 6.025316455696203,
      "grad_norm": 0.521119236946106,
      "learning_rate": 1.026956308337199e-05,
      "loss": 0.4206,
      "mean_token_accuracy": 0.9003894925117493,
      "num_tokens": 22257364.0,
      "step": 476
    },
    {
      "epoch": 6.037974683544304,
      "grad_norm": 0.5258929133415222,
      "learning_rate": 1.0012386092368475e-05,
      "loss": 0.3874,
      "mean_token_accuracy": 0.9093874096870422,
      "num_tokens": 22302190.0,
      "step": 477
    },
    {
      "epoch": 6.050632911392405,
      "grad_norm": 0.5309813618659973,
      "learning_rate": 9.75830047614117e-06,
      "loss": 0.4006,
      "mean_token_accuracy": 0.9050379395484924,
      "num_tokens": 22347967.0,
      "step": 478
    },
    {
      "epoch": 6.063291139240507,
      "grad_norm": 0.5413678288459778,
      "learning_rate": 9.507314963378745e-06,
      "loss": 0.3956,
      "mean_token_accuracy": 0.905818521976471,
      "num_tokens": 22393369.0,
      "step": 479
    },
    {
      "epoch": 6.075949367088608,
      "grad_norm": 0.5256166458129883,
      "learning_rate": 9.259438176270962e-06,
      "loss": 0.3977,
      "mean_token_accuracy": 0.9056070446968079,
      "num_tokens": 22441212.0,
      "step": 480
    },
    {
      "epoch": 6.0886075949367084,
      "grad_norm": 0.5342482328414917,
      "learning_rate": 9.014678630212469e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.9074738025665283,
      "num_tokens": 22489792.0,
      "step": 481
    },
    {
      "epoch": 6.10126582278481,
      "grad_norm": 0.5547141432762146,
      "learning_rate": 8.773044733510338e-06,
      "loss": 0.3841,
      "mean_token_accuracy": 0.9071083068847656,
      "num_tokens": 22537589.0,
      "step": 482
    },
    {
      "epoch": 6.113924050632911,
      "grad_norm": 0.6087285280227661,
      "learning_rate": 8.534544787095078e-06,
      "loss": 0.3873,
      "mean_token_accuracy": 0.9074901342391968,
      "num_tokens": 22583540.0,
      "step": 483
    },
    {
      "epoch": 6.1265822784810124,
      "grad_norm": 0.6063216328620911,
      "learning_rate": 8.299186984235585e-06,
      "loss": 0.3916,
      "mean_token_accuracy": 0.9041749238967896,
      "num_tokens": 22630982.0,
      "step": 484
    },
    {
      "epoch": 6.139240506329114,
      "grad_norm": 0.613871693611145,
      "learning_rate": 8.06697941025768e-06,
      "loss": 0.3752,
      "mean_token_accuracy": 0.9080203175544739,
      "num_tokens": 22675795.0,
      "step": 485
    },
    {
      "epoch": 6.151898734177215,
      "grad_norm": 0.6383016705513,
      "learning_rate": 7.837930042266262e-06,
      "loss": 0.4072,
      "mean_token_accuracy": 0.9005741477012634,
      "num_tokens": 22724800.0,
      "step": 486
    },
    {
      "epoch": 6.1645569620253164,
      "grad_norm": 0.6214612126350403,
      "learning_rate": 7.612046748871327e-06,
      "loss": 0.3898,
      "mean_token_accuracy": 0.9051695466041565,
      "num_tokens": 22772528.0,
      "step": 487
    },
    {
      "epoch": 6.177215189873418,
      "grad_norm": 0.6447484493255615,
      "learning_rate": 7.389337289917652e-06,
      "loss": 0.4098,
      "mean_token_accuracy": 0.9014731049537659,
      "num_tokens": 22820518.0,
      "step": 488
    },
    {
      "epoch": 6.189873417721519,
      "grad_norm": 0.6162357330322266,
      "learning_rate": 7.1698093162182125e-06,
      "loss": 0.3847,
      "mean_token_accuracy": 0.9040086269378662,
      "num_tokens": 22866857.0,
      "step": 489
    },
    {
      "epoch": 6.2025316455696204,
      "grad_norm": 0.5872465968132019,
      "learning_rate": 6.953470369291348e-06,
      "loss": 0.3906,
      "mean_token_accuracy": 0.9052353501319885,
      "num_tokens": 22915399.0,
      "step": 490
    },
    {
      "epoch": 6.215189873417722,
      "grad_norm": 0.5958002209663391,
      "learning_rate": 6.74032788110166e-06,
      "loss": 0.3833,
      "mean_token_accuracy": 0.9080662727355957,
      "num_tokens": 22961333.0,
      "step": 491
    },
    {
      "epoch": 6.227848101265823,
      "grad_norm": 0.5847265124320984,
      "learning_rate": 6.530389173804807e-06,
      "loss": 0.388,
      "mean_token_accuracy": 0.9040992259979248,
      "num_tokens": 23007601.0,
      "step": 492
    },
    {
      "epoch": 6.2405063291139244,
      "grad_norm": 0.5710330605506897,
      "learning_rate": 6.323661459495811e-06,
      "loss": 0.3927,
      "mean_token_accuracy": 0.9061951637268066,
      "num_tokens": 23057268.0,
      "step": 493
    },
    {
      "epoch": 6.253164556962025,
      "grad_norm": 0.5893294811248779,
      "learning_rate": 6.1201518399613635e-06,
      "loss": 0.3649,
      "mean_token_accuracy": 0.9110230207443237,
      "num_tokens": 23101467.0,
      "step": 494
    },
    {
      "epoch": 6.265822784810126,
      "grad_norm": 0.5623788833618164,
      "learning_rate": 5.919867306435934e-06,
      "loss": 0.3651,
      "mean_token_accuracy": 0.9124348163604736,
      "num_tokens": 23146389.0,
      "step": 495
    },
    {
      "epoch": 6.2784810126582276,
      "grad_norm": 0.557794451713562,
      "learning_rate": 5.722814739361459e-06,
      "loss": 0.3735,
      "mean_token_accuracy": 0.9109086394309998,
      "num_tokens": 23192664.0,
      "step": 496
    },
    {
      "epoch": 6.291139240506329,
      "grad_norm": 0.5795067548751831,
      "learning_rate": 5.529000908151105e-06,
      "loss": 0.3955,
      "mean_token_accuracy": 0.9051991701126099,
      "num_tokens": 23238466.0,
      "step": 497
    },
    {
      "epoch": 6.30379746835443,
      "grad_norm": 0.5689795017242432,
      "learning_rate": 5.338432470956589e-06,
      "loss": 0.3903,
      "mean_token_accuracy": 0.9060965776443481,
      "num_tokens": 23287016.0,
      "step": 498
    },
    {
      "epoch": 6.3164556962025316,
      "grad_norm": 0.5454722046852112,
      "learning_rate": 5.151115974439569e-06,
      "loss": 0.3707,
      "mean_token_accuracy": 0.9100461602210999,
      "num_tokens": 23334093.0,
      "step": 499
    },
    {
      "epoch": 6.329113924050633,
      "grad_norm": 0.5693135261535645,
      "learning_rate": 4.967057853546653e-06,
      "loss": 0.3775,
      "mean_token_accuracy": 0.9082093238830566,
      "num_tokens": 23379118.0,
      "step": 500
    },
    {
      "epoch": 6.341772151898734,
      "grad_norm": 0.5589075684547424,
      "learning_rate": 4.786264431288423e-06,
      "loss": 0.3701,
      "mean_token_accuracy": 0.9106748104095459,
      "num_tokens": 23423089.0,
      "step": 501
    },
    {
      "epoch": 6.3544303797468356,
      "grad_norm": 0.5521835684776306,
      "learning_rate": 4.608741918522097e-06,
      "loss": 0.3806,
      "mean_token_accuracy": 0.9093685746192932,
      "num_tokens": 23469649.0,
      "step": 502
    },
    {
      "epoch": 6.367088607594937,
      "grad_norm": 0.5633716583251953,
      "learning_rate": 4.434496413738332e-06,
      "loss": 0.392,
      "mean_token_accuracy": 0.9065303802490234,
      "num_tokens": 23516862.0,
      "step": 503
    },
    {
      "epoch": 6.379746835443038,
      "grad_norm": 0.5469903945922852,
      "learning_rate": 4.263533902851535e-06,
      "loss": 0.3696,
      "mean_token_accuracy": 0.9113619923591614,
      "num_tokens": 23563520.0,
      "step": 504
    },
    {
      "epoch": 6.3924050632911396,
      "grad_norm": 0.5530360341072083,
      "learning_rate": 4.095860258994388e-06,
      "loss": 0.4062,
      "mean_token_accuracy": 0.9027206897735596,
      "num_tokens": 23612505.0,
      "step": 505
    },
    {
      "epoch": 6.405063291139241,
      "grad_norm": 0.5586188435554504,
      "learning_rate": 3.931481242315993e-06,
      "loss": 0.3604,
      "mean_token_accuracy": 0.9130181670188904,
      "num_tokens": 23658981.0,
      "step": 506
    },
    {
      "epoch": 6.417721518987342,
      "grad_norm": 0.5679466724395752,
      "learning_rate": 3.770402499783976e-06,
      "loss": 0.3913,
      "mean_token_accuracy": 0.9053800106048584,
      "num_tokens": 23704807.0,
      "step": 507
    },
    {
      "epoch": 6.430379746835443,
      "grad_norm": 0.5666420459747314,
      "learning_rate": 3.6126295649906216e-06,
      "loss": 0.3983,
      "mean_token_accuracy": 0.9038554430007935,
      "num_tokens": 23752518.0,
      "step": 508
    },
    {
      "epoch": 6.443037974683544,
      "grad_norm": 0.5628715753555298,
      "learning_rate": 3.458167857962613e-06,
      "loss": 0.3987,
      "mean_token_accuracy": 0.9056045413017273,
      "num_tokens": 23799205.0,
      "step": 509
    },
    {
      "epoch": 6.455696202531645,
      "grad_norm": 0.5618675947189331,
      "learning_rate": 3.3070226849749366e-06,
      "loss": 0.372,
      "mean_token_accuracy": 0.9113752841949463,
      "num_tokens": 23845114.0,
      "step": 510
    },
    {
      "epoch": 6.468354430379747,
      "grad_norm": 0.5676432847976685,
      "learning_rate": 3.159199238368593e-06,
      "loss": 0.3799,
      "mean_token_accuracy": 0.9109406471252441,
      "num_tokens": 23891720.0,
      "step": 511
    },
    {
      "epoch": 6.481012658227848,
      "grad_norm": 0.5845745205879211,
      "learning_rate": 3.0147025963721433e-06,
      "loss": 0.4173,
      "mean_token_accuracy": 0.9012231826782227,
      "num_tokens": 23938465.0,
      "step": 512
    },
    {
      "epoch": 6.493670886075949,
      "grad_norm": 0.5692048668861389,
      "learning_rate": 2.8735377229273998e-06,
      "loss": 0.3965,
      "mean_token_accuracy": 0.9037320613861084,
      "num_tokens": 23987403.0,
      "step": 513
    },
    {
      "epoch": 6.506329113924051,
      "grad_norm": 0.5562811493873596,
      "learning_rate": 2.735709467518699e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.9068973064422607,
      "num_tokens": 24034877.0,
      "step": 514
    },
    {
      "epoch": 6.518987341772152,
      "grad_norm": 0.5730811357498169,
      "learning_rate": 2.6012225650064893e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.9057947993278503,
      "num_tokens": 24082104.0,
      "step": 515
    },
    {
      "epoch": 6.531645569620253,
      "grad_norm": 0.5873252153396606,
      "learning_rate": 2.470081635464594e-06,
      "loss": 0.4219,
      "mean_token_accuracy": 0.8977401256561279,
      "num_tokens": 24129958.0,
      "step": 516
    },
    {
      "epoch": 6.544303797468355,
      "grad_norm": 0.5700328350067139,
      "learning_rate": 2.342291184021461e-06,
      "loss": 0.3936,
      "mean_token_accuracy": 0.9058045744895935,
      "num_tokens": 24176606.0,
      "step": 517
    },
    {
      "epoch": 6.556962025316456,
      "grad_norm": 0.5592919588088989,
      "learning_rate": 2.2178556007054872e-06,
      "loss": 0.3819,
      "mean_token_accuracy": 0.9092505574226379,
      "num_tokens": 24224494.0,
      "step": 518
    },
    {
      "epoch": 6.569620253164557,
      "grad_norm": 0.5832672119140625,
      "learning_rate": 2.0967791602941154e-06,
      "loss": 0.3858,
      "mean_token_accuracy": 0.9063412547111511,
      "num_tokens": 24271142.0,
      "step": 519
    },
    {
      "epoch": 6.582278481012658,
      "grad_norm": 0.5810838937759399,
      "learning_rate": 1.979066022167042e-06,
      "loss": 0.385,
      "mean_token_accuracy": 0.9060366153717041,
      "num_tokens": 24318948.0,
      "step": 520
    },
    {
      "epoch": 6.594936708860759,
      "grad_norm": 0.6128122806549072,
      "learning_rate": 1.8647202301633194e-06,
      "loss": 0.39,
      "mean_token_accuracy": 0.9060620665550232,
      "num_tokens": 24362743.0,
      "step": 521
    },
    {
      "epoch": 6.6075949367088604,
      "grad_norm": 0.5534291863441467,
      "learning_rate": 1.7537457124423895e-06,
      "loss": 0.3817,
      "mean_token_accuracy": 0.9087843298912048,
      "num_tokens": 24412031.0,
      "step": 522
    },
    {
      "epoch": 6.620253164556962,
      "grad_norm": 0.5778301954269409,
      "learning_rate": 1.6461462813492034e-06,
      "loss": 0.4057,
      "mean_token_accuracy": 0.9008567929267883,
      "num_tokens": 24459713.0,
      "step": 523
    },
    {
      "epoch": 6.632911392405063,
      "grad_norm": 0.5713034868240356,
      "learning_rate": 1.5419256332832255e-06,
      "loss": 0.3706,
      "mean_token_accuracy": 0.9108288288116455,
      "num_tokens": 24505913.0,
      "step": 524
    },
    {
      "epoch": 6.6455696202531644,
      "grad_norm": 0.5820364952087402,
      "learning_rate": 1.4410873485714238e-06,
      "loss": 0.3786,
      "mean_token_accuracy": 0.9070352911949158,
      "num_tokens": 24550467.0,
      "step": 525
    },
    {
      "epoch": 6.658227848101266,
      "grad_norm": 0.5694465637207031,
      "learning_rate": 1.3436348913453578e-06,
      "loss": 0.391,
      "mean_token_accuracy": 0.9062048196792603,
      "num_tokens": 24598945.0,
      "step": 526
    },
    {
      "epoch": 6.670886075949367,
      "grad_norm": 0.5639371871948242,
      "learning_rate": 1.249571609422029e-06,
      "loss": 0.3754,
      "mean_token_accuracy": 0.9100679755210876,
      "num_tokens": 24646089.0,
      "step": 527
    },
    {
      "epoch": 6.6835443037974684,
      "grad_norm": 0.5671195387840271,
      "learning_rate": 1.158900734189039e-06,
      "loss": 0.3881,
      "mean_token_accuracy": 0.9070022702217102,
      "num_tokens": 24692595.0,
      "step": 528
    },
    {
      "epoch": 6.69620253164557,
      "grad_norm": 0.5708573460578918,
      "learning_rate": 1.0716253804934795e-06,
      "loss": 0.4032,
      "mean_token_accuracy": 0.9024674892425537,
      "num_tokens": 24741494.0,
      "step": 529
    },
    {
      "epoch": 6.708860759493671,
      "grad_norm": 0.5634827017784119,
      "learning_rate": 9.877485465349058e-07,
      "loss": 0.3845,
      "mean_token_accuracy": 0.9086854457855225,
      "num_tokens": 24787520.0,
      "step": 530
    },
    {
      "epoch": 6.7215189873417724,
      "grad_norm": 0.5839213132858276,
      "learning_rate": 9.072731137624413e-07,
      "loss": 0.3967,
      "mean_token_accuracy": 0.9044073820114136,
      "num_tokens": 24834324.0,
      "step": 531
    },
    {
      "epoch": 6.734177215189874,
      "grad_norm": 0.5715585350990295,
      "learning_rate": 8.3020184677568e-07,
      "loss": 0.3823,
      "mean_token_accuracy": 0.9075259566307068,
      "num_tokens": 24881558.0,
      "step": 532
    },
    {
      "epoch": 6.746835443037975,
      "grad_norm": 0.5677687525749207,
      "learning_rate": 7.56537393229817e-07,
      "loss": 0.3884,
      "mean_token_accuracy": 0.9045724868774414,
      "num_tokens": 24929386.0,
      "step": 533
    },
    {
      "epoch": 6.759493670886076,
      "grad_norm": 0.5910177826881409,
      "learning_rate": 6.862822837445881e-07,
      "loss": 0.3795,
      "mean_token_accuracy": 0.9077647924423218,
      "num_tokens": 24975723.0,
      "step": 534
    },
    {
      "epoch": 6.772151898734177,
      "grad_norm": 0.5636204481124878,
      "learning_rate": 6.194389318173954e-07,
      "loss": 0.3844,
      "mean_token_accuracy": 0.9066954255104065,
      "num_tokens": 25022655.0,
      "step": 535
    },
    {
      "epoch": 6.784810126582278,
      "grad_norm": 0.5545103549957275,
      "learning_rate": 5.560096337404397e-07,
      "loss": 0.392,
      "mean_token_accuracy": 0.9055996537208557,
      "num_tokens": 25072115.0,
      "step": 536
    },
    {
      "epoch": 6.7974683544303796,
      "grad_norm": 0.5688847303390503,
      "learning_rate": 4.959965685216949e-07,
      "loss": 0.4038,
      "mean_token_accuracy": 0.9030274748802185,
      "num_tokens": 25118852.0,
      "step": 537
    },
    {
      "epoch": 6.810126582278481,
      "grad_norm": 0.5681838989257812,
      "learning_rate": 4.3940179781019055e-07,
      "loss": 0.3654,
      "mean_token_accuracy": 0.9127160906791687,
      "num_tokens": 25163403.0,
      "step": 538
    },
    {
      "epoch": 6.822784810126582,
      "grad_norm": 0.5705924034118652,
      "learning_rate": 3.8622726582514537e-07,
      "loss": 0.41,
      "mean_token_accuracy": 0.904007613658905,
      "num_tokens": 25210752.0,
      "step": 539
    },
    {
      "epoch": 6.8354430379746836,
      "grad_norm": 0.570173978805542,
      "learning_rate": 3.364747992891215e-07,
      "loss": 0.3828,
      "mean_token_accuracy": 0.9091429710388184,
      "num_tokens": 25256206.0,
      "step": 540
    },
    {
      "epoch": 6.848101265822785,
      "grad_norm": 0.575129508972168,
      "learning_rate": 2.901461073653633e-07,
      "loss": 0.3892,
      "mean_token_accuracy": 0.9063740372657776,
      "num_tokens": 25301877.0,
      "step": 541
    },
    {
      "epoch": 6.860759493670886,
      "grad_norm": 0.5604121088981628,
      "learning_rate": 2.472427815989886e-07,
      "loss": 0.3715,
      "mean_token_accuracy": 0.9102283716201782,
      "num_tokens": 25348136.0,
      "step": 542
    },
    {
      "epoch": 6.8734177215189876,
      "grad_norm": 0.5795130729675293,
      "learning_rate": 2.0776629586239936e-07,
      "loss": 0.3988,
      "mean_token_accuracy": 0.90403151512146,
      "num_tokens": 25395403.0,
      "step": 543
    },
    {
      "epoch": 6.886075949367089,
      "grad_norm": 0.5694568157196045,
      "learning_rate": 1.7171800630458868e-07,
      "loss": 0.3567,
      "mean_token_accuracy": 0.915683925151825,
      "num_tokens": 25439053.0,
      "step": 544
    },
    {
      "epoch": 6.89873417721519,
      "grad_norm": 0.556625247001648,
      "learning_rate": 1.3909915130457808e-07,
      "loss": 0.3759,
      "mean_token_accuracy": 0.9094910621643066,
      "num_tokens": 25484781.0,
      "step": 545
    },
    {
      "epoch": 6.911392405063291,
      "grad_norm": 0.5877848267555237,
      "learning_rate": 1.0991085142886271e-07,
      "loss": 0.3996,
      "mean_token_accuracy": 0.9032759070396423,
      "num_tokens": 25529901.0,
      "step": 546
    },
    {
      "epoch": 6.924050632911392,
      "grad_norm": 0.5745137929916382,
      "learning_rate": 8.41541093929199e-08,
      "loss": 0.3932,
      "mean_token_accuracy": 0.9050623774528503,
      "num_tokens": 25578302.0,
      "step": 547
    },
    {
      "epoch": 6.936708860759493,
      "grad_norm": 0.5791630148887634,
      "learning_rate": 6.182981002679223e-08,
      "loss": 0.3789,
      "mean_token_accuracy": 0.907834529876709,
      "num_tokens": 25623589.0,
      "step": 548
    },
    {
      "epoch": 6.949367088607595,
      "grad_norm": 0.6007780432701111,
      "learning_rate": 4.293872024463408e-08,
      "loss": 0.4255,
      "mean_token_accuracy": 0.8964483737945557,
      "num_tokens": 25671040.0,
      "step": 549
    },
    {
      "epoch": 6.962025316455696,
      "grad_norm": 0.5847962498664856,
      "learning_rate": 2.7481489018410523e-08,
      "loss": 0.3983,
      "mean_token_accuracy": 0.9039490222930908,
      "num_tokens": 25718204.0,
      "step": 550
    },
    {
      "epoch": 6.974683544303797,
      "grad_norm": 0.5613832473754883,
      "learning_rate": 1.545864735558178e-08,
      "loss": 0.375,
      "mean_token_accuracy": 0.9110968708992004,
      "num_tokens": 25764993.0,
      "step": 551
    },
    {
      "epoch": 6.987341772151899,
      "grad_norm": 0.5971652269363403,
      "learning_rate": 6.8706082808955855e-09,
      "loss": 0.3968,
      "mean_token_accuracy": 0.9042324423789978,
      "num_tokens": 25810208.0,
      "step": 552
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.5324912667274475,
      "learning_rate": 1.7176668221208225e-09,
      "loss": 0.3848,
      "mean_token_accuracy": 0.9090686440467834,
      "num_tokens": 25859276.0,
      "step": 553
    },
    {
      "epoch": 7.0,
      "step": 553,
      "total_flos": 1.6654859065779814e+18,
      "train_loss": 1.0050470797752602,
      "train_runtime": 3322.7101,
      "train_samples_per_second": 10.534,
      "train_steps_per_second": 0.166
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 553,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6654859065779814e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}