{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.2422360248447206,
  "eval_steps": 500,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012422360248447204,
      "grad_norm": 0.00018831038323696703,
      "learning_rate": 6.060606060606061e-06,
      "loss": 2.4276,
      "step": 2
    },
    {
      "epoch": 0.024844720496894408,
      "grad_norm": 0.00010692311479942873,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 2.2565,
      "step": 4
    },
    {
      "epoch": 0.037267080745341616,
      "grad_norm": 6.173110887175426e-05,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 1.9191,
      "step": 6
    },
    {
      "epoch": 0.049689440993788817,
      "grad_norm": 4.326488488004543e-05,
      "learning_rate": 4.242424242424243e-05,
      "loss": 1.7997,
      "step": 8
    },
    {
      "epoch": 0.062111801242236024,
      "grad_norm": 3.188664049957879e-05,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 1.6296,
      "step": 10
    },
    {
      "epoch": 0.07453416149068323,
      "grad_norm": 2.6054423869936727e-05,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.5767,
      "step": 12
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 2.7649846742860973e-05,
      "learning_rate": 7.878787878787879e-05,
      "loss": 1.5131,
      "step": 14
    },
    {
      "epoch": 0.09937888198757763,
      "grad_norm": 1.6101763321785256e-05,
      "learning_rate": 9.090909090909092e-05,
      "loss": 1.4497,
      "step": 16
    },
    {
      "epoch": 0.11180124223602485,
      "grad_norm": 2.3354801669484004e-05,
      "learning_rate": 0.00010303030303030303,
      "loss": 1.4353,
      "step": 18
    },
    {
      "epoch": 0.12422360248447205,
      "grad_norm": 1.531304769741837e-05,
      "learning_rate": 0.00011515151515151516,
      "loss": 1.3655,
      "step": 20
    },
    {
      "epoch": 0.13664596273291926,
      "grad_norm": 9.777257218956947e-06,
      "learning_rate": 0.00012727272727272728,
      "loss": 1.3691,
      "step": 22
    },
    {
      "epoch": 0.14906832298136646,
      "grad_norm": 1.5149210412346292e-05,
      "learning_rate": 0.0001393939393939394,
      "loss": 1.3916,
      "step": 24
    },
    {
      "epoch": 0.16149068322981366,
      "grad_norm": 1.1748486940632574e-05,
      "learning_rate": 0.00015151515151515152,
      "loss": 1.341,
      "step": 26
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 3.532869595801458e-05,
      "learning_rate": 0.00016363636363636366,
      "loss": 1.3409,
      "step": 28
    },
    {
      "epoch": 0.18633540372670807,
      "grad_norm": 7.877199095673859e-06,
      "learning_rate": 0.00017575757575757578,
      "loss": 1.3728,
      "step": 30
    },
    {
      "epoch": 0.19875776397515527,
      "grad_norm": 1.1650959095277358e-05,
      "learning_rate": 0.0001878787878787879,
      "loss": 1.3579,
      "step": 32
    },
    {
      "epoch": 0.2111801242236025,
      "grad_norm": 3.298332740087062e-05,
      "learning_rate": 0.0002,
      "loss": 1.3835,
      "step": 34
    },
    {
      "epoch": 0.2236024844720497,
      "grad_norm": 9.103293450607453e-06,
      "learning_rate": 0.00019999471258384815,
      "loss": 1.3517,
      "step": 36
    },
    {
      "epoch": 0.2360248447204969,
      "grad_norm": 1.055269422067795e-05,
      "learning_rate": 0.000199978850894528,
      "loss": 1.3748,
      "step": 38
    },
    {
      "epoch": 0.2484472049689441,
      "grad_norm": 1.8264010577695444e-05,
      "learning_rate": 0.0001999524166093866,
      "loss": 1.383,
      "step": 40
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 9.094917004404124e-06,
      "learning_rate": 0.00019991541252380526,
      "loss": 1.4102,
      "step": 42
    },
    {
      "epoch": 0.2732919254658385,
      "grad_norm": 1.286895803787047e-05,
      "learning_rate": 0.00019986784255090397,
      "loss": 1.3811,
      "step": 44
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.386065014230553e-05,
      "learning_rate": 0.0001998097117211276,
      "loss": 1.3794,
      "step": 46
    },
    {
      "epoch": 0.2981366459627329,
      "grad_norm": 8.020670065889135e-06,
      "learning_rate": 0.00019974102618171394,
      "loss": 1.3584,
      "step": 48
    },
    {
      "epoch": 0.3105590062111801,
      "grad_norm": 8.217157301260158e-06,
      "learning_rate": 0.00019966179319604355,
      "loss": 1.3592,
      "step": 50
    },
    {
      "epoch": 0.32298136645962733,
      "grad_norm": 1.1515216101543047e-05,
      "learning_rate": 0.00019957202114287187,
      "loss": 1.3679,
      "step": 52
    },
    {
      "epoch": 0.33540372670807456,
      "grad_norm": 8.87443184183212e-06,
      "learning_rate": 0.0001994717195154429,
      "loss": 1.3779,
      "step": 54
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 7.773131073918194e-06,
      "learning_rate": 0.00019936089892048556,
      "loss": 1.3634,
      "step": 56
    },
    {
      "epoch": 0.36024844720496896,
      "grad_norm": 8.845816410030238e-06,
      "learning_rate": 0.00019923957107709195,
      "loss": 1.3312,
      "step": 58
    },
    {
      "epoch": 0.37267080745341613,
      "grad_norm": 8.347928087459877e-06,
      "learning_rate": 0.000199107748815478,
      "loss": 1.3626,
      "step": 60
    },
    {
      "epoch": 0.38509316770186336,
      "grad_norm": 9.047604180523194e-06,
      "learning_rate": 0.0001989654460756269,
      "loss": 1.3653,
      "step": 62
    },
    {
      "epoch": 0.39751552795031053,
      "grad_norm": 8.44144778966438e-06,
      "learning_rate": 0.00019881267790581466,
      "loss": 1.3712,
      "step": 64
    },
    {
      "epoch": 0.40993788819875776,
      "grad_norm": 8.453951522824354e-06,
      "learning_rate": 0.0001986494604610191,
      "loss": 1.3027,
      "step": 66
    },
    {
      "epoch": 0.422360248447205,
      "grad_norm": 7.831426046323031e-06,
      "learning_rate": 0.00019847581100121127,
      "loss": 1.3737,
      "step": 68
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 8.356658327102195e-06,
      "learning_rate": 0.00019829174788953038,
      "loss": 1.3522,
      "step": 70
    },
    {
      "epoch": 0.4472049689440994,
      "grad_norm": 8.671122486703098e-06,
      "learning_rate": 0.0001980972905903418,
      "loss": 1.3808,
      "step": 72
    },
    {
      "epoch": 0.45962732919254656,
      "grad_norm": 9.771009899850469e-06,
      "learning_rate": 0.00019789245966717883,
      "loss": 1.3695,
      "step": 74
    },
    {
      "epoch": 0.4720496894409938,
      "grad_norm": 6.7302303250471596e-06,
      "learning_rate": 0.00019767727678056805,
      "loss": 1.3754,
      "step": 76
    },
    {
      "epoch": 0.484472049689441,
      "grad_norm": 9.788966963242274e-06,
      "learning_rate": 0.00019745176468573893,
      "loss": 1.379,
      "step": 78
    },
    {
      "epoch": 0.4968944099378882,
      "grad_norm": 9.375480658491142e-06,
      "learning_rate": 0.00019721594723021732,
      "loss": 1.3484,
      "step": 80
    },
    {
      "epoch": 0.5093167701863354,
      "grad_norm": 3.550610927050002e-05,
      "learning_rate": 0.00019696984935130364,
      "loss": 1.377,
      "step": 82
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 9.707620847621001e-06,
      "learning_rate": 0.00019671349707343593,
      "loss": 1.3587,
      "step": 84
    },
    {
      "epoch": 0.5341614906832298,
      "grad_norm": 1.0621732144500129e-05,
      "learning_rate": 0.00019644691750543767,
      "loss": 1.393,
      "step": 86
    },
    {
      "epoch": 0.546583850931677,
      "grad_norm": 9.491611308476422e-06,
      "learning_rate": 0.0001961701388376511,
      "loss": 1.3715,
      "step": 88
    },
    {
      "epoch": 0.5590062111801242,
      "grad_norm": 8.982805411505979e-06,
      "learning_rate": 0.00019588319033895623,
      "loss": 1.3829,
      "step": 90
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.0132759598491248e-05,
      "learning_rate": 0.0001955861023536756,
      "loss": 1.3395,
      "step": 92
    },
    {
      "epoch": 0.5838509316770186,
      "grad_norm": 1.3477648280968424e-05,
      "learning_rate": 0.0001952789062983654,
      "loss": 1.3908,
      "step": 94
    },
    {
      "epoch": 0.5962732919254659,
      "grad_norm": 1.1435187843744643e-05,
      "learning_rate": 0.0001949616346584934,
      "loss": 1.3799,
      "step": 96
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 1.273245652555488e-05,
      "learning_rate": 0.00019463432098500337,
      "loss": 1.3973,
      "step": 98
    },
    {
      "epoch": 0.6211180124223602,
      "grad_norm": 1.1162846021761652e-05,
      "learning_rate": 0.00019429699989076746,
      "loss": 1.3564,
      "step": 100
    },
    {
      "epoch": 0.6335403726708074,
      "grad_norm": 1.1645292943285313e-05,
      "learning_rate": 0.00019394970704692566,
      "loss": 1.3535,
      "step": 102
    },
    {
      "epoch": 0.6459627329192547,
      "grad_norm": 3.541897240211256e-05,
      "learning_rate": 0.00019359247917911384,
      "loss": 1.3813,
      "step": 104
    },
    {
      "epoch": 0.6583850931677019,
      "grad_norm": 9.2498794401763e-06,
      "learning_rate": 0.00019322535406358,
      "loss": 1.3617,
      "step": 106
    },
    {
      "epoch": 0.6708074534161491,
      "grad_norm": 8.978020559879951e-06,
      "learning_rate": 0.00019284837052318933,
      "loss": 1.3813,
      "step": 108
    },
    {
      "epoch": 0.6832298136645962,
      "grad_norm": 1.3010416296310723e-05,
      "learning_rate": 0.00019246156842331918,
      "loss": 1.3726,
      "step": 110
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 1.2821891687053721e-05,
      "learning_rate": 0.00019206498866764288,
      "loss": 1.4042,
      "step": 112
    },
    {
      "epoch": 0.7080745341614907,
      "grad_norm": 9.70847668213537e-06,
      "learning_rate": 0.00019165867319380456,
      "loss": 1.4061,
      "step": 114
    },
    {
      "epoch": 0.7204968944099379,
      "grad_norm": 1.6388959920732304e-05,
      "learning_rate": 0.0001912426649689842,
      "loss": 1.3668,
      "step": 116
    },
    {
      "epoch": 0.7329192546583851,
      "grad_norm": 1.66231893672375e-05,
      "learning_rate": 0.00019081700798535397,
      "loss": 1.3712,
      "step": 118
    },
    {
      "epoch": 0.7453416149068323,
      "grad_norm": 1.864772457338404e-05,
      "learning_rate": 0.00019038174725542604,
      "loss": 1.3849,
      "step": 120
    },
    {
      "epoch": 0.7577639751552795,
      "grad_norm": 1.4855336303298827e-05,
      "learning_rate": 0.0001899369288072927,
      "loss": 1.4006,
      "step": 122
    },
    {
      "epoch": 0.7701863354037267,
      "grad_norm": 1.8388282114756294e-05,
      "learning_rate": 0.00018948259967975888,
      "loss": 1.4008,
      "step": 124
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 2.689829489099793e-05,
      "learning_rate": 0.00018901880791736793,
      "loss": 1.3808,
      "step": 126
    },
    {
      "epoch": 0.7950310559006211,
      "grad_norm": 2.5062678105314262e-05,
      "learning_rate": 0.000188545602565321,
      "loss": 1.3948,
      "step": 128
    },
    {
      "epoch": 0.8074534161490683,
      "grad_norm": 2.1250931240501814e-05,
      "learning_rate": 0.0001880630336642905,
      "loss": 1.402,
      "step": 130
    },
    {
      "epoch": 0.8198757763975155,
      "grad_norm": 2.403008875262458e-05,
      "learning_rate": 0.0001875711522451284,
      "loss": 1.384,
      "step": 132
    },
    {
      "epoch": 0.8322981366459627,
      "grad_norm": 2.597655839053914e-05,
      "learning_rate": 0.00018707001032347,
      "loss": 1.3541,
      "step": 134
    },
    {
      "epoch": 0.84472049689441,
      "grad_norm": 2.3845692339818925e-05,
      "learning_rate": 0.0001865596608942331,
      "loss": 1.396,
      "step": 136
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.8721084416029043e-05,
      "learning_rate": 0.00018604015792601396,
      "loss": 1.3427,
      "step": 138
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 3.2439467759104446e-05,
      "learning_rate": 0.0001855115563553803,
      "loss": 1.376,
      "step": 140
    },
    {
      "epoch": 0.8819875776397516,
      "grad_norm": 2.3974960640771315e-05,
      "learning_rate": 0.0001849739120810618,
      "loss": 1.4019,
      "step": 142
    },
    {
      "epoch": 0.8944099378881988,
      "grad_norm": 3.250224108342081e-05,
      "learning_rate": 0.00018442728195803881,
      "loss": 1.3915,
      "step": 144
    },
    {
      "epoch": 0.906832298136646,
      "grad_norm": 2.820833833538927e-05,
      "learning_rate": 0.0001838717237915302,
      "loss": 1.3943,
      "step": 146
    },
    {
      "epoch": 0.9192546583850931,
      "grad_norm": 4.0885188354877755e-05,
      "learning_rate": 0.00018330729633088045,
      "loss": 1.3818,
      "step": 148
    },
    {
      "epoch": 0.9316770186335404,
      "grad_norm": 5.243903797236271e-05,
      "learning_rate": 0.00018273405926334696,
      "loss": 1.4112,
      "step": 150
    },
    {
      "epoch": 0.9440993788819876,
      "grad_norm": 4.928431008011103e-05,
      "learning_rate": 0.0001821520732077883,
      "loss": 4.4804,
      "step": 152
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 7.073425513226539e-05,
      "learning_rate": 0.00018156139970825391,
      "loss": 1.4202,
      "step": 154
    },
    {
      "epoch": 0.968944099378882,
      "grad_norm": 4.9798843974713236e-05,
      "learning_rate": 0.00018096210122747584,
      "loss": 1.4218,
      "step": 156
    },
    {
      "epoch": 0.9813664596273292,
      "grad_norm": 6.847319309599698e-05,
      "learning_rate": 0.0001803542411402634,
      "loss": 1.401,
      "step": 158
    },
    {
      "epoch": 0.9937888198757764,
      "grad_norm": 5.58948922844138e-05,
      "learning_rate": 0.0001797378837268015,
      "loss": 1.3425,
      "step": 160
    },
    {
      "epoch": 1.0062111801242235,
      "grad_norm": 4.716762123280205e-05,
      "learning_rate": 0.00017911309416585296,
      "loss": 1.3779,
      "step": 162
    },
    {
      "epoch": 1.0186335403726707,
      "grad_norm": 5.352822699933313e-05,
      "learning_rate": 0.0001784799385278661,
      "loss": 1.341,
      "step": 164
    },
    {
      "epoch": 1.031055900621118,
      "grad_norm": 5.088459511171095e-05,
      "learning_rate": 0.0001778384837679879,
      "loss": 1.4021,
      "step": 166
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 5.318582770996727e-05,
      "learning_rate": 0.00017718879771898348,
      "loss": 1.3597,
      "step": 168
    },
    {
      "epoch": 1.0559006211180124,
      "grad_norm": 4.925034227198921e-05,
      "learning_rate": 0.00017653094908406301,
      "loss": 1.3425,
      "step": 170
    },
    {
      "epoch": 1.0683229813664596,
      "grad_norm": 5.734206206398085e-05,
      "learning_rate": 0.00017586500742961653,
      "loss": 1.389,
      "step": 172
    },
    {
      "epoch": 1.0807453416149069,
      "grad_norm": 5.8051424275618047e-05,
      "learning_rate": 0.00017519104317785717,
      "loss": 1.3776,
      "step": 174
    },
    {
      "epoch": 1.093167701863354,
      "grad_norm": 4.752572931465693e-05,
      "learning_rate": 0.00017450912759937434,
      "loss": 1.4045,
      "step": 176
    },
    {
      "epoch": 1.1055900621118013,
      "grad_norm": 4.9184389354195446e-05,
      "learning_rate": 0.00017381933280559693,
      "loss": 2.4013,
      "step": 178
    },
    {
      "epoch": 1.1180124223602483,
      "grad_norm": 6.995958392508328e-05,
      "learning_rate": 0.00017312173174116762,
      "loss": 1.6484,
      "step": 180
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 4.960764272254892e-05,
      "learning_rate": 0.0001724163981762291,
      "loss": 1.3301,
      "step": 182
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 4.448357503861189e-05,
      "learning_rate": 0.0001717034066986231,
      "loss": 1.4058,
      "step": 184
    },
    {
      "epoch": 1.15527950310559,
      "grad_norm": 4.6940349420765415e-05,
      "learning_rate": 0.00017098283270600267,
      "loss": 1.3731,
      "step": 186
    },
    {
      "epoch": 1.1677018633540373,
      "grad_norm": 4.82694485981483e-05,
      "learning_rate": 0.00017025475239785919,
      "loss": 1.3586,
      "step": 188
    },
    {
      "epoch": 1.1801242236024845,
      "grad_norm": 4.615723810275085e-05,
      "learning_rate": 0.00016951924276746425,
      "loss": 1.3777,
      "step": 190
    },
    {
      "epoch": 1.1925465838509317,
      "grad_norm": 4.163803532719612e-05,
      "learning_rate": 0.00016877638159372782,
      "loss": 1.3974,
      "step": 192
    },
    {
      "epoch": 1.204968944099379,
      "grad_norm": 4.0616945625515655e-05,
      "learning_rate": 0.00016802624743297333,
      "loss": 1.41,
      "step": 194
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 4.708253254648298e-05,
      "learning_rate": 0.00016726891961063028,
      "loss": 1.3499,
      "step": 196
    },
    {
      "epoch": 1.2298136645962732,
      "grad_norm": 4.392105984152295e-05,
      "learning_rate": 0.00016650447821284594,
      "loss": 1.3619,
      "step": 198
    },
    {
      "epoch": 1.2422360248447206,
      "grad_norm": 4.838509266846813e-05,
      "learning_rate": 0.00016573300407801616,
      "loss": 1.3634,
      "step": 200
    }
  ],
  "logging_steps": 2,
  "max_steps": 644,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.8994607076972954e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}