{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008,
      "grad_norm": 34.44804382324219,
      "learning_rate": 1e-05,
      "loss": 13.0101,
      "mean_token_accuracy": 0.4696590006351471,
      "step": 1
    },
    {
      "epoch": 0.016,
      "grad_norm": 30.779788970947266,
      "learning_rate": 2e-05,
      "loss": 12.3851,
      "mean_token_accuracy": 0.47303473204374313,
      "step": 2
    },
    {
      "epoch": 0.024,
      "grad_norm": 29.67559242248535,
      "learning_rate": 3e-05,
      "loss": 12.3488,
      "mean_token_accuracy": 0.49709559231996536,
      "step": 3
    },
    {
      "epoch": 0.032,
      "grad_norm": 26.862010955810547,
      "learning_rate": 4e-05,
      "loss": 11.6596,
      "mean_token_accuracy": 0.5584611147642136,
      "step": 4
    },
    {
      "epoch": 0.04,
      "grad_norm": 22.10072135925293,
      "learning_rate": 5e-05,
      "loss": 10.1384,
      "mean_token_accuracy": 0.5924926251173019,
      "step": 5
    },
    {
      "epoch": 0.048,
      "grad_norm": 20.171361923217773,
      "learning_rate": 4.98989898989899e-05,
      "loss": 9.5421,
      "mean_token_accuracy": 0.5888276249170303,
      "step": 6
    },
    {
      "epoch": 0.056,
      "grad_norm": 16.452842712402344,
      "learning_rate": 4.97979797979798e-05,
      "loss": 8.4344,
      "mean_token_accuracy": 0.632336363196373,
      "step": 7
    },
    {
      "epoch": 0.064,
      "grad_norm": 12.62137222290039,
      "learning_rate": 4.9696969696969694e-05,
      "loss": 7.7758,
      "mean_token_accuracy": 0.6633019298315048,
      "step": 8
    },
    {
      "epoch": 0.072,
      "grad_norm": 11.533007621765137,
      "learning_rate": 4.9595959595959594e-05,
      "loss": 7.5523,
      "mean_token_accuracy": 0.6721473336219788,
      "step": 9
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.251978874206543,
      "learning_rate": 4.94949494949495e-05,
      "loss": 7.5504,
      "mean_token_accuracy": 0.6924590468406677,
      "step": 10
    },
    {
      "epoch": 0.088,
      "grad_norm": 9.099457740783691,
      "learning_rate": 4.93939393939394e-05,
      "loss": 7.1883,
      "mean_token_accuracy": 0.694612979888916,
      "step": 11
    },
    {
      "epoch": 0.096,
      "grad_norm": 9.096614837646484,
      "learning_rate": 4.92929292929293e-05,
      "loss": 6.7714,
      "mean_token_accuracy": 0.7158520817756653,
      "step": 12
    },
    {
      "epoch": 0.104,
      "grad_norm": 8.9071044921875,
      "learning_rate": 4.919191919191919e-05,
      "loss": 6.7582,
      "mean_token_accuracy": 0.7046481072902679,
      "step": 13
    },
    {
      "epoch": 0.112,
      "grad_norm": 8.336892127990723,
      "learning_rate": 4.909090909090909e-05,
      "loss": 6.639,
      "mean_token_accuracy": 0.6894638538360596,
      "step": 14
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.760809421539307,
      "learning_rate": 4.898989898989899e-05,
      "loss": 6.1283,
      "mean_token_accuracy": 0.7323453426361084,
      "step": 15
    },
    {
      "epoch": 0.128,
      "grad_norm": 10.2608003616333,
      "learning_rate": 4.888888888888889e-05,
      "loss": 6.3731,
      "mean_token_accuracy": 0.7066609710454941,
      "step": 16
    },
    {
      "epoch": 0.136,
      "grad_norm": 9.326302528381348,
      "learning_rate": 4.878787878787879e-05,
      "loss": 6.0298,
      "mean_token_accuracy": 0.7226346284151077,
      "step": 17
    },
    {
      "epoch": 0.144,
      "grad_norm": 9.35575008392334,
      "learning_rate": 4.868686868686869e-05,
      "loss": 6.1363,
      "mean_token_accuracy": 0.7228655517101288,
      "step": 18
    },
    {
      "epoch": 0.152,
      "grad_norm": 7.2700018882751465,
      "learning_rate": 4.858585858585859e-05,
      "loss": 6.3529,
      "mean_token_accuracy": 0.7191445082426071,
      "step": 19
    },
    {
      "epoch": 0.16,
      "grad_norm": 7.965806484222412,
      "learning_rate": 4.848484848484849e-05,
      "loss": 6.1908,
      "mean_token_accuracy": 0.7225745767354965,
      "step": 20
    },
    {
      "epoch": 0.168,
      "grad_norm": 9.064255714416504,
      "learning_rate": 4.838383838383839e-05,
      "loss": 6.0033,
      "mean_token_accuracy": 0.7359140962362289,
      "step": 21
    },
    {
      "epoch": 0.176,
      "grad_norm": 6.869060039520264,
      "learning_rate": 4.828282828282829e-05,
      "loss": 5.1424,
      "mean_token_accuracy": 0.7650310397148132,
      "step": 22
    },
    {
      "epoch": 0.184,
      "grad_norm": 7.401363372802734,
      "learning_rate": 4.8181818181818186e-05,
      "loss": 6.2128,
      "mean_token_accuracy": 0.7163093239068985,
      "step": 23
    },
    {
      "epoch": 0.192,
      "grad_norm": 7.535747051239014,
      "learning_rate": 4.808080808080808e-05,
      "loss": 5.561,
      "mean_token_accuracy": 0.7451221346855164,
      "step": 24
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.211135864257812,
      "learning_rate": 4.797979797979798e-05,
      "loss": 5.265,
      "mean_token_accuracy": 0.7609298378229141,
      "step": 25
    },
    {
      "epoch": 0.208,
      "grad_norm": 6.720511436462402,
      "learning_rate": 4.787878787878788e-05,
      "loss": 5.3244,
      "mean_token_accuracy": 0.7525999993085861,
      "step": 26
    },
    {
      "epoch": 0.216,
      "grad_norm": 9.114042282104492,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 5.6173,
      "mean_token_accuracy": 0.7520534992218018,
      "step": 27
    },
    {
      "epoch": 0.224,
      "grad_norm": 7.770754337310791,
      "learning_rate": 4.7676767676767684e-05,
      "loss": 5.5093,
      "mean_token_accuracy": 0.7558625787496567,
      "step": 28
    },
    {
      "epoch": 0.232,
      "grad_norm": 7.749612808227539,
      "learning_rate": 4.7575757575757576e-05,
      "loss": 5.5386,
      "mean_token_accuracy": 0.7553819268941879,
      "step": 29
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.486739635467529,
      "learning_rate": 4.7474747474747476e-05,
      "loss": 5.73,
      "mean_token_accuracy": 0.732807844877243,
      "step": 30
    },
    {
      "epoch": 0.248,
      "grad_norm": 7.474062919616699,
      "learning_rate": 4.7373737373737375e-05,
      "loss": 6.0216,
      "mean_token_accuracy": 0.742279127240181,
      "step": 31
    },
    {
      "epoch": 0.256,
      "grad_norm": 7.018927574157715,
      "learning_rate": 4.7272727272727275e-05,
      "loss": 5.0907,
      "mean_token_accuracy": 0.7746435850858688,
      "step": 32
    },
    {
      "epoch": 0.264,
      "grad_norm": 6.719446659088135,
      "learning_rate": 4.7171717171717174e-05,
      "loss": 4.9967,
      "mean_token_accuracy": 0.7717019468545914,
      "step": 33
    },
    {
      "epoch": 0.272,
      "grad_norm": 6.464803695678711,
      "learning_rate": 4.7070707070707074e-05,
      "loss": 5.2966,
      "mean_token_accuracy": 0.7656355202198029,
      "step": 34
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.2371320724487305,
      "learning_rate": 4.696969696969697e-05,
      "loss": 5.6333,
      "mean_token_accuracy": 0.7543937265872955,
      "step": 35
    },
    {
      "epoch": 0.288,
      "grad_norm": 6.18785285949707,
      "learning_rate": 4.686868686868687e-05,
      "loss": 5.3846,
      "mean_token_accuracy": 0.7626298367977142,
      "step": 36
    },
    {
      "epoch": 0.296,
      "grad_norm": 5.636417865753174,
      "learning_rate": 4.676767676767677e-05,
      "loss": 4.7779,
      "mean_token_accuracy": 0.796937882900238,
      "step": 37
    },
    {
      "epoch": 0.304,
      "grad_norm": 6.951849460601807,
      "learning_rate": 4.666666666666667e-05,
      "loss": 5.3062,
      "mean_token_accuracy": 0.7514433115720749,
      "step": 38
    },
    {
      "epoch": 0.312,
      "grad_norm": 7.596467018127441,
      "learning_rate": 4.656565656565657e-05,
      "loss": 5.4916,
      "mean_token_accuracy": 0.738706648349762,
      "step": 39
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.036402225494385,
      "learning_rate": 4.6464646464646464e-05,
      "loss": 5.4346,
      "mean_token_accuracy": 0.7636198997497559,
      "step": 40
    },
    {
      "epoch": 0.328,
      "grad_norm": 7.065268039703369,
      "learning_rate": 4.636363636363636e-05,
      "loss": 4.9636,
      "mean_token_accuracy": 0.76950503885746,
      "step": 41
    },
    {
      "epoch": 0.336,
      "grad_norm": 5.93984842300415,
      "learning_rate": 4.626262626262626e-05,
      "loss": 4.9883,
      "mean_token_accuracy": 0.7728704810142517,
      "step": 42
    },
    {
      "epoch": 0.344,
      "grad_norm": 6.929603576660156,
      "learning_rate": 4.616161616161616e-05,
      "loss": 4.6927,
      "mean_token_accuracy": 0.7756397873163223,
      "step": 43
    },
    {
      "epoch": 0.352,
      "grad_norm": 7.2813496589660645,
      "learning_rate": 4.606060606060607e-05,
      "loss": 4.761,
      "mean_token_accuracy": 0.784713864326477,
      "step": 44
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.584592342376709,
      "learning_rate": 4.595959595959596e-05,
      "loss": 4.7331,
      "mean_token_accuracy": 0.787657305598259,
      "step": 45
    },
    {
      "epoch": 0.368,
      "grad_norm": 7.5701751708984375,
      "learning_rate": 4.585858585858586e-05,
      "loss": 4.1914,
      "mean_token_accuracy": 0.796615332365036,
      "step": 46
    },
    {
      "epoch": 0.376,
      "grad_norm": 6.217738628387451,
      "learning_rate": 4.575757575757576e-05,
      "loss": 4.6948,
      "mean_token_accuracy": 0.7924428284168243,
      "step": 47
    },
    {
      "epoch": 0.384,
      "grad_norm": 6.909704685211182,
      "learning_rate": 4.565656565656566e-05,
      "loss": 5.0747,
      "mean_token_accuracy": 0.7648352980613708,
      "step": 48
    },
    {
      "epoch": 0.392,
      "grad_norm": 7.039934158325195,
      "learning_rate": 4.555555555555556e-05,
      "loss": 5.168,
      "mean_token_accuracy": 0.7628057301044464,
      "step": 49
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.007484436035156,
      "learning_rate": 4.545454545454546e-05,
      "loss": 5.5139,
      "mean_token_accuracy": 0.7507320195436478,
      "step": 50
    },
    {
      "epoch": 0.408,
      "grad_norm": 7.163206100463867,
      "learning_rate": 4.535353535353535e-05,
      "loss": 4.6769,
      "mean_token_accuracy": 0.789847657084465,
      "step": 51
    },
    {
      "epoch": 0.416,
      "grad_norm": 6.482788562774658,
      "learning_rate": 4.525252525252526e-05,
      "loss": 4.0523,
      "mean_token_accuracy": 0.8091136366128922,
      "step": 52
    },
    {
      "epoch": 0.424,
      "grad_norm": 6.3435564041137695,
      "learning_rate": 4.515151515151516e-05,
      "loss": 5.0074,
      "mean_token_accuracy": 0.7676153928041458,
      "step": 53
    },
    {
      "epoch": 0.432,
      "grad_norm": 6.715310096740723,
      "learning_rate": 4.5050505050505056e-05,
      "loss": 4.7588,
      "mean_token_accuracy": 0.7704059928655624,
      "step": 54
    },
    {
      "epoch": 0.44,
      "grad_norm": 6.287456512451172,
      "learning_rate": 4.494949494949495e-05,
      "loss": 5.0595,
      "mean_token_accuracy": 0.7789322882890701,
      "step": 55
    },
    {
      "epoch": 0.448,
      "grad_norm": 6.842898845672607,
      "learning_rate": 4.484848484848485e-05,
      "loss": 4.3176,
      "mean_token_accuracy": 0.7993331402540207,
      "step": 56
    },
    {
      "epoch": 0.456,
      "grad_norm": 5.992519378662109,
      "learning_rate": 4.474747474747475e-05,
      "loss": 5.0297,
      "mean_token_accuracy": 0.779092013835907,
      "step": 57
    },
    {
      "epoch": 0.464,
      "grad_norm": 6.67594051361084,
      "learning_rate": 4.464646464646465e-05,
      "loss": 4.1185,
      "mean_token_accuracy": 0.7971874326467514,
      "step": 58
    },
    {
      "epoch": 0.472,
      "grad_norm": 6.712916374206543,
      "learning_rate": 4.454545454545455e-05,
      "loss": 4.6345,
      "mean_token_accuracy": 0.7811515778303146,
      "step": 59
    },
    {
      "epoch": 0.48,
      "grad_norm": 6.204624652862549,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 4.5869,
      "mean_token_accuracy": 0.8028187602758408,
      "step": 60
    },
    {
      "epoch": 0.488,
      "grad_norm": 7.798375129699707,
      "learning_rate": 4.4343434343434346e-05,
      "loss": 5.2535,
      "mean_token_accuracy": 0.7713681906461716,
      "step": 61
    },
    {
      "epoch": 0.496,
      "grad_norm": 6.647665500640869,
      "learning_rate": 4.4242424242424246e-05,
      "loss": 4.1662,
      "mean_token_accuracy": 0.809452161192894,
      "step": 62
    },
    {
      "epoch": 0.504,
      "grad_norm": 7.158292770385742,
      "learning_rate": 4.4141414141414145e-05,
      "loss": 4.5907,
      "mean_token_accuracy": 0.7843180149793625,
      "step": 63
    },
    {
      "epoch": 0.512,
      "grad_norm": 6.763492107391357,
      "learning_rate": 4.4040404040404044e-05,
      "loss": 4.6913,
      "mean_token_accuracy": 0.7733302861452103,
      "step": 64
    },
    {
      "epoch": 0.52,
      "grad_norm": 6.745001792907715,
      "learning_rate": 4.3939393939393944e-05,
      "loss": 4.7438,
      "mean_token_accuracy": 0.7817106246948242,
      "step": 65
    },
    {
      "epoch": 0.528,
      "grad_norm": 5.92717981338501,
      "learning_rate": 4.383838383838384e-05,
      "loss": 4.3055,
      "mean_token_accuracy": 0.7931149005889893,
      "step": 66
    },
    {
      "epoch": 0.536,
      "grad_norm": 7.291048049926758,
      "learning_rate": 4.3737373737373736e-05,
      "loss": 4.7281,
      "mean_token_accuracy": 0.7924042791128159,
      "step": 67
    },
    {
      "epoch": 0.544,
      "grad_norm": 6.6944899559021,
      "learning_rate": 4.3636363636363636e-05,
      "loss": 5.4566,
      "mean_token_accuracy": 0.7465656846761703,
      "step": 68
    },
    {
      "epoch": 0.552,
      "grad_norm": 6.43382453918457,
      "learning_rate": 4.3535353535353535e-05,
      "loss": 5.1924,
      "mean_token_accuracy": 0.765620231628418,
      "step": 69
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.896528244018555,
      "learning_rate": 4.343434343434344e-05,
      "loss": 4.535,
      "mean_token_accuracy": 0.780782625079155,
      "step": 70
    },
    {
      "epoch": 0.568,
      "grad_norm": 7.0929059982299805,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 4.7798,
      "mean_token_accuracy": 0.7812370210886002,
      "step": 71
    },
    {
      "epoch": 0.576,
      "grad_norm": 6.4265031814575195,
      "learning_rate": 4.3232323232323234e-05,
      "loss": 4.0221,
      "mean_token_accuracy": 0.8055804222822189,
      "step": 72
    },
    {
      "epoch": 0.584,
      "grad_norm": 6.464663982391357,
      "learning_rate": 4.313131313131313e-05,
      "loss": 5.3328,
      "mean_token_accuracy": 0.7507916688919067,
      "step": 73
    },
    {
      "epoch": 0.592,
      "grad_norm": 6.163589000701904,
      "learning_rate": 4.303030303030303e-05,
      "loss": 3.8647,
      "mean_token_accuracy": 0.8181672990322113,
      "step": 74
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.442709445953369,
      "learning_rate": 4.292929292929293e-05,
      "loss": 4.1589,
      "mean_token_accuracy": 0.7894054055213928,
      "step": 75
    },
    {
      "epoch": 0.608,
      "grad_norm": 6.7371673583984375,
      "learning_rate": 4.282828282828283e-05,
      "loss": 4.4978,
      "mean_token_accuracy": 0.7749454975128174,
      "step": 76
    },
    {
      "epoch": 0.616,
      "grad_norm": 6.681241035461426,
      "learning_rate": 4.2727272727272724e-05,
      "loss": 4.3404,
      "mean_token_accuracy": 0.8082751482725143,
      "step": 77
    },
    {
      "epoch": 0.624,
      "grad_norm": 6.194783687591553,
      "learning_rate": 4.262626262626263e-05,
      "loss": 4.1043,
      "mean_token_accuracy": 0.7980503439903259,
      "step": 78
    },
    {
      "epoch": 0.632,
      "grad_norm": 6.758218288421631,
      "learning_rate": 4.252525252525253e-05,
      "loss": 4.4278,
      "mean_token_accuracy": 0.7885774075984955,
      "step": 79
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.5866546630859375,
      "learning_rate": 4.242424242424243e-05,
      "loss": 4.227,
      "mean_token_accuracy": 0.7932698577642441,
      "step": 80
    },
    {
      "epoch": 0.648,
      "grad_norm": 7.192435264587402,
      "learning_rate": 4.232323232323233e-05,
      "loss": 4.0345,
      "mean_token_accuracy": 0.8199726790189743,
      "step": 81
    },
    {
      "epoch": 0.656,
      "grad_norm": 5.501450061798096,
      "learning_rate": 4.222222222222222e-05,
      "loss": 4.2629,
      "mean_token_accuracy": 0.7981265485286713,
      "step": 82
    },
    {
      "epoch": 0.664,
      "grad_norm": 6.11761474609375,
      "learning_rate": 4.212121212121212e-05,
      "loss": 4.2,
      "mean_token_accuracy": 0.8049240857362747,
      "step": 83
    },
    {
      "epoch": 0.672,
      "grad_norm": 6.632277011871338,
      "learning_rate": 4.202020202020202e-05,
      "loss": 4.5717,
      "mean_token_accuracy": 0.784144937992096,
      "step": 84
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.618972301483154,
      "learning_rate": 4.191919191919192e-05,
      "loss": 4.1424,
      "mean_token_accuracy": 0.8174345791339874,
      "step": 85
    },
    {
      "epoch": 0.688,
      "grad_norm": 6.348287105560303,
      "learning_rate": 4.181818181818182e-05,
      "loss": 3.9289,
      "mean_token_accuracy": 0.8103939890861511,
      "step": 86
    },
    {
      "epoch": 0.696,
      "grad_norm": 6.326577186584473,
      "learning_rate": 4.171717171717172e-05,
      "loss": 4.4158,
      "mean_token_accuracy": 0.7934867739677429,
      "step": 87
    },
    {
      "epoch": 0.704,
      "grad_norm": 5.938681602478027,
      "learning_rate": 4.161616161616162e-05,
      "loss": 4.3607,
      "mean_token_accuracy": 0.8023031949996948,
      "step": 88
    },
    {
      "epoch": 0.712,
      "grad_norm": 6.043774604797363,
      "learning_rate": 4.151515151515152e-05,
      "loss": 4.8988,
      "mean_token_accuracy": 0.7769377380609512,
      "step": 89
    },
    {
      "epoch": 0.72,
      "grad_norm": 6.513429641723633,
      "learning_rate": 4.141414141414142e-05,
      "loss": 4.469,
      "mean_token_accuracy": 0.7987037748098373,
      "step": 90
    },
    {
      "epoch": 0.728,
      "grad_norm": 6.306874752044678,
      "learning_rate": 4.131313131313132e-05,
      "loss": 3.8623,
      "mean_token_accuracy": 0.8089967370033264,
      "step": 91
    },
    {
      "epoch": 0.736,
      "grad_norm": 6.43092155456543,
      "learning_rate": 4.1212121212121216e-05,
      "loss": 4.1141,
      "mean_token_accuracy": 0.8077200204133987,
      "step": 92
    },
    {
      "epoch": 0.744,
      "grad_norm": 5.928807735443115,
      "learning_rate": 4.111111111111111e-05,
      "loss": 4.379,
      "mean_token_accuracy": 0.7994053959846497,
      "step": 93
    },
    {
      "epoch": 0.752,
      "grad_norm": 6.193979740142822,
      "learning_rate": 4.101010101010101e-05,
      "loss": 4.0719,
      "mean_token_accuracy": 0.7951067984104156,
      "step": 94
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.613356113433838,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 3.9982,
      "mean_token_accuracy": 0.8044578582048416,
      "step": 95
    },
    {
      "epoch": 0.768,
      "grad_norm": 5.828372478485107,
      "learning_rate": 4.0808080808080814e-05,
      "loss": 3.9207,
      "mean_token_accuracy": 0.804252415895462,
      "step": 96
    },
    {
      "epoch": 0.776,
      "grad_norm": 6.8737993240356445,
      "learning_rate": 4.070707070707071e-05,
      "loss": 4.018,
      "mean_token_accuracy": 0.7971819192171097,
      "step": 97
    },
    {
      "epoch": 0.784,
      "grad_norm": 6.014230251312256,
      "learning_rate": 4.0606060606060606e-05,
      "loss": 3.9449,
      "mean_token_accuracy": 0.8023800402879715,
      "step": 98
    },
    {
      "epoch": 0.792,
      "grad_norm": 6.424180030822754,
      "learning_rate": 4.0505050505050506e-05,
      "loss": 4.133,
      "mean_token_accuracy": 0.7898548692464828,
      "step": 99
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.941647529602051,
      "learning_rate": 4.0404040404040405e-05,
      "loss": 4.4152,
      "mean_token_accuracy": 0.7944744974374771,
      "step": 100
    },
    {
      "epoch": 0.808,
      "grad_norm": 5.532174110412598,
      "learning_rate": 4.0303030303030305e-05,
      "loss": 3.9543,
      "mean_token_accuracy": 0.8119575679302216,
      "step": 101
    },
    {
      "epoch": 0.816,
      "grad_norm": 6.766019344329834,
      "learning_rate": 4.0202020202020204e-05,
      "loss": 3.8357,
      "mean_token_accuracy": 0.8240242004394531,
      "step": 102
    },
    {
      "epoch": 0.824,
      "grad_norm": 5.829742908477783,
      "learning_rate": 4.01010101010101e-05,
      "loss": 4.1291,
      "mean_token_accuracy": 0.798798069357872,
      "step": 103
    },
    {
      "epoch": 0.832,
      "grad_norm": 6.3326416015625,
      "learning_rate": 4e-05,
      "loss": 4.5605,
      "mean_token_accuracy": 0.7854688614606857,
      "step": 104
    },
    {
      "epoch": 0.84,
      "grad_norm": 5.9824042320251465,
      "learning_rate": 3.98989898989899e-05,
      "loss": 4.0588,
      "mean_token_accuracy": 0.8019276112318039,
      "step": 105
    },
    {
      "epoch": 0.848,
      "grad_norm": 6.245088577270508,
      "learning_rate": 3.97979797979798e-05,
      "loss": 3.9216,
      "mean_token_accuracy": 0.8176022469997406,
      "step": 106
    },
    {
      "epoch": 0.856,
      "grad_norm": 6.343690395355225,
      "learning_rate": 3.96969696969697e-05,
      "loss": 4.5661,
      "mean_token_accuracy": 0.7900134176015854,
      "step": 107
    },
    {
      "epoch": 0.864,
      "grad_norm": 6.176941394805908,
      "learning_rate": 3.9595959595959594e-05,
      "loss": 4.3509,
      "mean_token_accuracy": 0.7881277054548264,
      "step": 108
    },
    {
      "epoch": 0.872,
      "grad_norm": 5.415067672729492,
      "learning_rate": 3.9494949494949494e-05,
      "loss": 3.9209,
      "mean_token_accuracy": 0.8084693402051926,
      "step": 109
    },
    {
      "epoch": 0.88,
      "grad_norm": 6.951507568359375,
      "learning_rate": 3.939393939393939e-05,
      "loss": 3.8214,
      "mean_token_accuracy": 0.8184851258993149,
      "step": 110
    },
    {
      "epoch": 0.888,
      "grad_norm": 6.3500213623046875,
      "learning_rate": 3.929292929292929e-05,
      "loss": 4.445,
      "mean_token_accuracy": 0.7920354455709457,
      "step": 111
    },
    {
      "epoch": 0.896,
      "grad_norm": 6.108852386474609,
      "learning_rate": 3.91919191919192e-05,
      "loss": 3.9563,
      "mean_token_accuracy": 0.8090476542711258,
      "step": 112
    },
    {
      "epoch": 0.904,
      "grad_norm": 5.749330043792725,
      "learning_rate": 3.909090909090909e-05,
      "loss": 3.7184,
      "mean_token_accuracy": 0.8214946985244751,
      "step": 113
    },
    {
      "epoch": 0.912,
      "grad_norm": 6.137975692749023,
      "learning_rate": 3.898989898989899e-05,
      "loss": 3.8174,
      "mean_token_accuracy": 0.8351726979017258,
      "step": 114
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.827854156494141,
      "learning_rate": 3.888888888888889e-05,
      "loss": 4.1516,
      "mean_token_accuracy": 0.8008674532175064,
      "step": 115
    },
    {
      "epoch": 0.928,
      "grad_norm": 6.231594562530518,
      "learning_rate": 3.878787878787879e-05,
      "loss": 3.9401,
      "mean_token_accuracy": 0.8067153543233871,
      "step": 116
    },
    {
      "epoch": 0.936,
      "grad_norm": 6.3449578285217285,
      "learning_rate": 3.868686868686869e-05,
      "loss": 4.5777,
      "mean_token_accuracy": 0.7864043414592743,
      "step": 117
    },
    {
      "epoch": 0.944,
      "grad_norm": 5.758909702301025,
      "learning_rate": 3.858585858585859e-05,
      "loss": 3.9313,
      "mean_token_accuracy": 0.8202812224626541,
      "step": 118
    },
    {
      "epoch": 0.952,
      "grad_norm": 5.801164627075195,
      "learning_rate": 3.848484848484848e-05,
      "loss": 3.4612,
      "mean_token_accuracy": 0.8463838249444962,
      "step": 119
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.331445217132568,
      "learning_rate": 3.838383838383838e-05,
      "loss": 4.1784,
      "mean_token_accuracy": 0.8084011971950531,
      "step": 120
    },
    {
      "epoch": 0.968,
      "grad_norm": 5.821787357330322,
      "learning_rate": 3.828282828282829e-05,
      "loss": 4.0625,
      "mean_token_accuracy": 0.8006434291601181,
      "step": 121
    },
    {
      "epoch": 0.976,
      "grad_norm": 6.507475852966309,
      "learning_rate": 3.818181818181819e-05,
      "loss": 3.7078,
      "mean_token_accuracy": 0.8223373293876648,
      "step": 122
    },
    {
      "epoch": 0.984,
      "grad_norm": 5.8519511222839355,
      "learning_rate": 3.8080808080808087e-05,
      "loss": 3.8093,
      "mean_token_accuracy": 0.8161063343286514,
      "step": 123
    },
    {
      "epoch": 0.992,
      "grad_norm": 6.515493869781494,
      "learning_rate": 3.797979797979798e-05,
      "loss": 3.8881,
      "mean_token_accuracy": 0.8130738884210587,
      "step": 124
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.974099159240723,
      "learning_rate": 3.787878787878788e-05,
      "loss": 3.8128,
      "mean_token_accuracy": 0.817224845290184,
      "step": 125
    },
    {
      "epoch": 1.008,
      "grad_norm": 5.158608436584473,
      "learning_rate": 3.777777777777778e-05,
      "loss": 3.5407,
      "mean_token_accuracy": 0.8298159688711166,
      "step": 126
    },
    {
      "epoch": 1.016,
      "grad_norm": 5.91457462310791,
      "learning_rate": 3.767676767676768e-05,
      "loss": 3.6346,
      "mean_token_accuracy": 0.8155378848314285,
      "step": 127
    },
    {
      "epoch": 1.024,
      "grad_norm": 5.550519943237305,
      "learning_rate": 3.757575757575758e-05,
      "loss": 3.8566,
      "mean_token_accuracy": 0.8141728788614273,
      "step": 128
    },
    {
      "epoch": 1.032,
      "grad_norm": 5.610849380493164,
      "learning_rate": 3.747474747474748e-05,
      "loss": 3.785,
      "mean_token_accuracy": 0.811878427863121,
      "step": 129
    },
    {
      "epoch": 1.04,
      "grad_norm": 6.2908854484558105,
      "learning_rate": 3.7373737373737376e-05,
      "loss": 3.0106,
      "mean_token_accuracy": 0.8458054810762405,
      "step": 130
    },
    {
      "epoch": 1.048,
      "grad_norm": 5.267597675323486,
      "learning_rate": 3.7272727272727276e-05,
      "loss": 3.4967,
      "mean_token_accuracy": 0.8339269608259201,
      "step": 131
    },
    {
      "epoch": 1.056,
      "grad_norm": 5.37116003036499,
      "learning_rate": 3.7171717171717175e-05,
      "loss": 3.4903,
      "mean_token_accuracy": 0.8220857381820679,
      "step": 132
    },
    {
      "epoch": 1.064,
      "grad_norm": 5.198507785797119,
      "learning_rate": 3.7070707070707075e-05,
      "loss": 2.9143,
      "mean_token_accuracy": 0.8564625233411789,
      "step": 133
    },
    {
      "epoch": 1.072,
      "grad_norm": 5.638608932495117,
      "learning_rate": 3.6969696969696974e-05,
      "loss": 3.6283,
      "mean_token_accuracy": 0.8230338096618652,
      "step": 134
    },
    {
      "epoch": 1.08,
      "grad_norm": 5.533100605010986,
      "learning_rate": 3.686868686868687e-05,
      "loss": 3.3163,
      "mean_token_accuracy": 0.8364241421222687,
      "step": 135
    },
    {
      "epoch": 1.088,
      "grad_norm": 5.794349670410156,
      "learning_rate": 3.6767676767676766e-05,
      "loss": 3.4776,
      "mean_token_accuracy": 0.822014644742012,
      "step": 136
    },
    {
      "epoch": 1.096,
      "grad_norm": 6.230714797973633,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 3.9834,
      "mean_token_accuracy": 0.808947280049324,
      "step": 137
    },
    {
      "epoch": 1.104,
      "grad_norm": 5.761725902557373,
      "learning_rate": 3.656565656565657e-05,
      "loss": 3.5205,
      "mean_token_accuracy": 0.8236398249864578,
      "step": 138
    },
    {
      "epoch": 1.112,
      "grad_norm": 5.620357990264893,
      "learning_rate": 3.6464646464646465e-05,
      "loss": 3.7957,
      "mean_token_accuracy": 0.8184810876846313,
      "step": 139
    },
    {
      "epoch": 1.12,
      "grad_norm": 5.732111930847168,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 3.3031,
      "mean_token_accuracy": 0.8268382251262665,
      "step": 140
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 5.512340545654297,
      "learning_rate": 3.6262626262626264e-05,
      "loss": 3.4623,
      "mean_token_accuracy": 0.8173917531967163,
      "step": 141
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 5.278458118438721,
      "learning_rate": 3.616161616161616e-05,
      "loss": 3.6187,
      "mean_token_accuracy": 0.8212647438049316,
      "step": 142
    },
    {
      "epoch": 1.144,
      "grad_norm": 6.000738620758057,
      "learning_rate": 3.606060606060606e-05,
      "loss": 3.4098,
      "mean_token_accuracy": 0.8118827193975449,
      "step": 143
    },
    {
      "epoch": 1.152,
      "grad_norm": 5.230849266052246,
      "learning_rate": 3.595959595959596e-05,
      "loss": 3.3312,
      "mean_token_accuracy": 0.8291827887296677,
      "step": 144
    },
    {
      "epoch": 1.16,
      "grad_norm": 6.520547389984131,
      "learning_rate": 3.5858585858585855e-05,
      "loss": 2.9547,
      "mean_token_accuracy": 0.8410928547382355,
      "step": 145
    },
    {
      "epoch": 1.168,
      "grad_norm": 6.905513286590576,
      "learning_rate": 3.575757575757576e-05,
      "loss": 3.3843,
      "mean_token_accuracy": 0.8292149305343628,
      "step": 146
    },
    {
      "epoch": 1.176,
      "grad_norm": 5.25789213180542,
      "learning_rate": 3.565656565656566e-05,
      "loss": 3.0306,
      "mean_token_accuracy": 0.8420234769582748,
      "step": 147
    },
    {
      "epoch": 1.184,
      "grad_norm": 6.301482677459717,
      "learning_rate": 3.555555555555556e-05,
      "loss": 3.563,
      "mean_token_accuracy": 0.816611647605896,
      "step": 148
    },
    {
      "epoch": 1.192,
      "grad_norm": 5.18954610824585,
      "learning_rate": 3.545454545454546e-05,
      "loss": 3.3712,
      "mean_token_accuracy": 0.8429348319768906,
      "step": 149
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.924179553985596,
      "learning_rate": 3.535353535353535e-05,
      "loss": 3.0173,
      "mean_token_accuracy": 0.8484609872102737,
      "step": 150
    },
    {
      "epoch": 1.208,
      "grad_norm": 6.125277996063232,
      "learning_rate": 3.525252525252525e-05,
      "loss": 4.017,
      "mean_token_accuracy": 0.8024349808692932,
      "step": 151
    },
    {
      "epoch": 1.216,
      "grad_norm": 6.013985633850098,
      "learning_rate": 3.515151515151515e-05,
      "loss": 3.1488,
      "mean_token_accuracy": 0.8332614004611969,
      "step": 152
    },
    {
      "epoch": 1.224,
      "grad_norm": 5.863841533660889,
      "learning_rate": 3.505050505050505e-05,
      "loss": 3.4748,
      "mean_token_accuracy": 0.8309570550918579,
      "step": 153
    },
    {
      "epoch": 1.232,
      "grad_norm": 5.565553665161133,
      "learning_rate": 3.494949494949495e-05,
      "loss": 3.4478,
      "mean_token_accuracy": 0.8285814672708511,
      "step": 154
    },
    {
      "epoch": 1.24,
      "grad_norm": 5.7123565673828125,
      "learning_rate": 3.484848484848485e-05,
      "loss": 3.1585,
      "mean_token_accuracy": 0.8400322198867798,
      "step": 155
    },
    {
      "epoch": 1.248,
      "grad_norm": 5.499763488769531,
      "learning_rate": 3.474747474747475e-05,
      "loss": 2.9475,
      "mean_token_accuracy": 0.8378583043813705,
      "step": 156
    },
    {
      "epoch": 1.256,
      "grad_norm": 5.909451961517334,
      "learning_rate": 3.464646464646465e-05,
      "loss": 3.0639,
      "mean_token_accuracy": 0.8469367325305939,
      "step": 157
    },
    {
      "epoch": 1.264,
      "grad_norm": 6.213925838470459,
      "learning_rate": 3.454545454545455e-05,
      "loss": 3.7884,
      "mean_token_accuracy": 0.8231126517057419,
      "step": 158
    },
    {
      "epoch": 1.272,
      "grad_norm": 5.21215295791626,
      "learning_rate": 3.444444444444445e-05,
      "loss": 3.4636,
      "mean_token_accuracy": 0.8168640285730362,
      "step": 159
    },
    {
      "epoch": 1.28,
      "grad_norm": 5.759207725524902,
      "learning_rate": 3.434343434343435e-05,
      "loss": 3.2428,
      "mean_token_accuracy": 0.8419955372810364,
      "step": 160
    },
    {
      "epoch": 1.288,
      "grad_norm": 5.337326526641846,
      "learning_rate": 3.424242424242424e-05,
      "loss": 3.5007,
      "mean_token_accuracy": 0.826006218791008,
      "step": 161
    },
    {
      "epoch": 1.296,
      "grad_norm": 5.429465293884277,
      "learning_rate": 3.414141414141414e-05,
      "loss": 3.1902,
      "mean_token_accuracy": 0.8436289280653,
      "step": 162
    },
    {
      "epoch": 1.304,
      "grad_norm": 5.5671234130859375,
      "learning_rate": 3.4040404040404045e-05,
      "loss": 3.1885,
      "mean_token_accuracy": 0.837678074836731,
      "step": 163
    },
    {
      "epoch": 1.312,
      "grad_norm": 5.217177391052246,
      "learning_rate": 3.3939393939393945e-05,
      "loss": 3.0344,
      "mean_token_accuracy": 0.8463016897439957,
      "step": 164
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.01118803024292,
      "learning_rate": 3.3838383838383844e-05,
      "loss": 3.6496,
      "mean_token_accuracy": 0.8142215311527252,
      "step": 165
    },
    {
      "epoch": 1.328,
      "grad_norm": 5.723564147949219,
      "learning_rate": 3.373737373737374e-05,
      "loss": 3.1859,
      "mean_token_accuracy": 0.8320183008909225,
      "step": 166
    },
    {
      "epoch": 1.336,
      "grad_norm": 5.838883399963379,
      "learning_rate": 3.3636363636363636e-05,
      "loss": 3.4536,
      "mean_token_accuracy": 0.8264990895986557,
      "step": 167
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 5.59492301940918,
      "learning_rate": 3.3535353535353536e-05,
      "loss": 3.1706,
      "mean_token_accuracy": 0.8360193520784378,
      "step": 168
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 5.386695861816406,
      "learning_rate": 3.3434343434343435e-05,
      "loss": 2.8613,
      "mean_token_accuracy": 0.8595142513513565,
      "step": 169
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 5.8718791007995605,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 3.405,
      "mean_token_accuracy": 0.8177822679281235,
      "step": 170
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 5.533784866333008,
      "learning_rate": 3.3232323232323234e-05,
      "loss": 2.7535,
      "mean_token_accuracy": 0.847559779882431,
      "step": 171
    },
    {
      "epoch": 1.376,
      "grad_norm": 6.221394062042236,
      "learning_rate": 3.3131313131313134e-05,
      "loss": 3.2758,
      "mean_token_accuracy": 0.8370038121938705,
      "step": 172
    },
    {
      "epoch": 1.384,
      "grad_norm": 5.768967628479004,
      "learning_rate": 3.303030303030303e-05,
      "loss": 3.144,
      "mean_token_accuracy": 0.8520988523960114,
      "step": 173
    },
    {
      "epoch": 1.392,
      "grad_norm": 5.794501781463623,
      "learning_rate": 3.292929292929293e-05,
      "loss": 3.294,
      "mean_token_accuracy": 0.8296584039926529,
      "step": 174
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.306114196777344,
      "learning_rate": 3.282828282828283e-05,
      "loss": 3.8122,
      "mean_token_accuracy": 0.8015681505203247,
      "step": 175
    },
    {
      "epoch": 1.408,
      "grad_norm": 5.309399127960205,
      "learning_rate": 3.272727272727273e-05,
      "loss": 3.5656,
      "mean_token_accuracy": 0.8148599863052368,
      "step": 176
    },
    {
      "epoch": 1.416,
      "grad_norm": 6.384012699127197,
      "learning_rate": 3.2626262626262624e-05,
      "loss": 3.3465,
      "mean_token_accuracy": 0.8153296113014221,
      "step": 177
    },
    {
      "epoch": 1.424,
      "grad_norm": 5.76847505569458,
      "learning_rate": 3.2525252525252524e-05,
      "loss": 3.5117,
      "mean_token_accuracy": 0.8243328183889389,
      "step": 178
    },
    {
      "epoch": 1.432,
      "grad_norm": 5.956698417663574,
      "learning_rate": 3.2424242424242423e-05,
      "loss": 3.3246,
      "mean_token_accuracy": 0.8309469819068909,
      "step": 179
    },
    {
      "epoch": 1.44,
      "grad_norm": 5.436334133148193,
      "learning_rate": 3.232323232323233e-05,
      "loss": 3.3948,
      "mean_token_accuracy": 0.8304746299982071,
      "step": 180
    },
    {
      "epoch": 1.448,
      "grad_norm": 6.713822364807129,
      "learning_rate": 3.222222222222223e-05,
      "loss": 3.7249,
      "mean_token_accuracy": 0.8197664022445679,
      "step": 181
    },
    {
      "epoch": 1.456,
      "grad_norm": 5.69707727432251,
      "learning_rate": 3.212121212121212e-05,
      "loss": 3.4366,
      "mean_token_accuracy": 0.8290866911411285,
      "step": 182
    },
    {
      "epoch": 1.464,
      "grad_norm": 5.674050807952881,
      "learning_rate": 3.202020202020202e-05,
      "loss": 3.4487,
      "mean_token_accuracy": 0.8316894471645355,
      "step": 183
    },
    {
      "epoch": 1.472,
      "grad_norm": 6.27432918548584,
      "learning_rate": 3.191919191919192e-05,
      "loss": 3.7579,
      "mean_token_accuracy": 0.8203246891498566,
      "step": 184
    },
    {
      "epoch": 1.48,
      "grad_norm": 6.963736534118652,
      "learning_rate": 3.181818181818182e-05,
      "loss": 3.3907,
      "mean_token_accuracy": 0.8244215101003647,
      "step": 185
    },
    {
      "epoch": 1.488,
      "grad_norm": 6.205438137054443,
      "learning_rate": 3.171717171717172e-05,
      "loss": 3.2918,
      "mean_token_accuracy": 0.8321795910596848,
      "step": 186
    },
    {
      "epoch": 1.496,
      "grad_norm": 6.026015281677246,
      "learning_rate": 3.161616161616161e-05,
      "loss": 3.7184,
      "mean_token_accuracy": 0.8272889703512192,
      "step": 187
    },
    {
      "epoch": 1.504,
      "grad_norm": 5.637466907501221,
      "learning_rate": 3.151515151515151e-05,
      "loss": 2.5835,
      "mean_token_accuracy": 0.8698715269565582,
      "step": 188
    },
    {
      "epoch": 1.512,
      "grad_norm": 6.474348545074463,
      "learning_rate": 3.141414141414142e-05,
      "loss": 3.2364,
      "mean_token_accuracy": 0.8249330222606659,
      "step": 189
    },
    {
      "epoch": 1.52,
      "grad_norm": 6.4876885414123535,
      "learning_rate": 3.131313131313132e-05,
      "loss": 3.3135,
      "mean_token_accuracy": 0.8382684588432312,
      "step": 190
    },
    {
      "epoch": 1.528,
      "grad_norm": 5.597368240356445,
      "learning_rate": 3.121212121212122e-05,
      "loss": 3.2368,
      "mean_token_accuracy": 0.8341715931892395,
      "step": 191
    },
    {
      "epoch": 1.536,
      "grad_norm": 6.087445259094238,
      "learning_rate": 3.111111111111111e-05,
      "loss": 3.4584,
      "mean_token_accuracy": 0.8352274149656296,
      "step": 192
    },
    {
      "epoch": 1.544,
      "grad_norm": 5.523479461669922,
      "learning_rate": 3.101010101010101e-05,
      "loss": 3.0476,
      "mean_token_accuracy": 0.8508585840463638,
      "step": 193
    },
    {
      "epoch": 1.552,
      "grad_norm": 5.6211442947387695,
      "learning_rate": 3.090909090909091e-05,
      "loss": 3.2642,
      "mean_token_accuracy": 0.8426928371191025,
      "step": 194
    },
    {
      "epoch": 1.56,
      "grad_norm": 5.739730358123779,
      "learning_rate": 3.080808080808081e-05,
      "loss": 3.6835,
      "mean_token_accuracy": 0.8226543515920639,
      "step": 195
    },
    {
      "epoch": 1.568,
      "grad_norm": 6.00140380859375,
      "learning_rate": 3.070707070707071e-05,
      "loss": 2.9485,
      "mean_token_accuracy": 0.8533317595720291,
      "step": 196
    },
    {
      "epoch": 1.576,
      "grad_norm": 5.541549205780029,
      "learning_rate": 3.060606060606061e-05,
      "loss": 3.0892,
      "mean_token_accuracy": 0.8474195152521133,
      "step": 197
    },
    {
      "epoch": 1.584,
      "grad_norm": 5.923431873321533,
      "learning_rate": 3.050505050505051e-05,
      "loss": 3.018,
      "mean_token_accuracy": 0.84869784116745,
      "step": 198
    },
    {
      "epoch": 1.592,
      "grad_norm": 6.204696178436279,
      "learning_rate": 3.0404040404040406e-05,
      "loss": 3.5126,
      "mean_token_accuracy": 0.8261184245347977,
      "step": 199
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.3025712966918945,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 3.4185,
      "mean_token_accuracy": 0.8210860043764114,
      "step": 200
    },
    {
      "epoch": 1.608,
      "grad_norm": 5.689709663391113,
      "learning_rate": 3.0202020202020205e-05,
      "loss": 3.5322,
      "mean_token_accuracy": 0.8191230297088623,
      "step": 201
    },
    {
      "epoch": 1.616,
      "grad_norm": 5.707929611206055,
      "learning_rate": 3.01010101010101e-05,
      "loss": 3.024,
      "mean_token_accuracy": 0.8488472253084183,
      "step": 202
    },
    {
      "epoch": 1.624,
      "grad_norm": 5.80771017074585,
      "learning_rate": 3e-05,
      "loss": 3.5556,
      "mean_token_accuracy": 0.8266987651586533,
      "step": 203
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 6.03700590133667,
      "learning_rate": 2.98989898989899e-05,
      "loss": 3.1062,
      "mean_token_accuracy": 0.8308413475751877,
      "step": 204
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 6.733646392822266,
      "learning_rate": 2.9797979797979796e-05,
      "loss": 3.1995,
      "mean_token_accuracy": 0.8294665813446045,
      "step": 205
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 5.883334636688232,
      "learning_rate": 2.96969696969697e-05,
      "loss": 2.8124,
      "mean_token_accuracy": 0.8439156115055084,
      "step": 206
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 5.941483497619629,
      "learning_rate": 2.95959595959596e-05,
      "loss": 3.0179,
      "mean_token_accuracy": 0.8416165858507156,
      "step": 207
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 6.391957759857178,
      "learning_rate": 2.9494949494949498e-05,
      "loss": 3.7068,
      "mean_token_accuracy": 0.8122821152210236,
      "step": 208
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 6.237335205078125,
      "learning_rate": 2.9393939393939394e-05,
      "loss": 3.32,
      "mean_token_accuracy": 0.8313788622617722,
      "step": 209
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 6.5731987953186035,
      "learning_rate": 2.9292929292929294e-05,
      "loss": 3.5619,
      "mean_token_accuracy": 0.816206768155098,
      "step": 210
    },
    {
      "epoch": 1.688,
      "grad_norm": 5.061030387878418,
      "learning_rate": 2.9191919191919193e-05,
      "loss": 2.8566,
      "mean_token_accuracy": 0.8475048393011093,
      "step": 211
    },
    {
      "epoch": 1.696,
      "grad_norm": 5.921192646026611,
      "learning_rate": 2.909090909090909e-05,
      "loss": 3.3349,
      "mean_token_accuracy": 0.827596977353096,
      "step": 212
    },
    {
      "epoch": 1.704,
      "grad_norm": 5.721929550170898,
      "learning_rate": 2.898989898989899e-05,
      "loss": 2.7363,
      "mean_token_accuracy": 0.8517429679632187,
      "step": 213
    },
    {
      "epoch": 1.712,
      "grad_norm": 6.763914585113525,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 2.8292,
      "mean_token_accuracy": 0.8498516529798508,
      "step": 214
    },
    {
      "epoch": 1.72,
      "grad_norm": 6.817018032073975,
      "learning_rate": 2.878787878787879e-05,
      "loss": 3.2958,
      "mean_token_accuracy": 0.8348537087440491,
      "step": 215
    },
    {
      "epoch": 1.728,
      "grad_norm": 5.659510612487793,
      "learning_rate": 2.868686868686869e-05,
      "loss": 2.8683,
      "mean_token_accuracy": 0.8435689210891724,
      "step": 216
    },
    {
      "epoch": 1.736,
      "grad_norm": 6.099102020263672,
      "learning_rate": 2.8585858585858587e-05,
      "loss": 2.9825,
      "mean_token_accuracy": 0.8394478559494019,
      "step": 217
    },
    {
      "epoch": 1.744,
      "grad_norm": 6.094569206237793,
      "learning_rate": 2.8484848484848486e-05,
      "loss": 3.2325,
      "mean_token_accuracy": 0.8246908932924271,
      "step": 218
    },
    {
      "epoch": 1.752,
      "grad_norm": 6.866254806518555,
      "learning_rate": 2.8383838383838386e-05,
      "loss": 3.1284,
      "mean_token_accuracy": 0.8342044651508331,
      "step": 219
    },
    {
      "epoch": 1.76,
      "grad_norm": 6.149699687957764,
      "learning_rate": 2.8282828282828282e-05,
      "loss": 3.6781,
      "mean_token_accuracy": 0.8211702555418015,
      "step": 220
    },
    {
      "epoch": 1.768,
      "grad_norm": 5.862130641937256,
      "learning_rate": 2.818181818181818e-05,
      "loss": 3.3624,
      "mean_token_accuracy": 0.8324427157640457,
      "step": 221
    },
    {
      "epoch": 1.776,
      "grad_norm": 5.395320415496826,
      "learning_rate": 2.808080808080808e-05,
      "loss": 3.3338,
      "mean_token_accuracy": 0.8399553894996643,
      "step": 222
    },
    {
      "epoch": 1.784,
      "grad_norm": 5.550381183624268,
      "learning_rate": 2.7979797979797984e-05,
      "loss": 3.4036,
      "mean_token_accuracy": 0.8223441988229752,
      "step": 223
    },
    {
      "epoch": 1.792,
      "grad_norm": 5.406087875366211,
      "learning_rate": 2.7878787878787883e-05,
      "loss": 2.6777,
      "mean_token_accuracy": 0.8600380420684814,
      "step": 224
    },
    {
      "epoch": 1.8,
      "grad_norm": 5.7635698318481445,
      "learning_rate": 2.777777777777778e-05,
      "loss": 3.5319,
      "mean_token_accuracy": 0.8255642205476761,
      "step": 225
    },
    {
      "epoch": 1.808,
      "grad_norm": 5.934004783630371,
      "learning_rate": 2.767676767676768e-05,
      "loss": 3.2659,
      "mean_token_accuracy": 0.8305630534887314,
      "step": 226
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 6.8721184730529785,
      "learning_rate": 2.7575757575757578e-05,
      "loss": 3.1587,
      "mean_token_accuracy": 0.8299548774957657,
      "step": 227
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 5.556332588195801,
      "learning_rate": 2.7474747474747474e-05,
      "loss": 2.8912,
      "mean_token_accuracy": 0.8541744500398636,
      "step": 228
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 5.591052532196045,
      "learning_rate": 2.7373737373737374e-05,
      "loss": 3.1768,
      "mean_token_accuracy": 0.840098574757576,
      "step": 229
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 7.923620700836182,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 3.1165,
      "mean_token_accuracy": 0.8416318744421005,
      "step": 230
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 6.180293083190918,
      "learning_rate": 2.717171717171717e-05,
      "loss": 3.1878,
      "mean_token_accuracy": 0.8414299786090851,
      "step": 231
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 6.479142189025879,
      "learning_rate": 2.7070707070707075e-05,
      "loss": 3.519,
      "mean_token_accuracy": 0.8237949758768082,
      "step": 232
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 6.211218357086182,
      "learning_rate": 2.696969696969697e-05,
      "loss": 3.6602,
      "mean_token_accuracy": 0.8112812489271164,
      "step": 233
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 6.262876033782959,
      "learning_rate": 2.686868686868687e-05,
      "loss": 3.8658,
      "mean_token_accuracy": 0.8071164190769196,
      "step": 234
    },
    {
      "epoch": 1.88,
      "grad_norm": 5.474344730377197,
      "learning_rate": 2.676767676767677e-05,
      "loss": 2.971,
      "mean_token_accuracy": 0.8377246409654617,
      "step": 235
    },
    {
      "epoch": 1.888,
      "grad_norm": 6.193055629730225,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 3.478,
      "mean_token_accuracy": 0.823166161775589,
      "step": 236
    },
    {
      "epoch": 1.896,
      "grad_norm": 5.627188682556152,
      "learning_rate": 2.6565656565656566e-05,
      "loss": 3.1995,
      "mean_token_accuracy": 0.8456520289182663,
      "step": 237
    },
    {
      "epoch": 1.904,
      "grad_norm": 6.110049247741699,
      "learning_rate": 2.6464646464646466e-05,
      "loss": 3.0544,
      "mean_token_accuracy": 0.8367758989334106,
      "step": 238
    },
    {
      "epoch": 1.912,
      "grad_norm": 5.3701982498168945,
      "learning_rate": 2.636363636363636e-05,
      "loss": 2.6659,
      "mean_token_accuracy": 0.8609540313482285,
      "step": 239
    },
    {
      "epoch": 1.92,
      "grad_norm": 5.352710247039795,
      "learning_rate": 2.6262626262626268e-05,
      "loss": 3.082,
      "mean_token_accuracy": 0.842663049697876,
      "step": 240
    },
    {
      "epoch": 1.928,
      "grad_norm": 6.08484411239624,
      "learning_rate": 2.6161616161616164e-05,
      "loss": 3.3737,
      "mean_token_accuracy": 0.8245035409927368,
      "step": 241
    },
    {
      "epoch": 1.936,
      "grad_norm": 5.956201076507568,
      "learning_rate": 2.6060606060606063e-05,
      "loss": 2.9072,
      "mean_token_accuracy": 0.8548275828361511,
      "step": 242
    },
    {
      "epoch": 1.944,
      "grad_norm": 5.499819278717041,
      "learning_rate": 2.5959595959595963e-05,
      "loss": 3.355,
      "mean_token_accuracy": 0.820908397436142,
      "step": 243
    },
    {
      "epoch": 1.952,
      "grad_norm": 6.338292598724365,
      "learning_rate": 2.585858585858586e-05,
      "loss": 3.7158,
      "mean_token_accuracy": 0.8252478241920471,
      "step": 244
    },
    {
      "epoch": 1.96,
      "grad_norm": 6.105760097503662,
      "learning_rate": 2.575757575757576e-05,
      "loss": 3.1973,
      "mean_token_accuracy": 0.8429799377918243,
      "step": 245
    },
    {
      "epoch": 1.968,
      "grad_norm": 5.636971950531006,
      "learning_rate": 2.5656565656565658e-05,
      "loss": 2.6206,
      "mean_token_accuracy": 0.8573070019483566,
      "step": 246
    },
    {
      "epoch": 1.976,
      "grad_norm": 6.419624328613281,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 3.5778,
      "mean_token_accuracy": 0.8265507072210312,
      "step": 247
    },
    {
      "epoch": 1.984,
      "grad_norm": 5.554307460784912,
      "learning_rate": 2.5454545454545454e-05,
      "loss": 2.9344,
      "mean_token_accuracy": 0.8515568971633911,
      "step": 248
    },
    {
      "epoch": 1.992,
      "grad_norm": 6.059617519378662,
      "learning_rate": 2.5353535353535356e-05,
      "loss": 3.4458,
      "mean_token_accuracy": 0.8348643332719803,
      "step": 249
    },
    {
      "epoch": 2.0,
      "grad_norm": 5.995134353637695,
      "learning_rate": 2.5252525252525256e-05,
      "loss": 2.9078,
      "mean_token_accuracy": 0.8390886038541794,
      "step": 250
    },
    {
      "epoch": 2.008,
      "grad_norm": 5.29079532623291,
      "learning_rate": 2.5151515151515155e-05,
      "loss": 2.5449,
      "mean_token_accuracy": 0.8623387068510056,
      "step": 251
    },
    {
      "epoch": 2.016,
      "grad_norm": 6.307847499847412,
      "learning_rate": 2.505050505050505e-05,
      "loss": 2.6746,
      "mean_token_accuracy": 0.8547452688217163,
      "step": 252
    },
    {
      "epoch": 2.024,
      "grad_norm": 5.258662700653076,
      "learning_rate": 2.494949494949495e-05,
      "loss": 2.8106,
      "mean_token_accuracy": 0.8465935438871384,
      "step": 253
    },
    {
      "epoch": 2.032,
      "grad_norm": 5.160598278045654,
      "learning_rate": 2.4848484848484847e-05,
      "loss": 2.5327,
      "mean_token_accuracy": 0.8596736937761307,
      "step": 254
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.5139055252075195,
      "learning_rate": 2.474747474747475e-05,
      "loss": 2.5406,
      "mean_token_accuracy": 0.8599353432655334,
      "step": 255
    },
    {
      "epoch": 2.048,
      "grad_norm": 5.596591949462891,
      "learning_rate": 2.464646464646465e-05,
      "loss": 2.5955,
      "mean_token_accuracy": 0.8636805862188339,
      "step": 256
    },
    {
      "epoch": 2.056,
      "grad_norm": 5.260463237762451,
      "learning_rate": 2.4545454545454545e-05,
      "loss": 2.8003,
      "mean_token_accuracy": 0.8472322225570679,
      "step": 257
    },
    {
      "epoch": 2.064,
      "grad_norm": 5.485147476196289,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 2.4907,
      "mean_token_accuracy": 0.8619499355554581,
      "step": 258
    },
    {
      "epoch": 2.072,
      "grad_norm": 5.589687824249268,
      "learning_rate": 2.4343434343434344e-05,
      "loss": 2.563,
      "mean_token_accuracy": 0.8651385009288788,
      "step": 259
    },
    {
      "epoch": 2.08,
      "grad_norm": 4.876934051513672,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 2.0974,
      "mean_token_accuracy": 0.8819858431816101,
      "step": 260
    },
    {
      "epoch": 2.088,
      "grad_norm": 6.266694068908691,
      "learning_rate": 2.4141414141414143e-05,
      "loss": 2.6164,
      "mean_token_accuracy": 0.8617752641439438,
      "step": 261
    },
    {
      "epoch": 2.096,
      "grad_norm": 5.54473352432251,
      "learning_rate": 2.404040404040404e-05,
      "loss": 2.9005,
      "mean_token_accuracy": 0.8471231460571289,
      "step": 262
    },
    {
      "epoch": 2.104,
      "grad_norm": 5.299022197723389,
      "learning_rate": 2.393939393939394e-05,
      "loss": 2.3437,
      "mean_token_accuracy": 0.872315376996994,
      "step": 263
    },
    {
      "epoch": 2.112,
      "grad_norm": 5.8907856941223145,
      "learning_rate": 2.3838383838383842e-05,
      "loss": 2.5687,
      "mean_token_accuracy": 0.8683241903781891,
      "step": 264
    },
    {
      "epoch": 2.12,
      "grad_norm": 5.84760046005249,
      "learning_rate": 2.3737373737373738e-05,
      "loss": 2.6167,
      "mean_token_accuracy": 0.8569298684597015,
      "step": 265
    },
    {
      "epoch": 2.128,
      "grad_norm": 5.831202507019043,
      "learning_rate": 2.3636363636363637e-05,
      "loss": 3.0151,
      "mean_token_accuracy": 0.8395980596542358,
      "step": 266
    },
    {
      "epoch": 2.136,
      "grad_norm": 5.698826313018799,
      "learning_rate": 2.3535353535353537e-05,
      "loss": 3.1071,
      "mean_token_accuracy": 0.850052535533905,
      "step": 267
    },
    {
      "epoch": 2.144,
      "grad_norm": 6.112237930297852,
      "learning_rate": 2.3434343434343436e-05,
      "loss": 2.624,
      "mean_token_accuracy": 0.8661712259054184,
      "step": 268
    },
    {
      "epoch": 2.152,
      "grad_norm": 6.229471683502197,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 2.5656,
      "mean_token_accuracy": 0.8691338896751404,
      "step": 269
    },
    {
      "epoch": 2.16,
      "grad_norm": 6.126317501068115,
      "learning_rate": 2.3232323232323232e-05,
      "loss": 2.6405,
      "mean_token_accuracy": 0.8534891903400421,
      "step": 270
    },
    {
      "epoch": 2.168,
      "grad_norm": 6.28452205657959,
      "learning_rate": 2.313131313131313e-05,
      "loss": 2.6796,
      "mean_token_accuracy": 0.8531413078308105,
      "step": 271
    },
    {
      "epoch": 2.176,
      "grad_norm": 5.863641262054443,
      "learning_rate": 2.3030303030303034e-05,
      "loss": 2.7015,
      "mean_token_accuracy": 0.8519039303064346,
      "step": 272
    },
    {
      "epoch": 2.184,
      "grad_norm": 6.535420894622803,
      "learning_rate": 2.292929292929293e-05,
      "loss": 3.0103,
      "mean_token_accuracy": 0.8415365815162659,
      "step": 273
    },
    {
      "epoch": 2.192,
      "grad_norm": 5.875096321105957,
      "learning_rate": 2.282828282828283e-05,
      "loss": 2.8179,
      "mean_token_accuracy": 0.8494950979948044,
      "step": 274
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.35692834854126,
      "learning_rate": 2.272727272727273e-05,
      "loss": 2.1563,
      "mean_token_accuracy": 0.8810548335313797,
      "step": 275
    },
    {
      "epoch": 2.208,
      "grad_norm": 6.05448579788208,
      "learning_rate": 2.262626262626263e-05,
      "loss": 2.4902,
      "mean_token_accuracy": 0.8656753301620483,
      "step": 276
    },
    {
      "epoch": 2.216,
      "grad_norm": 5.888792037963867,
      "learning_rate": 2.2525252525252528e-05,
      "loss": 2.5769,
      "mean_token_accuracy": 0.8682566732168198,
      "step": 277
    },
    {
      "epoch": 2.224,
      "grad_norm": 5.912112712860107,
      "learning_rate": 2.2424242424242424e-05,
      "loss": 2.5715,
      "mean_token_accuracy": 0.8543179333209991,
      "step": 278
    },
    {
      "epoch": 2.232,
      "grad_norm": 6.485673427581787,
      "learning_rate": 2.2323232323232324e-05,
      "loss": 2.6801,
      "mean_token_accuracy": 0.8597021549940109,
      "step": 279
    },
    {
      "epoch": 2.24,
      "grad_norm": 5.757596969604492,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 2.5495,
      "mean_token_accuracy": 0.8788398951292038,
      "step": 280
    },
    {
      "epoch": 2.248,
      "grad_norm": 5.577345848083496,
      "learning_rate": 2.2121212121212123e-05,
      "loss": 2.4548,
      "mean_token_accuracy": 0.8728667497634888,
      "step": 281
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 8.356124877929688,
      "learning_rate": 2.2020202020202022e-05,
      "loss": 2.9846,
      "mean_token_accuracy": 0.8483117371797562,
      "step": 282
    },
    {
      "epoch": 2.2640000000000002,
      "grad_norm": 5.887564659118652,
      "learning_rate": 2.191919191919192e-05,
      "loss": 2.7833,
      "mean_token_accuracy": 0.8488295823335648,
      "step": 283
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 6.558447360992432,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 2.3944,
      "mean_token_accuracy": 0.8682046681642532,
      "step": 284
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 6.828135013580322,
      "learning_rate": 2.171717171717172e-05,
      "loss": 2.8403,
      "mean_token_accuracy": 0.8527731895446777,
      "step": 285
    },
    {
      "epoch": 2.288,
      "grad_norm": 5.772212505340576,
      "learning_rate": 2.1616161616161617e-05,
      "loss": 2.8966,
      "mean_token_accuracy": 0.8490441888570786,
      "step": 286
    },
    {
      "epoch": 2.296,
      "grad_norm": 5.538363933563232,
      "learning_rate": 2.1515151515151516e-05,
      "loss": 2.4595,
      "mean_token_accuracy": 0.8743099421262741,
      "step": 287
    },
    {
      "epoch": 2.304,
      "grad_norm": 6.172807216644287,
      "learning_rate": 2.1414141414141416e-05,
      "loss": 2.3328,
      "mean_token_accuracy": 0.8704792261123657,
      "step": 288
    },
    {
      "epoch": 2.312,
      "grad_norm": 6.232668399810791,
      "learning_rate": 2.1313131313131315e-05,
      "loss": 2.9026,
      "mean_token_accuracy": 0.8546594232320786,
      "step": 289
    },
    {
      "epoch": 2.32,
      "grad_norm": 6.070257663726807,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 2.4686,
      "mean_token_accuracy": 0.8643316328525543,
      "step": 290
    },
    {
      "epoch": 2.328,
      "grad_norm": 5.737977504730225,
      "learning_rate": 2.111111111111111e-05,
      "loss": 2.9917,
      "mean_token_accuracy": 0.8514521420001984,
      "step": 291
    },
    {
      "epoch": 2.336,
      "grad_norm": 6.409512519836426,
      "learning_rate": 2.101010101010101e-05,
      "loss": 2.6572,
      "mean_token_accuracy": 0.8566225320100784,
      "step": 292
    },
    {
      "epoch": 2.344,
      "grad_norm": 6.57427978515625,
      "learning_rate": 2.090909090909091e-05,
      "loss": 2.466,
      "mean_token_accuracy": 0.853950634598732,
      "step": 293
    },
    {
      "epoch": 2.352,
      "grad_norm": 6.017867088317871,
      "learning_rate": 2.080808080808081e-05,
      "loss": 2.8763,
      "mean_token_accuracy": 0.8532497733831406,
      "step": 294
    },
    {
      "epoch": 2.36,
      "grad_norm": 5.640237808227539,
      "learning_rate": 2.070707070707071e-05,
      "loss": 2.6013,
      "mean_token_accuracy": 0.8739284723997116,
      "step": 295
    },
    {
      "epoch": 2.368,
      "grad_norm": 6.341038703918457,
      "learning_rate": 2.0606060606060608e-05,
      "loss": 2.6485,
      "mean_token_accuracy": 0.8521928191184998,
      "step": 296
    },
    {
      "epoch": 2.376,
      "grad_norm": 6.142149925231934,
      "learning_rate": 2.0505050505050504e-05,
      "loss": 2.3397,
      "mean_token_accuracy": 0.8850863575935364,
      "step": 297
    },
    {
      "epoch": 2.384,
      "grad_norm": 6.308354377746582,
      "learning_rate": 2.0404040404040407e-05,
      "loss": 2.5977,
      "mean_token_accuracy": 0.8648134768009186,
      "step": 298
    },
    {
      "epoch": 2.392,
      "grad_norm": 5.566152572631836,
      "learning_rate": 2.0303030303030303e-05,
      "loss": 2.7123,
      "mean_token_accuracy": 0.8633100241422653,
      "step": 299
    },
    {
      "epoch": 2.4,
      "grad_norm": 5.725775241851807,
      "learning_rate": 2.0202020202020203e-05,
      "loss": 2.6108,
      "mean_token_accuracy": 0.8738928139209747,
      "step": 300
    },
    {
      "epoch": 2.408,
      "grad_norm": 6.435664176940918,
      "learning_rate": 2.0101010101010102e-05,
      "loss": 2.7976,
      "mean_token_accuracy": 0.8510608673095703,
      "step": 301
    },
    {
      "epoch": 2.416,
      "grad_norm": 7.4542622566223145,
      "learning_rate": 2e-05,
      "loss": 2.8097,
      "mean_token_accuracy": 0.8516113609075546,
      "step": 302
    },
    {
      "epoch": 2.424,
      "grad_norm": 5.655482769012451,
      "learning_rate": 1.98989898989899e-05,
      "loss": 2.3856,
      "mean_token_accuracy": 0.8649388700723648,
      "step": 303
    },
    {
      "epoch": 2.432,
      "grad_norm": 6.412247180938721,
      "learning_rate": 1.9797979797979797e-05,
      "loss": 2.8675,
      "mean_token_accuracy": 0.8470883667469025,
      "step": 304
    },
    {
      "epoch": 2.44,
      "grad_norm": 7.186456680297852,
      "learning_rate": 1.9696969696969697e-05,
      "loss": 3.1156,
      "mean_token_accuracy": 0.8328486531972885,
      "step": 305
    },
    {
      "epoch": 2.448,
      "grad_norm": 5.615805625915527,
      "learning_rate": 1.95959595959596e-05,
      "loss": 2.6565,
      "mean_token_accuracy": 0.8666907250881195,
      "step": 306
    },
    {
      "epoch": 2.456,
      "grad_norm": 6.449917793273926,
      "learning_rate": 1.9494949494949496e-05,
      "loss": 2.9544,
      "mean_token_accuracy": 0.8483339697122574,
      "step": 307
    },
    {
      "epoch": 2.464,
      "grad_norm": 5.747985363006592,
      "learning_rate": 1.9393939393939395e-05,
      "loss": 2.1422,
      "mean_token_accuracy": 0.8723134845495224,
      "step": 308
    },
    {
      "epoch": 2.472,
      "grad_norm": 6.027318000793457,
      "learning_rate": 1.9292929292929295e-05,
      "loss": 2.5305,
      "mean_token_accuracy": 0.8660032451152802,
      "step": 309
    },
    {
      "epoch": 2.48,
      "grad_norm": 7.454213619232178,
      "learning_rate": 1.919191919191919e-05,
      "loss": 2.5679,
      "mean_token_accuracy": 0.8533085733652115,
      "step": 310
    },
    {
      "epoch": 2.488,
      "grad_norm": 5.4801716804504395,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 2.6006,
      "mean_token_accuracy": 0.8550811111927032,
      "step": 311
    },
    {
      "epoch": 2.496,
      "grad_norm": 6.020359039306641,
      "learning_rate": 1.898989898989899e-05,
      "loss": 2.2521,
      "mean_token_accuracy": 0.8731967061758041,
      "step": 312
    },
    {
      "epoch": 2.504,
      "grad_norm": 5.78825569152832,
      "learning_rate": 1.888888888888889e-05,
      "loss": 2.4461,
      "mean_token_accuracy": 0.8596260696649551,
      "step": 313
    },
    {
      "epoch": 2.512,
      "grad_norm": 5.771088123321533,
      "learning_rate": 1.878787878787879e-05,
      "loss": 2.3915,
      "mean_token_accuracy": 0.8663473874330521,
      "step": 314
    },
    {
      "epoch": 2.52,
      "grad_norm": 5.939653396606445,
      "learning_rate": 1.8686868686868688e-05,
      "loss": 2.7159,
      "mean_token_accuracy": 0.8722144514322281,
      "step": 315
    },
    {
      "epoch": 2.528,
      "grad_norm": 6.489267349243164,
      "learning_rate": 1.8585858585858588e-05,
      "loss": 2.6137,
      "mean_token_accuracy": 0.8580323159694672,
      "step": 316
    },
    {
      "epoch": 2.536,
      "grad_norm": 5.97398042678833,
      "learning_rate": 1.8484848484848487e-05,
      "loss": 2.7352,
      "mean_token_accuracy": 0.8655538409948349,
      "step": 317
    },
    {
      "epoch": 2.544,
      "grad_norm": 5.858009338378906,
      "learning_rate": 1.8383838383838383e-05,
      "loss": 2.2991,
      "mean_token_accuracy": 0.8808925747871399,
      "step": 318
    },
    {
      "epoch": 2.552,
      "grad_norm": 5.720334053039551,
      "learning_rate": 1.8282828282828286e-05,
      "loss": 2.6282,
      "mean_token_accuracy": 0.8669092357158661,
      "step": 319
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.569767475128174,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 2.5663,
      "mean_token_accuracy": 0.8692184388637543,
      "step": 320
    },
    {
      "epoch": 2.568,
      "grad_norm": 6.509884357452393,
      "learning_rate": 1.808080808080808e-05,
      "loss": 2.8698,
      "mean_token_accuracy": 0.8457562029361725,
      "step": 321
    },
    {
      "epoch": 2.576,
      "grad_norm": 5.976521968841553,
      "learning_rate": 1.797979797979798e-05,
      "loss": 2.7501,
      "mean_token_accuracy": 0.846638560295105,
      "step": 322
    },
    {
      "epoch": 2.584,
      "grad_norm": 5.878550052642822,
      "learning_rate": 1.787878787878788e-05,
      "loss": 2.2021,
      "mean_token_accuracy": 0.8834634870290756,
      "step": 323
    },
    {
      "epoch": 2.592,
      "grad_norm": 5.905157566070557,
      "learning_rate": 1.777777777777778e-05,
      "loss": 2.5808,
      "mean_token_accuracy": 0.8616163432598114,
      "step": 324
    },
    {
      "epoch": 2.6,
      "grad_norm": 5.510688304901123,
      "learning_rate": 1.7676767676767676e-05,
      "loss": 2.265,
      "mean_token_accuracy": 0.8738918304443359,
      "step": 325
    },
    {
      "epoch": 2.608,
      "grad_norm": 5.952402591705322,
      "learning_rate": 1.7575757575757576e-05,
      "loss": 2.7743,
      "mean_token_accuracy": 0.85427226126194,
      "step": 326
    },
    {
      "epoch": 2.616,
      "grad_norm": 6.647154808044434,
      "learning_rate": 1.7474747474747475e-05,
      "loss": 2.9181,
      "mean_token_accuracy": 0.8349238932132721,
      "step": 327
    },
    {
      "epoch": 2.624,
      "grad_norm": 7.24924898147583,
      "learning_rate": 1.7373737373737375e-05,
      "loss": 3.1174,
      "mean_token_accuracy": 0.8411069959402084,
      "step": 328
    },
    {
      "epoch": 2.632,
      "grad_norm": 5.238758563995361,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 2.4068,
      "mean_token_accuracy": 0.8669510632753372,
      "step": 329
    },
    {
      "epoch": 2.64,
      "grad_norm": 5.997193336486816,
      "learning_rate": 1.7171717171717173e-05,
      "loss": 3.0963,
      "mean_token_accuracy": 0.8448814451694489,
      "step": 330
    },
    {
      "epoch": 2.648,
      "grad_norm": 6.348751544952393,
      "learning_rate": 1.707070707070707e-05,
      "loss": 2.8593,
      "mean_token_accuracy": 0.8491098284721375,
      "step": 331
    },
    {
      "epoch": 2.656,
      "grad_norm": 5.824342250823975,
      "learning_rate": 1.6969696969696972e-05,
      "loss": 2.5448,
      "mean_token_accuracy": 0.8616123348474503,
      "step": 332
    },
    {
      "epoch": 2.664,
      "grad_norm": 7.513331890106201,
      "learning_rate": 1.686868686868687e-05,
      "loss": 2.5108,
      "mean_token_accuracy": 0.8665321916341782,
      "step": 333
    },
    {
      "epoch": 2.672,
      "grad_norm": 5.547114372253418,
      "learning_rate": 1.6767676767676768e-05,
      "loss": 2.674,
      "mean_token_accuracy": 0.8699924349784851,
      "step": 334
    },
    {
      "epoch": 2.68,
      "grad_norm": 6.60705041885376,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 2.9442,
      "mean_token_accuracy": 0.8435302972793579,
      "step": 335
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 6.116333961486816,
      "learning_rate": 1.6565656565656567e-05,
      "loss": 2.3428,
      "mean_token_accuracy": 0.8697264492511749,
      "step": 336
    },
    {
      "epoch": 2.6959999999999997,
      "grad_norm": 6.642714500427246,
      "learning_rate": 1.6464646464646466e-05,
      "loss": 2.7629,
      "mean_token_accuracy": 0.863296702504158,
      "step": 337
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 6.764316558837891,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 3.0041,
      "mean_token_accuracy": 0.8556036204099655,
      "step": 338
    },
    {
      "epoch": 2.7119999999999997,
      "grad_norm": 5.682474613189697,
      "learning_rate": 1.6262626262626262e-05,
      "loss": 2.4677,
      "mean_token_accuracy": 0.8652613461017609,
      "step": 339
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 6.347187519073486,
      "learning_rate": 1.6161616161616165e-05,
      "loss": 2.4674,
      "mean_token_accuracy": 0.8690385818481445,
      "step": 340
    },
    {
      "epoch": 2.7279999999999998,
      "grad_norm": 5.507145404815674,
      "learning_rate": 1.606060606060606e-05,
      "loss": 2.692,
      "mean_token_accuracy": 0.8610135316848755,
      "step": 341
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 6.108325481414795,
      "learning_rate": 1.595959595959596e-05,
      "loss": 2.4909,
      "mean_token_accuracy": 0.8648791164159775,
      "step": 342
    },
    {
      "epoch": 2.7439999999999998,
      "grad_norm": 6.017334938049316,
      "learning_rate": 1.585858585858586e-05,
      "loss": 2.2442,
      "mean_token_accuracy": 0.8753398060798645,
      "step": 343
    },
    {
      "epoch": 2.752,
      "grad_norm": 6.193384170532227,
      "learning_rate": 1.5757575757575756e-05,
      "loss": 2.3729,
      "mean_token_accuracy": 0.8614612221717834,
      "step": 344
    },
    {
      "epoch": 2.76,
      "grad_norm": 6.495743751525879,
      "learning_rate": 1.565656565656566e-05,
      "loss": 2.8584,
      "mean_token_accuracy": 0.8486870527267456,
      "step": 345
    },
    {
      "epoch": 2.768,
      "grad_norm": 5.537295818328857,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 2.3311,
      "mean_token_accuracy": 0.8707073032855988,
      "step": 346
    },
    {
      "epoch": 2.776,
      "grad_norm": 5.175790309906006,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 2.4548,
      "mean_token_accuracy": 0.8665929436683655,
      "step": 347
    },
    {
      "epoch": 2.784,
      "grad_norm": 5.646080017089844,
      "learning_rate": 1.5353535353535354e-05,
      "loss": 2.4042,
      "mean_token_accuracy": 0.8688449859619141,
      "step": 348
    },
    {
      "epoch": 2.792,
      "grad_norm": 5.514770984649658,
      "learning_rate": 1.5252525252525255e-05,
      "loss": 2.3592,
      "mean_token_accuracy": 0.868006706237793,
      "step": 349
    },
    {
      "epoch": 2.8,
      "grad_norm": 5.443539619445801,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 2.7749,
      "mean_token_accuracy": 0.8517529368400574,
      "step": 350
    },
    {
      "epoch": 2.808,
      "grad_norm": 6.177525043487549,
      "learning_rate": 1.505050505050505e-05,
      "loss": 2.7422,
      "mean_token_accuracy": 0.853731244802475,
      "step": 351
    },
    {
      "epoch": 2.816,
      "grad_norm": 5.842599868774414,
      "learning_rate": 1.494949494949495e-05,
      "loss": 2.7828,
      "mean_token_accuracy": 0.8474886268377304,
      "step": 352
    },
    {
      "epoch": 2.824,
      "grad_norm": 6.956360816955566,
      "learning_rate": 1.484848484848485e-05,
      "loss": 2.6584,
      "mean_token_accuracy": 0.8608374446630478,
      "step": 353
    },
    {
      "epoch": 2.832,
      "grad_norm": 6.075902938842773,
      "learning_rate": 1.4747474747474749e-05,
      "loss": 1.7885,
      "mean_token_accuracy": 0.8908968865871429,
      "step": 354
    },
    {
      "epoch": 2.84,
      "grad_norm": 5.676952838897705,
      "learning_rate": 1.4646464646464647e-05,
      "loss": 2.2233,
      "mean_token_accuracy": 0.8666234165430069,
      "step": 355
    },
    {
      "epoch": 2.848,
      "grad_norm": 6.072289943695068,
      "learning_rate": 1.4545454545454545e-05,
      "loss": 2.9788,
      "mean_token_accuracy": 0.8446744084358215,
      "step": 356
    },
    {
      "epoch": 2.856,
      "grad_norm": 6.316305637359619,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 2.6826,
      "mean_token_accuracy": 0.8572750687599182,
      "step": 357
    },
    {
      "epoch": 2.864,
      "grad_norm": 6.896461486816406,
      "learning_rate": 1.4343434343434345e-05,
      "loss": 2.3163,
      "mean_token_accuracy": 0.874391183257103,
      "step": 358
    },
    {
      "epoch": 2.872,
      "grad_norm": 6.410128116607666,
      "learning_rate": 1.4242424242424243e-05,
      "loss": 2.66,
      "mean_token_accuracy": 0.8778804391622543,
      "step": 359
    },
    {
      "epoch": 2.88,
      "grad_norm": 5.7101664543151855,
      "learning_rate": 1.4141414141414141e-05,
      "loss": 2.5463,
      "mean_token_accuracy": 0.8713642507791519,
      "step": 360
    },
    {
      "epoch": 2.888,
      "grad_norm": 5.941382884979248,
      "learning_rate": 1.404040404040404e-05,
      "loss": 2.4611,
      "mean_token_accuracy": 0.8689820915460587,
      "step": 361
    },
    {
      "epoch": 2.896,
      "grad_norm": 5.766918182373047,
      "learning_rate": 1.3939393939393942e-05,
      "loss": 2.3588,
      "mean_token_accuracy": 0.8715604543685913,
      "step": 362
    },
    {
      "epoch": 2.904,
      "grad_norm": 6.412060260772705,
      "learning_rate": 1.383838383838384e-05,
      "loss": 2.9418,
      "mean_token_accuracy": 0.8422341346740723,
      "step": 363
    },
    {
      "epoch": 2.912,
      "grad_norm": 6.339536666870117,
      "learning_rate": 1.3737373737373737e-05,
      "loss": 2.8499,
      "mean_token_accuracy": 0.8519201278686523,
      "step": 364
    },
    {
      "epoch": 2.92,
      "grad_norm": 5.80408239364624,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 2.6526,
      "mean_token_accuracy": 0.8615650832653046,
      "step": 365
    },
    {
      "epoch": 2.928,
      "grad_norm": 5.643406391143799,
      "learning_rate": 1.3535353535353538e-05,
      "loss": 2.4131,
      "mean_token_accuracy": 0.8788148909807205,
      "step": 366
    },
    {
      "epoch": 2.936,
      "grad_norm": 6.2641167640686035,
      "learning_rate": 1.3434343434343436e-05,
      "loss": 2.7667,
      "mean_token_accuracy": 0.8569764792919159,
      "step": 367
    },
    {
      "epoch": 2.944,
      "grad_norm": 6.191479682922363,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.3207,
      "mean_token_accuracy": 0.8694847077131271,
      "step": 368
    },
    {
      "epoch": 2.952,
      "grad_norm": 5.668253421783447,
      "learning_rate": 1.3232323232323233e-05,
      "loss": 2.4538,
      "mean_token_accuracy": 0.8613507598638535,
      "step": 369
    },
    {
      "epoch": 2.96,
      "grad_norm": 5.68297004699707,
      "learning_rate": 1.3131313131313134e-05,
      "loss": 2.4336,
      "mean_token_accuracy": 0.8691912293434143,
      "step": 370
    },
    {
      "epoch": 2.968,
      "grad_norm": 6.308213710784912,
      "learning_rate": 1.3030303030303032e-05,
      "loss": 2.9029,
      "mean_token_accuracy": 0.8455093801021576,
      "step": 371
    },
    {
      "epoch": 2.976,
      "grad_norm": 5.907562255859375,
      "learning_rate": 1.292929292929293e-05,
      "loss": 2.4898,
      "mean_token_accuracy": 0.8742183148860931,
      "step": 372
    },
    {
      "epoch": 2.984,
      "grad_norm": 6.077236652374268,
      "learning_rate": 1.2828282828282829e-05,
      "loss": 2.5998,
      "mean_token_accuracy": 0.8597747683525085,
      "step": 373
    },
    {
      "epoch": 2.992,
      "grad_norm": 5.781510829925537,
      "learning_rate": 1.2727272727272727e-05,
      "loss": 2.5208,
      "mean_token_accuracy": 0.8634484708309174,
      "step": 374
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.007801532745361,
      "learning_rate": 1.2626262626262628e-05,
      "loss": 2.5073,
      "mean_token_accuracy": 0.8571873605251312,
      "step": 375
    },
    {
      "epoch": 3.008,
      "grad_norm": 5.24554443359375,
      "learning_rate": 1.2525252525252526e-05,
      "loss": 2.189,
      "mean_token_accuracy": 0.8963351398706436,
      "step": 376
    },
    {
      "epoch": 3.016,
      "grad_norm": 5.7106614112854,
      "learning_rate": 1.2424242424242424e-05,
      "loss": 2.1298,
      "mean_token_accuracy": 0.8783598244190216,
      "step": 377
    },
    {
      "epoch": 3.024,
      "grad_norm": 5.2301025390625,
      "learning_rate": 1.2323232323232325e-05,
      "loss": 1.9574,
      "mean_token_accuracy": 0.8882095962762833,
      "step": 378
    },
    {
      "epoch": 3.032,
      "grad_norm": 5.160372257232666,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 2.315,
      "mean_token_accuracy": 0.8809339702129364,
      "step": 379
    },
    {
      "epoch": 3.04,
      "grad_norm": 5.574481010437012,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 2.1501,
      "mean_token_accuracy": 0.8889680802822113,
      "step": 380
    },
    {
      "epoch": 3.048,
      "grad_norm": 6.913398265838623,
      "learning_rate": 1.202020202020202e-05,
      "loss": 2.3129,
      "mean_token_accuracy": 0.8849718272686005,
      "step": 381
    },
    {
      "epoch": 3.056,
      "grad_norm": 6.229403972625732,
      "learning_rate": 1.1919191919191921e-05,
      "loss": 2.3757,
      "mean_token_accuracy": 0.8662046045064926,
      "step": 382
    },
    {
      "epoch": 3.064,
      "grad_norm": 5.60806131362915,
      "learning_rate": 1.1818181818181819e-05,
      "loss": 2.6174,
      "mean_token_accuracy": 0.8517654091119766,
      "step": 383
    },
    {
      "epoch": 3.072,
      "grad_norm": 5.600075721740723,
      "learning_rate": 1.1717171717171718e-05,
      "loss": 2.3305,
      "mean_token_accuracy": 0.8705442994832993,
      "step": 384
    },
    {
      "epoch": 3.08,
      "grad_norm": 6.062666416168213,
      "learning_rate": 1.1616161616161616e-05,
      "loss": 2.2109,
      "mean_token_accuracy": 0.8755911886692047,
      "step": 385
    },
    {
      "epoch": 3.088,
      "grad_norm": 7.154458999633789,
      "learning_rate": 1.1515151515151517e-05,
      "loss": 2.6082,
      "mean_token_accuracy": 0.860401377081871,
      "step": 386
    },
    {
      "epoch": 3.096,
      "grad_norm": 6.247750759124756,
      "learning_rate": 1.1414141414141415e-05,
      "loss": 2.2877,
      "mean_token_accuracy": 0.880204901099205,
      "step": 387
    },
    {
      "epoch": 3.104,
      "grad_norm": 5.955714702606201,
      "learning_rate": 1.1313131313131314e-05,
      "loss": 1.8738,
      "mean_token_accuracy": 0.9012208431959152,
      "step": 388
    },
    {
      "epoch": 3.112,
      "grad_norm": 5.796663284301758,
      "learning_rate": 1.1212121212121212e-05,
      "loss": 1.6587,
      "mean_token_accuracy": 0.9029347151517868,
      "step": 389
    },
    {
      "epoch": 3.12,
      "grad_norm": 5.392340660095215,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 2.3056,
      "mean_token_accuracy": 0.8753447830677032,
      "step": 390
    },
    {
      "epoch": 3.128,
      "grad_norm": 7.139640808105469,
      "learning_rate": 1.1010101010101011e-05,
      "loss": 2.8439,
      "mean_token_accuracy": 0.8585825264453888,
      "step": 391
    },
    {
      "epoch": 3.136,
      "grad_norm": 5.9656829833984375,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 2.1093,
      "mean_token_accuracy": 0.8788597285747528,
      "step": 392
    },
    {
      "epoch": 3.144,
      "grad_norm": 5.839116096496582,
      "learning_rate": 1.0808080808080808e-05,
      "loss": 2.2813,
      "mean_token_accuracy": 0.885636180639267,
      "step": 393
    },
    {
      "epoch": 3.152,
      "grad_norm": 5.032593727111816,
      "learning_rate": 1.0707070707070708e-05,
      "loss": 1.9325,
      "mean_token_accuracy": 0.8939605355262756,
      "step": 394
    },
    {
      "epoch": 3.16,
      "grad_norm": 6.447021961212158,
      "learning_rate": 1.0606060606060607e-05,
      "loss": 2.0899,
      "mean_token_accuracy": 0.8907236605882645,
      "step": 395
    },
    {
      "epoch": 3.168,
      "grad_norm": 6.839473724365234,
      "learning_rate": 1.0505050505050505e-05,
      "loss": 1.8443,
      "mean_token_accuracy": 0.8959701657295227,
      "step": 396
    },
    {
      "epoch": 3.176,
      "grad_norm": 6.076663017272949,
      "learning_rate": 1.0404040404040405e-05,
      "loss": 2.0503,
      "mean_token_accuracy": 0.8750788569450378,
      "step": 397
    },
    {
      "epoch": 3.184,
      "grad_norm": 5.917912483215332,
      "learning_rate": 1.0303030303030304e-05,
      "loss": 2.1561,
      "mean_token_accuracy": 0.8795887529850006,
      "step": 398
    },
    {
      "epoch": 3.192,
      "grad_norm": 6.047354698181152,
      "learning_rate": 1.0202020202020204e-05,
      "loss": 2.4048,
      "mean_token_accuracy": 0.8771449476480484,
      "step": 399
    },
    {
      "epoch": 3.2,
      "grad_norm": 6.2761125564575195,
      "learning_rate": 1.0101010101010101e-05,
      "loss": 2.0975,
      "mean_token_accuracy": 0.8842770159244537,
      "step": 400
    },
    {
      "epoch": 3.208,
      "grad_norm": 6.07623291015625,
      "learning_rate": 1e-05,
      "loss": 1.8668,
      "mean_token_accuracy": 0.8863209933042526,
      "step": 401
    },
    {
      "epoch": 3.216,
      "grad_norm": 6.246818542480469,
      "learning_rate": 9.898989898989899e-06,
      "loss": 2.245,
      "mean_token_accuracy": 0.8780923783779144,
      "step": 402
    },
    {
      "epoch": 3.224,
      "grad_norm": 6.323997974395752,
      "learning_rate": 9.7979797979798e-06,
      "loss": 2.1369,
      "mean_token_accuracy": 0.8779752850532532,
      "step": 403
    },
    {
      "epoch": 3.232,
      "grad_norm": 7.1778411865234375,
      "learning_rate": 9.696969696969698e-06,
      "loss": 2.5985,
      "mean_token_accuracy": 0.8683747202157974,
      "step": 404
    },
    {
      "epoch": 3.24,
      "grad_norm": 6.361847877502441,
      "learning_rate": 9.595959595959595e-06,
      "loss": 2.3375,
      "mean_token_accuracy": 0.8720128089189529,
      "step": 405
    },
    {
      "epoch": 3.248,
      "grad_norm": 6.138617992401123,
      "learning_rate": 9.494949494949495e-06,
      "loss": 2.0752,
      "mean_token_accuracy": 0.8752965927124023,
      "step": 406
    },
    {
      "epoch": 3.2560000000000002,
      "grad_norm": 5.688358783721924,
      "learning_rate": 9.393939393939394e-06,
      "loss": 1.9569,
      "mean_token_accuracy": 0.8975227028131485,
      "step": 407
    },
    {
      "epoch": 3.2640000000000002,
      "grad_norm": 5.894192218780518,
      "learning_rate": 9.292929292929294e-06,
      "loss": 2.0029,
      "mean_token_accuracy": 0.886854737997055,
      "step": 408
    },
    {
      "epoch": 3.2720000000000002,
      "grad_norm": 6.036264419555664,
      "learning_rate": 9.191919191919192e-06,
      "loss": 2.2071,
      "mean_token_accuracy": 0.8780567795038223,
      "step": 409
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 5.998513221740723,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.9301,
      "mean_token_accuracy": 0.8917412608861923,
      "step": 410
    },
    {
      "epoch": 3.288,
      "grad_norm": 6.909607887268066,
      "learning_rate": 8.98989898989899e-06,
      "loss": 2.1684,
      "mean_token_accuracy": 0.871549054980278,
      "step": 411
    },
    {
      "epoch": 3.296,
      "grad_norm": 6.868885040283203,
      "learning_rate": 8.88888888888889e-06,
      "loss": 2.3904,
      "mean_token_accuracy": 0.8834272921085358,
      "step": 412
    },
    {
      "epoch": 3.304,
      "grad_norm": 6.619902610778809,
      "learning_rate": 8.787878787878788e-06,
      "loss": 2.3145,
      "mean_token_accuracy": 0.8664613366127014,
      "step": 413
    },
    {
      "epoch": 3.312,
      "grad_norm": 7.675636291503906,
      "learning_rate": 8.686868686868687e-06,
      "loss": 2.4144,
      "mean_token_accuracy": 0.8673148602247238,
      "step": 414
    },
    {
      "epoch": 3.32,
      "grad_norm": 6.839119911193848,
      "learning_rate": 8.585858585858587e-06,
      "loss": 2.445,
      "mean_token_accuracy": 0.8668957501649857,
      "step": 415
    },
    {
      "epoch": 3.328,
      "grad_norm": 6.37800407409668,
      "learning_rate": 8.484848484848486e-06,
      "loss": 2.0239,
      "mean_token_accuracy": 0.8818509876728058,
      "step": 416
    },
    {
      "epoch": 3.336,
      "grad_norm": 6.840362071990967,
      "learning_rate": 8.383838383838384e-06,
      "loss": 2.2702,
      "mean_token_accuracy": 0.8701023757457733,
      "step": 417
    },
    {
      "epoch": 3.344,
      "grad_norm": 6.7885003089904785,
      "learning_rate": 8.282828282828283e-06,
      "loss": 2.4609,
      "mean_token_accuracy": 0.8644936680793762,
      "step": 418
    },
    {
      "epoch": 3.352,
      "grad_norm": 7.857728481292725,
      "learning_rate": 8.181818181818183e-06,
      "loss": 1.8772,
      "mean_token_accuracy": 0.8961075842380524,
      "step": 419
    },
    {
      "epoch": 3.36,
      "grad_norm": 6.570309162139893,
      "learning_rate": 8.080808080808082e-06,
      "loss": 1.8877,
      "mean_token_accuracy": 0.8875200748443604,
      "step": 420
    },
    {
      "epoch": 3.368,
      "grad_norm": 6.203190803527832,
      "learning_rate": 7.97979797979798e-06,
      "loss": 2.529,
      "mean_token_accuracy": 0.8654111176729202,
      "step": 421
    },
    {
      "epoch": 3.376,
      "grad_norm": 5.97314977645874,
      "learning_rate": 7.878787878787878e-06,
      "loss": 1.9645,
      "mean_token_accuracy": 0.8833965063095093,
      "step": 422
    },
    {
      "epoch": 3.384,
      "grad_norm": 6.25310754776001,
      "learning_rate": 7.777777777777777e-06,
      "loss": 2.2233,
      "mean_token_accuracy": 0.8773595839738846,
      "step": 423
    },
    {
      "epoch": 3.392,
      "grad_norm": 7.28585958480835,
      "learning_rate": 7.676767676767677e-06,
      "loss": 2.0955,
      "mean_token_accuracy": 0.8743046373128891,
      "step": 424
    },
    {
      "epoch": 3.4,
      "grad_norm": 5.4623260498046875,
      "learning_rate": 7.5757575757575764e-06,
      "loss": 2.0769,
      "mean_token_accuracy": 0.886017695069313,
      "step": 425
    },
    {
      "epoch": 3.408,
      "grad_norm": 6.525033950805664,
      "learning_rate": 7.474747474747475e-06,
      "loss": 1.8359,
      "mean_token_accuracy": 0.8860061317682266,
      "step": 426
    },
    {
      "epoch": 3.416,
      "grad_norm": 6.582717418670654,
      "learning_rate": 7.3737373737373745e-06,
      "loss": 1.9729,
      "mean_token_accuracy": 0.8987778276205063,
      "step": 427
    },
    {
      "epoch": 3.424,
      "grad_norm": 6.132490634918213,
      "learning_rate": 7.272727272727272e-06,
      "loss": 1.9637,
      "mean_token_accuracy": 0.883312463760376,
      "step": 428
    },
    {
      "epoch": 3.432,
      "grad_norm": 5.695188045501709,
      "learning_rate": 7.171717171717173e-06,
      "loss": 2.0213,
      "mean_token_accuracy": 0.8891171365976334,
      "step": 429
    },
    {
      "epoch": 3.44,
      "grad_norm": 8.365019798278809,
      "learning_rate": 7.0707070707070704e-06,
      "loss": 2.4396,
      "mean_token_accuracy": 0.8708128333091736,
      "step": 430
    },
    {
      "epoch": 3.448,
      "grad_norm": 5.691287040710449,
      "learning_rate": 6.969696969696971e-06,
      "loss": 2.0038,
      "mean_token_accuracy": 0.9007010161876678,
      "step": 431
    },
    {
      "epoch": 3.456,
      "grad_norm": 6.807380676269531,
      "learning_rate": 6.8686868686868685e-06,
      "loss": 1.9772,
      "mean_token_accuracy": 0.8897037208080292,
      "step": 432
    },
    {
      "epoch": 3.464,
      "grad_norm": 6.571257591247559,
      "learning_rate": 6.767676767676769e-06,
      "loss": 2.5581,
      "mean_token_accuracy": 0.8635726571083069,
      "step": 433
    },
    {
      "epoch": 3.472,
      "grad_norm": 5.491265296936035,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.9038,
      "mean_token_accuracy": 0.9004794210195541,
      "step": 434
    },
    {
      "epoch": 3.48,
      "grad_norm": 5.837332725524902,
      "learning_rate": 6.565656565656567e-06,
      "loss": 1.8962,
      "mean_token_accuracy": 0.890899047255516,
      "step": 435
    },
    {
      "epoch": 3.488,
      "grad_norm": 6.126434803009033,
      "learning_rate": 6.464646464646465e-06,
      "loss": 2.0859,
      "mean_token_accuracy": 0.8814673125743866,
      "step": 436
    },
    {
      "epoch": 3.496,
      "grad_norm": 5.65269660949707,
      "learning_rate": 6.363636363636363e-06,
      "loss": 1.7333,
      "mean_token_accuracy": 0.900140568614006,
      "step": 437
    },
    {
      "epoch": 3.504,
      "grad_norm": 6.0961079597473145,
      "learning_rate": 6.262626262626263e-06,
      "loss": 1.9624,
      "mean_token_accuracy": 0.8910833448171616,
      "step": 438
    },
    {
      "epoch": 3.512,
      "grad_norm": 6.404435157775879,
      "learning_rate": 6.161616161616162e-06,
      "loss": 2.3729,
      "mean_token_accuracy": 0.8687343299388885,
      "step": 439
    },
    {
      "epoch": 3.52,
      "grad_norm": 6.7053093910217285,
      "learning_rate": 6.060606060606061e-06,
      "loss": 2.2333,
      "mean_token_accuracy": 0.8874952048063278,
      "step": 440
    },
    {
      "epoch": 3.528,
      "grad_norm": 6.296670436859131,
      "learning_rate": 5.9595959595959605e-06,
      "loss": 2.1395,
      "mean_token_accuracy": 0.8732293993234634,
      "step": 441
    },
    {
      "epoch": 3.536,
      "grad_norm": 6.392675399780273,
      "learning_rate": 5.858585858585859e-06,
      "loss": 2.0524,
      "mean_token_accuracy": 0.8836774080991745,
      "step": 442
    },
    {
      "epoch": 3.544,
      "grad_norm": 6.301563262939453,
      "learning_rate": 5.7575757575757586e-06,
      "loss": 1.8739,
      "mean_token_accuracy": 0.8968744426965714,
      "step": 443
    },
    {
      "epoch": 3.552,
      "grad_norm": 6.225803852081299,
      "learning_rate": 5.656565656565657e-06,
      "loss": 2.0003,
      "mean_token_accuracy": 0.8827953040599823,
      "step": 444
    },
    {
      "epoch": 3.56,
      "grad_norm": 6.4414801597595215,
      "learning_rate": 5.555555555555556e-06,
      "loss": 2.1538,
      "mean_token_accuracy": 0.8831271827220917,
      "step": 445
    },
    {
      "epoch": 3.568,
      "grad_norm": 8.136292457580566,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 1.7717,
      "mean_token_accuracy": 0.8966427743434906,
      "step": 446
    },
    {
      "epoch": 3.576,
      "grad_norm": 6.468795299530029,
      "learning_rate": 5.353535353535354e-06,
      "loss": 1.9399,
      "mean_token_accuracy": 0.8909775465726852,
      "step": 447
    },
    {
      "epoch": 3.584,
      "grad_norm": 6.0381083488464355,
      "learning_rate": 5.2525252525252526e-06,
      "loss": 2.0466,
      "mean_token_accuracy": 0.8934204578399658,
      "step": 448
    },
    {
      "epoch": 3.592,
      "grad_norm": 6.083271026611328,
      "learning_rate": 5.151515151515152e-06,
      "loss": 2.1554,
      "mean_token_accuracy": 0.88754041492939,
      "step": 449
    },
    {
      "epoch": 3.6,
      "grad_norm": 6.717824935913086,
      "learning_rate": 5.050505050505051e-06,
      "loss": 2.1326,
      "mean_token_accuracy": 0.8911218792200089,
      "step": 450
    },
    {
      "epoch": 3.608,
      "grad_norm": 6.14173698425293,
      "learning_rate": 4.949494949494949e-06,
      "loss": 1.6302,
      "mean_token_accuracy": 0.9065367430448532,
      "step": 451
    },
    {
      "epoch": 3.616,
      "grad_norm": 6.943065643310547,
      "learning_rate": 4.848484848484849e-06,
      "loss": 2.6283,
      "mean_token_accuracy": 0.8691755533218384,
      "step": 452
    },
    {
      "epoch": 3.624,
      "grad_norm": 6.845998287200928,
      "learning_rate": 4.747474747474747e-06,
      "loss": 2.1922,
      "mean_token_accuracy": 0.8834270387887955,
      "step": 453
    },
    {
      "epoch": 3.632,
      "grad_norm": 6.045809268951416,
      "learning_rate": 4.646464646464647e-06,
      "loss": 1.8713,
      "mean_token_accuracy": 0.9000080078840256,
      "step": 454
    },
    {
      "epoch": 3.64,
      "grad_norm": 7.274755477905273,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 2.0781,
      "mean_token_accuracy": 0.8945260941982269,
      "step": 455
    },
    {
      "epoch": 3.648,
      "grad_norm": 6.384909152984619,
      "learning_rate": 4.444444444444445e-06,
      "loss": 2.1529,
      "mean_token_accuracy": 0.8765928000211716,
      "step": 456
    },
    {
      "epoch": 3.656,
      "grad_norm": 5.835945129394531,
      "learning_rate": 4.343434343434344e-06,
      "loss": 1.626,
      "mean_token_accuracy": 0.9043529778718948,
      "step": 457
    },
    {
      "epoch": 3.664,
      "grad_norm": 6.979379177093506,
      "learning_rate": 4.242424242424243e-06,
      "loss": 2.3105,
      "mean_token_accuracy": 0.8783639371395111,
      "step": 458
    },
    {
      "epoch": 3.672,
      "grad_norm": 7.117347240447998,
      "learning_rate": 4.141414141414142e-06,
      "loss": 2.1868,
      "mean_token_accuracy": 0.884697675704956,
      "step": 459
    },
    {
      "epoch": 3.68,
      "grad_norm": 6.831372261047363,
      "learning_rate": 4.040404040404041e-06,
      "loss": 2.1852,
      "mean_token_accuracy": 0.8815398663282394,
      "step": 460
    },
    {
      "epoch": 3.6879999999999997,
      "grad_norm": 6.203273773193359,
      "learning_rate": 3.939393939393939e-06,
      "loss": 2.1329,
      "mean_token_accuracy": 0.8797174096107483,
      "step": 461
    },
    {
      "epoch": 3.6959999999999997,
      "grad_norm": 5.709746837615967,
      "learning_rate": 3.8383838383838385e-06,
      "loss": 2.0121,
      "mean_token_accuracy": 0.8906229883432388,
      "step": 462
    },
    {
      "epoch": 3.7039999999999997,
      "grad_norm": 6.700322151184082,
      "learning_rate": 3.7373737373737375e-06,
      "loss": 2.2537,
      "mean_token_accuracy": 0.8837107121944427,
      "step": 463
    },
    {
      "epoch": 3.7119999999999997,
      "grad_norm": 5.748446464538574,
      "learning_rate": 3.636363636363636e-06,
      "loss": 2.199,
      "mean_token_accuracy": 0.8685683310031891,
      "step": 464
    },
    {
      "epoch": 3.7199999999999998,
      "grad_norm": 6.6395697593688965,
      "learning_rate": 3.5353535353535352e-06,
      "loss": 2.4631,
      "mean_token_accuracy": 0.8770331889390945,
      "step": 465
    },
    {
      "epoch": 3.7279999999999998,
      "grad_norm": 6.521454334259033,
      "learning_rate": 3.4343434343434343e-06,
      "loss": 2.7614,
      "mean_token_accuracy": 0.8615328371524811,
      "step": 466
    },
    {
      "epoch": 3.7359999999999998,
      "grad_norm": 5.956634998321533,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.2834,
      "mean_token_accuracy": 0.8763253539800644,
      "step": 467
    },
    {
      "epoch": 3.7439999999999998,
      "grad_norm": 7.035250186920166,
      "learning_rate": 3.2323232323232324e-06,
      "loss": 2.033,
      "mean_token_accuracy": 0.8936846852302551,
      "step": 468
    },
    {
      "epoch": 3.752,
      "grad_norm": 6.001895904541016,
      "learning_rate": 3.1313131313131314e-06,
      "loss": 2.005,
      "mean_token_accuracy": 0.8896369785070419,
      "step": 469
    },
    {
      "epoch": 3.76,
      "grad_norm": 6.189558506011963,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 2.1698,
      "mean_token_accuracy": 0.8766046166419983,
      "step": 470
    },
    {
      "epoch": 3.768,
      "grad_norm": 6.994729995727539,
      "learning_rate": 2.9292929292929295e-06,
      "loss": 2.108,
      "mean_token_accuracy": 0.8722548186779022,
      "step": 471
    },
    {
      "epoch": 3.776,
      "grad_norm": 5.766332626342773,
      "learning_rate": 2.8282828282828286e-06,
      "loss": 1.7626,
      "mean_token_accuracy": 0.8968187123537064,
      "step": 472
    },
    {
      "epoch": 3.784,
      "grad_norm": 6.6420392990112305,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 2.1263,
      "mean_token_accuracy": 0.8799799233675003,
      "step": 473
    },
    {
      "epoch": 3.792,
      "grad_norm": 6.921932220458984,
      "learning_rate": 2.6262626262626263e-06,
      "loss": 1.8222,
      "mean_token_accuracy": 0.905672699213028,
      "step": 474
    },
    {
      "epoch": 3.8,
      "grad_norm": 6.345583915710449,
      "learning_rate": 2.5252525252525253e-06,
      "loss": 2.3984,
      "mean_token_accuracy": 0.8654076457023621,
      "step": 475
    },
    {
      "epoch": 3.808,
      "grad_norm": 6.0304484367370605,
      "learning_rate": 2.4242424242424244e-06,
      "loss": 2.5167,
      "mean_token_accuracy": 0.8724007159471512,
      "step": 476
    },
    {
      "epoch": 3.816,
      "grad_norm": 6.487928867340088,
      "learning_rate": 2.3232323232323234e-06,
      "loss": 2.5901,
      "mean_token_accuracy": 0.8569000661373138,
      "step": 477
    },
    {
      "epoch": 3.824,
      "grad_norm": 6.210375785827637,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 1.7313,
      "mean_token_accuracy": 0.8982816338539124,
      "step": 478
    },
    {
      "epoch": 3.832,
      "grad_norm": 6.7894182205200195,
      "learning_rate": 2.1212121212121216e-06,
      "loss": 2.121,
      "mean_token_accuracy": 0.8915885388851166,
      "step": 479
    },
    {
      "epoch": 3.84,
      "grad_norm": 6.732741355895996,
      "learning_rate": 2.0202020202020206e-06,
      "loss": 2.0609,
      "mean_token_accuracy": 0.8849688917398453,
      "step": 480
    },
    {
      "epoch": 3.848,
      "grad_norm": 6.280490875244141,
      "learning_rate": 1.9191919191919192e-06,
      "loss": 2.0463,
      "mean_token_accuracy": 0.8867028504610062,
      "step": 481
    },
    {
      "epoch": 3.856,
      "grad_norm": 6.721777439117432,
      "learning_rate": 1.818181818181818e-06,
      "loss": 2.2023,
      "mean_token_accuracy": 0.8840513229370117,
      "step": 482
    },
    {
      "epoch": 3.864,
      "grad_norm": 6.34207010269165,
      "learning_rate": 1.7171717171717171e-06,
      "loss": 2.2238,
      "mean_token_accuracy": 0.8773442953824997,
      "step": 483
    },
    {
      "epoch": 3.872,
      "grad_norm": 6.753549098968506,
      "learning_rate": 1.6161616161616162e-06,
      "loss": 2.1247,
      "mean_token_accuracy": 0.8759674429893494,
      "step": 484
    },
    {
      "epoch": 3.88,
      "grad_norm": 6.267600059509277,
      "learning_rate": 1.5151515151515152e-06,
      "loss": 2.1506,
      "mean_token_accuracy": 0.8819513469934464,
      "step": 485
    },
    {
      "epoch": 3.888,
      "grad_norm": 6.509019374847412,
      "learning_rate": 1.4141414141414143e-06,
      "loss": 2.6317,
      "mean_token_accuracy": 0.8672285228967667,
      "step": 486
    },
    {
      "epoch": 3.896,
      "grad_norm": 6.893660068511963,
      "learning_rate": 1.3131313131313131e-06,
      "loss": 2.1853,
      "mean_token_accuracy": 0.8704394996166229,
      "step": 487
    },
    {
      "epoch": 3.904,
      "grad_norm": 6.314718723297119,
      "learning_rate": 1.2121212121212122e-06,
      "loss": 1.7886,
      "mean_token_accuracy": 0.8912533521652222,
      "step": 488
    },
    {
      "epoch": 3.912,
      "grad_norm": 7.181535243988037,
      "learning_rate": 1.1111111111111112e-06,
      "loss": 2.2651,
      "mean_token_accuracy": 0.8664899617433548,
      "step": 489
    },
    {
      "epoch": 3.92,
      "grad_norm": 6.317063331604004,
      "learning_rate": 1.0101010101010103e-06,
      "loss": 2.3394,
      "mean_token_accuracy": 0.8771510571241379,
      "step": 490
    },
    {
      "epoch": 3.928,
      "grad_norm": 6.195638656616211,
      "learning_rate": 9.09090909090909e-07,
      "loss": 2.0807,
      "mean_token_accuracy": 0.8755189925432205,
      "step": 491
    },
    {
      "epoch": 3.936,
      "grad_norm": 6.2105817794799805,
      "learning_rate": 8.080808080808081e-07,
      "loss": 1.7974,
      "mean_token_accuracy": 0.9031495600938797,
      "step": 492
    },
    {
      "epoch": 3.944,
      "grad_norm": 6.697377681732178,
      "learning_rate": 7.070707070707071e-07,
      "loss": 2.0917,
      "mean_token_accuracy": 0.8885557353496552,
      "step": 493
    },
    {
      "epoch": 3.952,
      "grad_norm": 6.320727825164795,
      "learning_rate": 6.060606060606061e-07,
      "loss": 2.0905,
      "mean_token_accuracy": 0.8876091539859772,
      "step": 494
    },
    {
      "epoch": 3.96,
      "grad_norm": 5.805496692657471,
      "learning_rate": 5.050505050505052e-07,
      "loss": 2.1719,
      "mean_token_accuracy": 0.8817218542098999,
      "step": 495
    },
    {
      "epoch": 3.968,
      "grad_norm": 6.263917446136475,
      "learning_rate": 4.0404040404040405e-07,
      "loss": 1.7863,
      "mean_token_accuracy": 0.8985198885202408,
      "step": 496
    },
    {
      "epoch": 3.976,
      "grad_norm": 5.801756858825684,
      "learning_rate": 3.0303030303030305e-07,
      "loss": 2.0907,
      "mean_token_accuracy": 0.8869260847568512,
      "step": 497
    },
    {
      "epoch": 3.984,
      "grad_norm": 7.1110310554504395,
      "learning_rate": 2.0202020202020202e-07,
      "loss": 2.3715,
      "mean_token_accuracy": 0.8689504116773605,
      "step": 498
    },
    {
      "epoch": 3.992,
      "grad_norm": 7.1602911949157715,
      "learning_rate": 1.0101010101010101e-07,
      "loss": 2.4322,
      "mean_token_accuracy": 0.8692635595798492,
      "step": 499
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.362170696258545,
      "learning_rate": 0.0,
      "loss": 1.9344,
      "mean_token_accuracy": 0.8930138498544693,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7732474675200000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}