{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.004201680672269,
  "eval_steps": 500,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004201680672268907,
      "grad_norm": 1.2855414152145386,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 2.5024,
      "step": 1
    },
    {
      "epoch": 0.008403361344537815,
      "grad_norm": 1.2932429313659668,
      "learning_rate": 4.800000000000001e-06,
      "loss": 2.5834,
      "step": 2
    },
    {
      "epoch": 0.012605042016806723,
      "grad_norm": 1.3544902801513672,
      "learning_rate": 7.2e-06,
      "loss": 2.4398,
      "step": 3
    },
    {
      "epoch": 0.01680672268907563,
      "grad_norm": 1.6224277019500732,
      "learning_rate": 9.600000000000001e-06,
      "loss": 2.6242,
      "step": 4
    },
    {
      "epoch": 0.02100840336134454,
      "grad_norm": 1.5453084707260132,
      "learning_rate": 1.2e-05,
      "loss": 2.2337,
      "step": 5
    },
    {
      "epoch": 0.025210084033613446,
      "grad_norm": 1.8252912759780884,
      "learning_rate": 1.44e-05,
      "loss": 2.3317,
      "step": 6
    },
    {
      "epoch": 0.029411764705882353,
      "grad_norm": 1.2351248264312744,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 2.156,
      "step": 7
    },
    {
      "epoch": 0.03361344537815126,
      "grad_norm": 0.6219255924224854,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.0386,
      "step": 8
    },
    {
      "epoch": 0.037815126050420166,
      "grad_norm": 0.23430386185646057,
      "learning_rate": 2.16e-05,
      "loss": 1.9042,
      "step": 9
    },
    {
      "epoch": 0.04201680672268908,
      "grad_norm": 0.5495060086250305,
      "learning_rate": 2.4e-05,
      "loss": 2.1308,
      "step": 10
    },
    {
      "epoch": 0.046218487394957986,
      "grad_norm": 0.7401595115661621,
      "learning_rate": 2.64e-05,
      "loss": 2.1082,
      "step": 11
    },
    {
      "epoch": 0.05042016806722689,
      "grad_norm": 1.3093836307525635,
      "learning_rate": 2.88e-05,
      "loss": 1.9264,
      "step": 12
    },
    {
      "epoch": 0.0546218487394958,
      "grad_norm": 1.1299002170562744,
      "learning_rate": 3.12e-05,
      "loss": 2.004,
      "step": 13
    },
    {
      "epoch": 0.058823529411764705,
      "grad_norm": 1.1077698469161987,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 2.1267,
      "step": 14
    },
    {
      "epoch": 0.06302521008403361,
      "grad_norm": 0.813391923904419,
      "learning_rate": 3.6e-05,
      "loss": 1.9106,
      "step": 15
    },
    {
      "epoch": 0.06722689075630252,
      "grad_norm": 0.48726755380630493,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 2.055,
      "step": 16
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 0.35943350195884705,
      "learning_rate": 4.08e-05,
      "loss": 2.0779,
      "step": 17
    },
    {
      "epoch": 0.07563025210084033,
      "grad_norm": 0.27760088443756104,
      "learning_rate": 4.32e-05,
      "loss": 2.0189,
      "step": 18
    },
    {
      "epoch": 0.07983193277310924,
      "grad_norm": 0.2671177089214325,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 1.9943,
      "step": 19
    },
    {
      "epoch": 0.08403361344537816,
      "grad_norm": 0.2401943951845169,
      "learning_rate": 4.8e-05,
      "loss": 2.032,
      "step": 20
    },
    {
      "epoch": 0.08823529411764706,
      "grad_norm": 0.18362487852573395,
      "learning_rate": 5.04e-05,
      "loss": 1.9384,
      "step": 21
    },
    {
      "epoch": 0.09243697478991597,
      "grad_norm": 0.18155354261398315,
      "learning_rate": 5.28e-05,
      "loss": 1.9294,
      "step": 22
    },
    {
      "epoch": 0.09663865546218488,
      "grad_norm": 0.311981201171875,
      "learning_rate": 5.520000000000001e-05,
      "loss": 2.1202,
      "step": 23
    },
    {
      "epoch": 0.10084033613445378,
      "grad_norm": 0.46018195152282715,
      "learning_rate": 5.76e-05,
      "loss": 2.0634,
      "step": 24
    },
    {
      "epoch": 0.10504201680672269,
      "grad_norm": 0.725566565990448,
      "learning_rate": 6e-05,
      "loss": 2.0021,
      "step": 25
    },
    {
      "epoch": 0.1092436974789916,
      "grad_norm": 0.3912893235683441,
      "learning_rate": 5.999927215958923e-05,
      "loss": 2.0414,
      "step": 26
    },
    {
      "epoch": 0.1134453781512605,
      "grad_norm": 0.3835509121417999,
      "learning_rate": 5.999708867367372e-05,
      "loss": 2.0237,
      "step": 27
    },
    {
      "epoch": 0.11764705882352941,
      "grad_norm": 0.39111143350601196,
      "learning_rate": 5.999344964820206e-05,
      "loss": 2.0386,
      "step": 28
    },
    {
      "epoch": 0.12184873949579832,
      "grad_norm": 0.31010451912879944,
      "learning_rate": 5.998835525974959e-05,
      "loss": 2.0222,
      "step": 29
    },
    {
      "epoch": 0.12605042016806722,
      "grad_norm": 0.36277830600738525,
      "learning_rate": 5.9981805755509766e-05,
      "loss": 1.9205,
      "step": 30
    },
    {
      "epoch": 0.13025210084033614,
      "grad_norm": 0.24080567061901093,
      "learning_rate": 5.9973801453282154e-05,
      "loss": 1.8759,
      "step": 31
    },
    {
      "epoch": 0.13445378151260504,
      "grad_norm": 0.22599194943904877,
      "learning_rate": 5.9964342741457086e-05,
      "loss": 1.9547,
      "step": 32
    },
    {
      "epoch": 0.13865546218487396,
      "grad_norm": 0.33472496271133423,
      "learning_rate": 5.9953430078996735e-05,
      "loss": 1.7979,
      "step": 33
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.275795578956604,
      "learning_rate": 5.994106399541288e-05,
      "loss": 1.9714,
      "step": 34
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 0.1729828417301178,
      "learning_rate": 5.992724509074121e-05,
      "loss": 2.0898,
      "step": 35
    },
    {
      "epoch": 0.15126050420168066,
      "grad_norm": 0.11706062406301498,
      "learning_rate": 5.991197403551221e-05,
      "loss": 1.8647,
      "step": 36
    },
    {
      "epoch": 0.15546218487394958,
      "grad_norm": 0.18395456671714783,
      "learning_rate": 5.989525157071863e-05,
      "loss": 1.9258,
      "step": 37
    },
    {
      "epoch": 0.15966386554621848,
      "grad_norm": 0.14259740710258484,
      "learning_rate": 5.98770785077795e-05,
      "loss": 1.8656,
      "step": 38
    },
    {
      "epoch": 0.1638655462184874,
      "grad_norm": 0.15977182984352112,
      "learning_rate": 5.98574557285008e-05,
      "loss": 2.0456,
      "step": 39
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 0.1527746468782425,
      "learning_rate": 5.9836384185032645e-05,
      "loss": 1.8477,
      "step": 40
    },
    {
      "epoch": 0.1722689075630252,
      "grad_norm": 0.18175554275512695,
      "learning_rate": 5.9813864899823085e-05,
      "loss": 1.8427,
      "step": 41
    },
    {
      "epoch": 0.17647058823529413,
      "grad_norm": 0.19280438125133514,
      "learning_rate": 5.978989896556851e-05,
      "loss": 1.9003,
      "step": 42
    },
    {
      "epoch": 0.18067226890756302,
      "grad_norm": 0.1241818442940712,
      "learning_rate": 5.976448754516063e-05,
      "loss": 1.7676,
      "step": 43
    },
    {
      "epoch": 0.18487394957983194,
      "grad_norm": 0.23351380228996277,
      "learning_rate": 5.973763187163e-05,
      "loss": 1.8321,
      "step": 44
    },
    {
      "epoch": 0.18907563025210083,
      "grad_norm": 0.11710195988416672,
      "learning_rate": 5.970933324808626e-05,
      "loss": 1.9762,
      "step": 45
    },
    {
      "epoch": 0.19327731092436976,
      "grad_norm": 0.1551787108182907,
      "learning_rate": 5.967959304765487e-05,
      "loss": 2.003,
      "step": 46
    },
    {
      "epoch": 0.19747899159663865,
      "grad_norm": 0.12680837512016296,
      "learning_rate": 5.964841271341046e-05,
      "loss": 1.909,
      "step": 47
    },
    {
      "epoch": 0.20168067226890757,
      "grad_norm": 0.14685605466365814,
      "learning_rate": 5.961579375830686e-05,
      "loss": 1.9564,
      "step": 48
    },
    {
      "epoch": 0.20588235294117646,
      "grad_norm": 0.2090328335762024,
      "learning_rate": 5.958173776510365e-05,
      "loss": 2.0247,
      "step": 49
    },
    {
      "epoch": 0.21008403361344538,
      "grad_norm": 0.15236154198646545,
      "learning_rate": 5.9546246386289364e-05,
      "loss": 1.8914,
      "step": 50
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.13934959471225739,
      "learning_rate": 5.950932134400132e-05,
      "loss": 1.9527,
      "step": 51
    },
    {
      "epoch": 0.2184873949579832,
      "grad_norm": 0.14951136708259583,
      "learning_rate": 5.947096442994205e-05,
      "loss": 2.0099,
      "step": 52
    },
    {
      "epoch": 0.22268907563025211,
      "grad_norm": 0.16193124651908875,
      "learning_rate": 5.9431177505292346e-05,
      "loss": 1.9779,
      "step": 53
    },
    {
      "epoch": 0.226890756302521,
      "grad_norm": 0.12261511385440826,
      "learning_rate": 5.9389962500621e-05,
      "loss": 1.7994,
      "step": 54
    },
    {
      "epoch": 0.23109243697478993,
      "grad_norm": 0.13601715862751007,
      "learning_rate": 5.934732141579106e-05,
      "loss": 1.8998,
      "step": 55
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 0.16675348579883575,
      "learning_rate": 5.930325631986285e-05,
      "loss": 1.9909,
      "step": 56
    },
    {
      "epoch": 0.23949579831932774,
      "grad_norm": 0.16077667474746704,
      "learning_rate": 5.925776935099353e-05,
      "loss": 1.7902,
      "step": 57
    },
    {
      "epoch": 0.24369747899159663,
      "grad_norm": 0.12847718596458435,
      "learning_rate": 5.9210862716333373e-05,
      "loss": 1.8161,
      "step": 58
    },
    {
      "epoch": 0.24789915966386555,
      "grad_norm": 0.15590892732143402,
      "learning_rate": 5.916253869191867e-05,
      "loss": 2.0115,
      "step": 59
    },
    {
      "epoch": 0.25210084033613445,
      "grad_norm": 0.16642436385154724,
      "learning_rate": 5.911279962256126e-05,
      "loss": 2.0116,
      "step": 60
    },
    {
      "epoch": 0.25630252100840334,
      "grad_norm": 0.12253208458423615,
      "learning_rate": 5.9061647921734794e-05,
      "loss": 1.9297,
      "step": 61
    },
    {
      "epoch": 0.2605042016806723,
      "grad_norm": 0.13511379063129425,
      "learning_rate": 5.900908607145761e-05,
      "loss": 1.8262,
      "step": 62
    },
    {
      "epoch": 0.2647058823529412,
      "grad_norm": 0.1308983564376831,
      "learning_rate": 5.8955116622172274e-05,
      "loss": 1.9066,
      "step": 63
    },
    {
      "epoch": 0.2689075630252101,
      "grad_norm": 0.1284906417131424,
      "learning_rate": 5.8899742192621875e-05,
      "loss": 1.8297,
      "step": 64
    },
    {
      "epoch": 0.27310924369747897,
      "grad_norm": 0.14260078966617584,
      "learning_rate": 5.88429654697229e-05,
      "loss": 2.0597,
      "step": 65
    },
    {
      "epoch": 0.2773109243697479,
      "grad_norm": 0.13550403714179993,
      "learning_rate": 5.878478920843492e-05,
      "loss": 2.1711,
      "step": 66
    },
    {
      "epoch": 0.2815126050420168,
      "grad_norm": 0.1789373755455017,
      "learning_rate": 5.872521623162686e-05,
      "loss": 1.6575,
      "step": 67
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.12788023054599762,
      "learning_rate": 5.866424942994004e-05,
      "loss": 1.9221,
      "step": 68
    },
    {
      "epoch": 0.28991596638655465,
      "grad_norm": 0.18658475577831268,
      "learning_rate": 5.8601891761647914e-05,
      "loss": 1.9013,
      "step": 69
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.18295790255069733,
      "learning_rate": 5.853814625251257e-05,
      "loss": 1.9291,
      "step": 70
    },
    {
      "epoch": 0.29831932773109243,
      "grad_norm": 0.12854939699172974,
      "learning_rate": 5.847301599563783e-05,
      "loss": 2.0076,
      "step": 71
    },
    {
      "epoch": 0.3025210084033613,
      "grad_norm": 0.17280517518520355,
      "learning_rate": 5.840650415131922e-05,
      "loss": 1.8706,
      "step": 72
    },
    {
      "epoch": 0.3067226890756303,
      "grad_norm": 0.15926142036914825,
      "learning_rate": 5.8338613946890617e-05,
      "loss": 1.9854,
      "step": 73
    },
    {
      "epoch": 0.31092436974789917,
      "grad_norm": 0.2687736451625824,
      "learning_rate": 5.826934867656765e-05,
      "loss": 1.6384,
      "step": 74
    },
    {
      "epoch": 0.31512605042016806,
      "grad_norm": 0.11397445946931839,
      "learning_rate": 5.819871170128781e-05,
      "loss": 1.8928,
      "step": 75
    },
    {
      "epoch": 0.31932773109243695,
      "grad_norm": 0.12926487624645233,
      "learning_rate": 5.812670644854745e-05,
      "loss": 1.8773,
      "step": 76
    },
    {
      "epoch": 0.3235294117647059,
      "grad_norm": 0.14260263741016388,
      "learning_rate": 5.8053336412235445e-05,
      "loss": 2.0524,
      "step": 77
    },
    {
      "epoch": 0.3277310924369748,
      "grad_norm": 0.14220871031284332,
      "learning_rate": 5.7978605152463594e-05,
      "loss": 1.7581,
      "step": 78
    },
    {
      "epoch": 0.3319327731092437,
      "grad_norm": 0.1228056475520134,
      "learning_rate": 5.7902516295393945e-05,
      "loss": 1.9749,
      "step": 79
    },
    {
      "epoch": 0.33613445378151263,
      "grad_norm": 0.12861187756061554,
      "learning_rate": 5.782507353306285e-05,
      "loss": 1.7885,
      "step": 80
    },
    {
      "epoch": 0.3403361344537815,
      "grad_norm": 0.12451905757188797,
      "learning_rate": 5.774628062320175e-05,
      "loss": 1.899,
      "step": 81
    },
    {
      "epoch": 0.3445378151260504,
      "grad_norm": 0.1419980823993683,
      "learning_rate": 5.7666141389054923e-05,
      "loss": 1.8726,
      "step": 82
    },
    {
      "epoch": 0.3487394957983193,
      "grad_norm": 0.18138334155082703,
      "learning_rate": 5.75846597191939e-05,
      "loss": 1.8988,
      "step": 83
    },
    {
      "epoch": 0.35294117647058826,
      "grad_norm": 0.2072378247976303,
      "learning_rate": 5.750183956732882e-05,
      "loss": 1.8585,
      "step": 84
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.15298499166965485,
      "learning_rate": 5.741768495211658e-05,
      "loss": 1.9546,
      "step": 85
    },
    {
      "epoch": 0.36134453781512604,
      "grad_norm": 0.15661871433258057,
      "learning_rate": 5.733219995696581e-05,
      "loss": 1.8402,
      "step": 86
    },
    {
      "epoch": 0.36554621848739494,
      "grad_norm": 0.13822637498378754,
      "learning_rate": 5.7245388729838796e-05,
      "loss": 2.0029,
      "step": 87
    },
    {
      "epoch": 0.3697478991596639,
      "grad_norm": 0.14563272893428802,
      "learning_rate": 5.715725548305014e-05,
      "loss": 2.0999,
      "step": 88
    },
    {
      "epoch": 0.3739495798319328,
      "grad_norm": 0.15243449807167053,
      "learning_rate": 5.706780449306241e-05,
      "loss": 2.1257,
      "step": 89
    },
    {
      "epoch": 0.37815126050420167,
      "grad_norm": 0.16341632604599,
      "learning_rate": 5.697704010027864e-05,
      "loss": 2.0414,
      "step": 90
    },
    {
      "epoch": 0.38235294117647056,
      "grad_norm": 0.15627144277095795,
      "learning_rate": 5.688496670883167e-05,
      "loss": 1.9314,
      "step": 91
    },
    {
      "epoch": 0.3865546218487395,
      "grad_norm": 0.12584370374679565,
      "learning_rate": 5.6791588786370515e-05,
      "loss": 1.796,
      "step": 92
    },
    {
      "epoch": 0.3907563025210084,
      "grad_norm": 0.15590286254882812,
      "learning_rate": 5.6696910863843546e-05,
      "loss": 1.9411,
      "step": 93
    },
    {
      "epoch": 0.3949579831932773,
      "grad_norm": 0.12517981231212616,
      "learning_rate": 5.660093753527861e-05,
      "loss": 1.9865,
      "step": 94
    },
    {
      "epoch": 0.39915966386554624,
      "grad_norm": 0.1500655859708786,
      "learning_rate": 5.650367345756019e-05,
      "loss": 1.7765,
      "step": 95
    },
    {
      "epoch": 0.40336134453781514,
      "grad_norm": 0.1287829726934433,
      "learning_rate": 5.640512335020336e-05,
      "loss": 1.9496,
      "step": 96
    },
    {
      "epoch": 0.40756302521008403,
      "grad_norm": 0.18750888109207153,
      "learning_rate": 5.6305291995124827e-05,
      "loss": 1.7406,
      "step": 97
    },
    {
      "epoch": 0.4117647058823529,
      "grad_norm": 0.12096212059259415,
      "learning_rate": 5.620418423641089e-05,
      "loss": 2.0335,
      "step": 98
    },
    {
      "epoch": 0.41596638655462187,
      "grad_norm": 0.13212282955646515,
      "learning_rate": 5.610180498008239e-05,
      "loss": 1.8856,
      "step": 99
    },
    {
      "epoch": 0.42016806722689076,
      "grad_norm": 0.14193859696388245,
      "learning_rate": 5.599815919385666e-05,
      "loss": 1.9243,
      "step": 100
    },
    {
      "epoch": 0.42436974789915966,
      "grad_norm": 0.144227996468544,
      "learning_rate": 5.589325190690648e-05,
      "loss": 1.9885,
      "step": 101
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.14424113929271698,
      "learning_rate": 5.578708820961604e-05,
      "loss": 1.8761,
      "step": 102
    },
    {
      "epoch": 0.4327731092436975,
      "grad_norm": 0.14241594076156616,
      "learning_rate": 5.567967325333393e-05,
      "loss": 1.8844,
      "step": 103
    },
    {
      "epoch": 0.4369747899159664,
      "grad_norm": 0.23358209431171417,
      "learning_rate": 5.557101225012321e-05,
      "loss": 1.7688,
      "step": 104
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 0.13762815296649933,
      "learning_rate": 5.546111047250852e-05,
      "loss": 1.9983,
      "step": 105
    },
    {
      "epoch": 0.44537815126050423,
      "grad_norm": 0.15425845980644226,
      "learning_rate": 5.5349973253220164e-05,
      "loss": 1.6622,
      "step": 106
    },
    {
      "epoch": 0.4495798319327731,
      "grad_norm": 0.142578125,
      "learning_rate": 5.5237605984935435e-05,
      "loss": 1.9803,
      "step": 107
    },
    {
      "epoch": 0.453781512605042,
      "grad_norm": 0.13661649823188782,
      "learning_rate": 5.512401412001692e-05,
      "loss": 1.8166,
      "step": 108
    },
    {
      "epoch": 0.4579831932773109,
      "grad_norm": 0.1451268047094345,
      "learning_rate": 5.5009203170247934e-05,
      "loss": 1.7331,
      "step": 109
    },
    {
      "epoch": 0.46218487394957986,
      "grad_norm": 0.14962796866893768,
      "learning_rate": 5.4893178706565053e-05,
      "loss": 1.9284,
      "step": 110
    },
    {
      "epoch": 0.46638655462184875,
      "grad_norm": 0.14141930639743805,
      "learning_rate": 5.477594635878783e-05,
      "loss": 2.0571,
      "step": 111
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 0.1254950761795044,
      "learning_rate": 5.465751181534562e-05,
      "loss": 1.9862,
      "step": 112
    },
    {
      "epoch": 0.47478991596638653,
      "grad_norm": 0.1633477658033371,
      "learning_rate": 5.4537880823001535e-05,
      "loss": 1.8533,
      "step": 113
    },
    {
      "epoch": 0.4789915966386555,
      "grad_norm": 0.1374650001525879,
      "learning_rate": 5.4417059186573615e-05,
      "loss": 1.9262,
      "step": 114
    },
    {
      "epoch": 0.4831932773109244,
      "grad_norm": 0.16950736939907074,
      "learning_rate": 5.429505276865315e-05,
      "loss": 1.9775,
      "step": 115
    },
    {
      "epoch": 0.48739495798319327,
      "grad_norm": 0.14926834404468536,
      "learning_rate": 5.417186748932024e-05,
      "loss": 1.7784,
      "step": 116
    },
    {
      "epoch": 0.49159663865546216,
      "grad_norm": 0.16016586124897003,
      "learning_rate": 5.4047509325856514e-05,
      "loss": 1.9489,
      "step": 117
    },
    {
      "epoch": 0.4957983193277311,
      "grad_norm": 0.14702598750591278,
      "learning_rate": 5.392198431245507e-05,
      "loss": 1.717,
      "step": 118
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.13934609293937683,
      "learning_rate": 5.379529853992774e-05,
      "loss": 1.9416,
      "step": 119
    },
    {
      "epoch": 0.5042016806722689,
      "grad_norm": 0.1452355682849884,
      "learning_rate": 5.3667458155409505e-05,
      "loss": 1.927,
      "step": 120
    },
    {
      "epoch": 0.5084033613445378,
      "grad_norm": 0.14766626060009003,
      "learning_rate": 5.3538469362060226e-05,
      "loss": 1.7578,
      "step": 121
    },
    {
      "epoch": 0.5126050420168067,
      "grad_norm": 0.19430850446224213,
      "learning_rate": 5.340833841876367e-05,
      "loss": 1.8113,
      "step": 122
    },
    {
      "epoch": 0.5168067226890757,
      "grad_norm": 0.16308219730854034,
      "learning_rate": 5.327707163982377e-05,
      "loss": 1.7603,
      "step": 123
    },
    {
      "epoch": 0.5210084033613446,
      "grad_norm": 0.1342669427394867,
      "learning_rate": 5.314467539465829e-05,
      "loss": 2.0409,
      "step": 124
    },
    {
      "epoch": 0.5252100840336135,
      "grad_norm": 0.12464125454425812,
      "learning_rate": 5.3011156107489726e-05,
      "loss": 1.92,
      "step": 125
    },
    {
      "epoch": 0.5294117647058824,
      "grad_norm": 0.1477716565132141,
      "learning_rate": 5.2876520257033594e-05,
      "loss": 1.7391,
      "step": 126
    },
    {
      "epoch": 0.5336134453781513,
      "grad_norm": 0.15375854074954987,
      "learning_rate": 5.274077437618408e-05,
      "loss": 1.9802,
      "step": 127
    },
    {
      "epoch": 0.5378151260504201,
      "grad_norm": 0.17319805920124054,
      "learning_rate": 5.260392505169702e-05,
      "loss": 1.9145,
      "step": 128
    },
    {
      "epoch": 0.542016806722689,
      "grad_norm": 0.13694263994693756,
      "learning_rate": 5.246597892387033e-05,
      "loss": 1.8611,
      "step": 129
    },
    {
      "epoch": 0.5462184873949579,
      "grad_norm": 0.14227637648582458,
      "learning_rate": 5.2326942686221754e-05,
      "loss": 1.9588,
      "step": 130
    },
    {
      "epoch": 0.5504201680672269,
      "grad_norm": 0.13238848745822906,
      "learning_rate": 5.218682308516413e-05,
      "loss": 2.0077,
      "step": 131
    },
    {
      "epoch": 0.5546218487394958,
      "grad_norm": 0.1516510248184204,
      "learning_rate": 5.2045626919677975e-05,
      "loss": 2.0937,
      "step": 132
    },
    {
      "epoch": 0.5588235294117647,
      "grad_norm": 0.15112727880477905,
      "learning_rate": 5.1903361040981624e-05,
      "loss": 1.8754,
      "step": 133
    },
    {
      "epoch": 0.5630252100840336,
      "grad_norm": 0.1264156699180603,
      "learning_rate": 5.17600323521988e-05,
      "loss": 1.8644,
      "step": 134
    },
    {
      "epoch": 0.5672268907563025,
      "grad_norm": 0.19539165496826172,
      "learning_rate": 5.161564780802361e-05,
      "loss": 1.9111,
      "step": 135
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.1472083181142807,
      "learning_rate": 5.147021441438312e-05,
      "loss": 1.8196,
      "step": 136
    },
    {
      "epoch": 0.5756302521008403,
      "grad_norm": 0.16398820281028748,
      "learning_rate": 5.132373922809738e-05,
      "loss": 1.9675,
      "step": 137
    },
    {
      "epoch": 0.5798319327731093,
      "grad_norm": 0.17623396217823029,
      "learning_rate": 5.117622935653706e-05,
      "loss": 1.8302,
      "step": 138
    },
    {
      "epoch": 0.5840336134453782,
      "grad_norm": 0.1819780021905899,
      "learning_rate": 5.1027691957278515e-05,
      "loss": 1.6874,
      "step": 139
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.1527121514081955,
      "learning_rate": 5.087813423775653e-05,
      "loss": 1.7856,
      "step": 140
    },
    {
      "epoch": 0.592436974789916,
      "grad_norm": 0.18883730471134186,
      "learning_rate": 5.0727563454914566e-05,
      "loss": 2.0023,
      "step": 141
    },
    {
      "epoch": 0.5966386554621849,
      "grad_norm": 0.1411568820476532,
      "learning_rate": 5.057598691485265e-05,
      "loss": 1.939,
      "step": 142
    },
    {
      "epoch": 0.6008403361344538,
      "grad_norm": 0.1455598771572113,
      "learning_rate": 5.042341197247288e-05,
      "loss": 1.9601,
      "step": 143
    },
    {
      "epoch": 0.6050420168067226,
      "grad_norm": 0.191257044672966,
      "learning_rate": 5.026984603112248e-05,
      "loss": 1.8474,
      "step": 144
    },
    {
      "epoch": 0.6092436974789915,
      "grad_norm": 0.1409291923046112,
      "learning_rate": 5.011529654223465e-05,
      "loss": 1.9115,
      "step": 145
    },
    {
      "epoch": 0.6134453781512605,
      "grad_norm": 0.1498507708311081,
      "learning_rate": 4.995977100496696e-05,
      "loss": 1.8116,
      "step": 146
    },
    {
      "epoch": 0.6176470588235294,
      "grad_norm": 0.1509602963924408,
      "learning_rate": 4.980327696583746e-05,
      "loss": 1.9574,
      "step": 147
    },
    {
      "epoch": 0.6218487394957983,
      "grad_norm": 0.15841971337795258,
      "learning_rate": 4.964582201835856e-05,
      "loss": 1.9305,
      "step": 148
    },
    {
      "epoch": 0.6260504201680672,
      "grad_norm": 0.154954195022583,
      "learning_rate": 4.948741380266846e-05,
      "loss": 1.6991,
      "step": 149
    },
    {
      "epoch": 0.6302521008403361,
      "grad_norm": 0.1550845056772232,
      "learning_rate": 4.9328060005160575e-05,
      "loss": 1.8016,
      "step": 150
    },
    {
      "epoch": 0.634453781512605,
      "grad_norm": 0.15357814729213715,
      "learning_rate": 4.916776835811046e-05,
      "loss": 2.1041,
      "step": 151
    },
    {
      "epoch": 0.6386554621848739,
      "grad_norm": 0.16955037415027618,
      "learning_rate": 4.9006546639300656e-05,
      "loss": 2.0618,
      "step": 152
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 0.1576966941356659,
      "learning_rate": 4.884440267164331e-05,
      "loss": 1.7933,
      "step": 153
    },
    {
      "epoch": 0.6470588235294118,
      "grad_norm": 0.15891976654529572,
      "learning_rate": 4.868134432280056e-05,
      "loss": 1.7879,
      "step": 154
    },
    {
      "epoch": 0.6512605042016807,
      "grad_norm": 0.16660089790821075,
      "learning_rate": 4.851737950480276e-05,
      "loss": 1.8625,
      "step": 155
    },
    {
      "epoch": 0.6554621848739496,
      "grad_norm": 0.23937979340553284,
      "learning_rate": 4.835251617366463e-05,
      "loss": 1.8112,
      "step": 156
    },
    {
      "epoch": 0.6596638655462185,
      "grad_norm": 0.1381472498178482,
      "learning_rate": 4.8186762328999144e-05,
      "loss": 1.9289,
      "step": 157
    },
    {
      "epoch": 0.6638655462184874,
      "grad_norm": 0.1760849952697754,
      "learning_rate": 4.802012601362938e-05,
      "loss": 1.9384,
      "step": 158
    },
    {
      "epoch": 0.6680672268907563,
      "grad_norm": 0.18892566859722137,
      "learning_rate": 4.785261531319829e-05,
      "loss": 1.8544,
      "step": 159
    },
    {
      "epoch": 0.6722689075630253,
      "grad_norm": 0.18952800333499908,
      "learning_rate": 4.7684238355776366e-05,
      "loss": 1.7216,
      "step": 160
    },
    {
      "epoch": 0.6764705882352942,
      "grad_norm": 0.1533077359199524,
      "learning_rate": 4.7515003311467164e-05,
      "loss": 1.924,
      "step": 161
    },
    {
      "epoch": 0.680672268907563,
      "grad_norm": 0.20900940895080566,
      "learning_rate": 4.7344918392010985e-05,
      "loss": 1.8922,
      "step": 162
    },
    {
      "epoch": 0.6848739495798319,
      "grad_norm": 0.22550180554389954,
      "learning_rate": 4.7173991850386325e-05,
      "loss": 1.9899,
      "step": 163
    },
    {
      "epoch": 0.6890756302521008,
      "grad_norm": 0.15734222531318665,
      "learning_rate": 4.7002231980409466e-05,
      "loss": 1.9437,
      "step": 164
    },
    {
      "epoch": 0.6932773109243697,
      "grad_norm": 0.15735475718975067,
      "learning_rate": 4.6829647116332045e-05,
      "loss": 1.9653,
      "step": 165
    },
    {
      "epoch": 0.6974789915966386,
      "grad_norm": 0.19688263535499573,
      "learning_rate": 4.665624563243661e-05,
      "loss": 1.9832,
      "step": 166
    },
    {
      "epoch": 0.7016806722689075,
      "grad_norm": 0.1678629070520401,
      "learning_rate": 4.648203594263031e-05,
      "loss": 1.7721,
      "step": 167
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 0.16972413659095764,
      "learning_rate": 4.630702650003664e-05,
      "loss": 1.8245,
      "step": 168
    },
    {
      "epoch": 0.7100840336134454,
      "grad_norm": 0.17227919399738312,
      "learning_rate": 4.613122579658522e-05,
      "loss": 1.8539,
      "step": 169
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.20971490442752838,
      "learning_rate": 4.595464236259979e-05,
      "loss": 1.7229,
      "step": 170
    },
    {
      "epoch": 0.7184873949579832,
      "grad_norm": 0.2913236618041992,
      "learning_rate": 4.577728476638432e-05,
      "loss": 1.8346,
      "step": 171
    },
    {
      "epoch": 0.7226890756302521,
      "grad_norm": 0.16503480076789856,
      "learning_rate": 4.559916161380718e-05,
      "loss": 1.7839,
      "step": 172
    },
    {
      "epoch": 0.726890756302521,
      "grad_norm": 0.2223767340183258,
      "learning_rate": 4.542028154788359e-05,
      "loss": 1.8485,
      "step": 173
    },
    {
      "epoch": 0.7310924369747899,
      "grad_norm": 0.3713437616825104,
      "learning_rate": 4.5240653248356274e-05,
      "loss": 1.8596,
      "step": 174
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 0.1753188520669937,
      "learning_rate": 4.506028543127425e-05,
      "loss": 1.7406,
      "step": 175
    },
    {
      "epoch": 0.7394957983193278,
      "grad_norm": 0.15702295303344727,
      "learning_rate": 4.4879186848569916e-05,
      "loss": 2.1146,
      "step": 176
    },
    {
      "epoch": 0.7436974789915967,
      "grad_norm": 0.17345549166202545,
      "learning_rate": 4.469736628763441e-05,
      "loss": 1.9998,
      "step": 177
    },
    {
      "epoch": 0.7478991596638656,
      "grad_norm": 0.3005940914154053,
      "learning_rate": 4.4514832570891164e-05,
      "loss": 1.8081,
      "step": 178
    },
    {
      "epoch": 0.7521008403361344,
      "grad_norm": 0.19223691523075104,
      "learning_rate": 4.433159455536789e-05,
      "loss": 1.7473,
      "step": 179
    },
    {
      "epoch": 0.7563025210084033,
      "grad_norm": 0.15088941156864166,
      "learning_rate": 4.414766113226673e-05,
      "loss": 1.8934,
      "step": 180
    },
    {
      "epoch": 0.7605042016806722,
      "grad_norm": 0.28544843196868896,
      "learning_rate": 4.396304122653292e-05,
      "loss": 1.7362,
      "step": 181
    },
    {
      "epoch": 0.7647058823529411,
      "grad_norm": 0.27440345287323,
      "learning_rate": 4.377774379642165e-05,
      "loss": 1.7409,
      "step": 182
    },
    {
      "epoch": 0.7689075630252101,
      "grad_norm": 0.16212020814418793,
      "learning_rate": 4.359177783306344e-05,
      "loss": 1.9367,
      "step": 183
    },
    {
      "epoch": 0.773109243697479,
      "grad_norm": 0.15055981278419495,
      "learning_rate": 4.3405152360027815e-05,
      "loss": 1.9787,
      "step": 184
    },
    {
      "epoch": 0.7773109243697479,
      "grad_norm": 0.28331828117370605,
      "learning_rate": 4.3217876432885525e-05,
      "loss": 1.6631,
      "step": 185
    },
    {
      "epoch": 0.7815126050420168,
      "grad_norm": 0.18612366914749146,
      "learning_rate": 4.302995913876907e-05,
      "loss": 2.0595,
      "step": 186
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 0.16861191391944885,
      "learning_rate": 4.284140959593184e-05,
      "loss": 1.6634,
      "step": 187
    },
    {
      "epoch": 0.7899159663865546,
      "grad_norm": 0.20200952887535095,
      "learning_rate": 4.265223695330559e-05,
      "loss": 1.8512,
      "step": 188
    },
    {
      "epoch": 0.7941176470588235,
      "grad_norm": 0.17656445503234863,
      "learning_rate": 4.24624503900566e-05,
      "loss": 1.8023,
      "step": 189
    },
    {
      "epoch": 0.7983193277310925,
      "grad_norm": 0.15888820588588715,
      "learning_rate": 4.22720591151402e-05,
      "loss": 1.9317,
      "step": 190
    },
    {
      "epoch": 0.8025210084033614,
      "grad_norm": 0.19560733437538147,
      "learning_rate": 4.2081072366853984e-05,
      "loss": 1.527,
      "step": 191
    },
    {
      "epoch": 0.8067226890756303,
      "grad_norm": 0.19005230069160461,
      "learning_rate": 4.18894994123895e-05,
      "loss": 1.9566,
      "step": 192
    },
    {
      "epoch": 0.8109243697478992,
      "grad_norm": 0.162759929895401,
      "learning_rate": 4.169734954738261e-05,
      "loss": 2.0595,
      "step": 193
    },
    {
      "epoch": 0.8151260504201681,
      "grad_norm": 0.26847419142723083,
      "learning_rate": 4.150463209546243e-05,
      "loss": 1.7949,
      "step": 194
    },
    {
      "epoch": 0.819327731092437,
      "grad_norm": 0.18686293065547943,
      "learning_rate": 4.13113564077989e-05,
      "loss": 1.6174,
      "step": 195
    },
    {
      "epoch": 0.8235294117647058,
      "grad_norm": 0.15110686421394348,
      "learning_rate": 4.1117531862649096e-05,
      "loss": 2.0025,
      "step": 196
    },
    {
      "epoch": 0.8277310924369747,
      "grad_norm": 0.17056453227996826,
      "learning_rate": 4.092316786490213e-05,
      "loss": 1.9113,
      "step": 197
    },
    {
      "epoch": 0.8319327731092437,
      "grad_norm": 0.33667296171188354,
      "learning_rate": 4.072827384562277e-05,
      "loss": 1.808,
      "step": 198
    },
    {
      "epoch": 0.8361344537815126,
      "grad_norm": 0.19670777022838593,
      "learning_rate": 4.0532859261593916e-05,
      "loss": 2.0879,
      "step": 199
    },
    {
      "epoch": 0.8403361344537815,
      "grad_norm": 0.19230350852012634,
      "learning_rate": 4.033693359485763e-05,
      "loss": 2.0093,
      "step": 200
    },
    {
      "epoch": 0.8445378151260504,
      "grad_norm": 0.14327159523963928,
      "learning_rate": 4.014050635225508e-05,
      "loss": 1.8698,
      "step": 201
    },
    {
      "epoch": 0.8487394957983193,
      "grad_norm": 0.16874836385250092,
      "learning_rate": 3.994358706496529e-05,
      "loss": 2.1746,
      "step": 202
    },
    {
      "epoch": 0.8529411764705882,
      "grad_norm": 0.2902648448944092,
      "learning_rate": 3.9746185288042574e-05,
      "loss": 1.6873,
      "step": 203
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.14672698080539703,
      "learning_rate": 3.954831059995296e-05,
      "loss": 2.0229,
      "step": 204
    },
    {
      "epoch": 0.8613445378151261,
      "grad_norm": 0.17595812678337097,
      "learning_rate": 3.9349972602109404e-05,
      "loss": 1.8638,
      "step": 205
    },
    {
      "epoch": 0.865546218487395,
      "grad_norm": 0.17806874215602875,
      "learning_rate": 3.9151180918405886e-05,
      "loss": 1.8911,
      "step": 206
    },
    {
      "epoch": 0.8697478991596639,
      "grad_norm": 0.16666433215141296,
      "learning_rate": 3.895194519475046e-05,
      "loss": 1.912,
      "step": 207
    },
    {
      "epoch": 0.8739495798319328,
      "grad_norm": 0.1936957836151123,
      "learning_rate": 3.8752275098597186e-05,
      "loss": 1.8171,
      "step": 208
    },
    {
      "epoch": 0.8781512605042017,
      "grad_norm": 0.1712755411863327,
      "learning_rate": 3.8552180318477055e-05,
      "loss": 2.0602,
      "step": 209
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 0.16776567697525024,
      "learning_rate": 3.8351670563527854e-05,
      "loss": 1.6852,
      "step": 210
    },
    {
      "epoch": 0.8865546218487395,
      "grad_norm": 0.15940265357494354,
      "learning_rate": 3.815075556302309e-05,
      "loss": 1.7701,
      "step": 211
    },
    {
      "epoch": 0.8907563025210085,
      "grad_norm": 0.2017056941986084,
      "learning_rate": 3.794944506589986e-05,
      "loss": 1.6285,
      "step": 212
    },
    {
      "epoch": 0.8949579831932774,
      "grad_norm": 0.1665181964635849,
      "learning_rate": 3.774774884028582e-05,
      "loss": 1.8832,
      "step": 213
    },
    {
      "epoch": 0.8991596638655462,
      "grad_norm": 0.1938188076019287,
      "learning_rate": 3.754567667302521e-05,
      "loss": 1.7577,
      "step": 214
    },
    {
      "epoch": 0.9033613445378151,
      "grad_norm": 0.1959097981452942,
      "learning_rate": 3.7343238369203985e-05,
      "loss": 1.7827,
      "step": 215
    },
    {
      "epoch": 0.907563025210084,
      "grad_norm": 0.15705636143684387,
      "learning_rate": 3.714044375167403e-05,
      "loss": 1.8884,
      "step": 216
    },
    {
      "epoch": 0.9117647058823529,
      "grad_norm": 0.21525953710079193,
      "learning_rate": 3.693730266057653e-05,
      "loss": 1.9235,
      "step": 217
    },
    {
      "epoch": 0.9159663865546218,
      "grad_norm": 0.16254454851150513,
      "learning_rate": 3.67338249528645e-05,
      "loss": 1.9587,
      "step": 218
    },
    {
      "epoch": 0.9201680672268907,
      "grad_norm": 0.13950613141059875,
      "learning_rate": 3.6530020501824465e-05,
      "loss": 2.0398,
      "step": 219
    },
    {
      "epoch": 0.9243697478991597,
      "grad_norm": 0.171157568693161,
      "learning_rate": 3.6325899196597493e-05,
      "loss": 1.8841,
      "step": 220
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 0.1359075903892517,
      "learning_rate": 3.612147094169921e-05,
      "loss": 2.1739,
      "step": 221
    },
    {
      "epoch": 0.9327731092436975,
      "grad_norm": 0.22331862151622772,
      "learning_rate": 3.591674565653926e-05,
      "loss": 2.0195,
      "step": 222
    },
    {
      "epoch": 0.9369747899159664,
      "grad_norm": 0.16365067660808563,
      "learning_rate": 3.571173327494005e-05,
      "loss": 1.9375,
      "step": 223
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 0.24077700078487396,
      "learning_rate": 3.5506443744654644e-05,
      "loss": 1.8117,
      "step": 224
    },
    {
      "epoch": 0.9453781512605042,
      "grad_norm": 0.1790030151605606,
      "learning_rate": 3.53008870268841e-05,
      "loss": 1.5849,
      "step": 225
    },
    {
      "epoch": 0.9495798319327731,
      "grad_norm": 0.20200079679489136,
      "learning_rate": 3.5095073095794156e-05,
      "loss": 2.0211,
      "step": 226
    },
    {
      "epoch": 0.9537815126050421,
      "grad_norm": 0.17911580204963684,
      "learning_rate": 3.4889011938031216e-05,
      "loss": 1.8153,
      "step": 227
    },
    {
      "epoch": 0.957983193277311,
      "grad_norm": 0.23249390721321106,
      "learning_rate": 3.4682713552237795e-05,
      "loss": 1.74,
      "step": 228
    },
    {
      "epoch": 0.9621848739495799,
      "grad_norm": 0.17647947371006012,
      "learning_rate": 3.447618794856735e-05,
      "loss": 1.8374,
      "step": 229
    },
    {
      "epoch": 0.9663865546218487,
      "grad_norm": 0.1574244350194931,
      "learning_rate": 3.426944514819856e-05,
      "loss": 2.0606,
      "step": 230
    },
    {
      "epoch": 0.9705882352941176,
      "grad_norm": 0.17625559866428375,
      "learning_rate": 3.406249518284907e-05,
      "loss": 1.841,
      "step": 231
    },
    {
      "epoch": 0.9747899159663865,
      "grad_norm": 0.16222096979618073,
      "learning_rate": 3.385534809428875e-05,
      "loss": 1.9827,
      "step": 232
    },
    {
      "epoch": 0.9789915966386554,
      "grad_norm": 0.18522681295871735,
      "learning_rate": 3.3648013933852384e-05,
      "loss": 1.8832,
      "step": 233
    },
    {
      "epoch": 0.9831932773109243,
      "grad_norm": 0.1793953776359558,
      "learning_rate": 3.3440502761952016e-05,
      "loss": 1.9176,
      "step": 234
    },
    {
      "epoch": 0.9873949579831933,
      "grad_norm": 0.16444647312164307,
      "learning_rate": 3.323282464758874e-05,
      "loss": 1.8963,
      "step": 235
    },
    {
      "epoch": 0.9915966386554622,
      "grad_norm": 0.1859455704689026,
      "learning_rate": 3.302498966786418e-05,
      "loss": 1.9036,
      "step": 236
    },
    {
      "epoch": 0.9957983193277311,
      "grad_norm": 0.16911397874355316,
      "learning_rate": 3.281700790749145e-05,
      "loss": 1.7915,
      "step": 237
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1729690283536911,
      "learning_rate": 3.260888945830589e-05,
      "loss": 1.8512,
      "step": 238
    },
    {
      "epoch": 1.004201680672269,
      "grad_norm": 0.19748295843601227,
      "learning_rate": 3.240064441877533e-05,
      "loss": 2.0865,
      "step": 239
    },
    {
      "epoch": 1.004201680672269,
      "grad_norm": 0.2490590363740921,
      "learning_rate": 3.219228289351011e-05,
      "loss": 1.6512,
      "step": 240
    }
  ],
  "logging_steps": 1,
  "max_steps": 476,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 60,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4225750429270016e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}